======================================================================= 1 === Date: Tue, 1 Mar 1994 02:35:12 GMT From: zmbenhal@netcom.com (Zeyd M. Ben-Halim) Subject: Re: Trestle problems under Linux 0.99pl15 In article <1994Feb26.125121.8435@infodev.cam.ac.uk> rrw1000@cus.cam.ac.uk (Ric hard Watts) writes: > Hello everyone :-). I'm having a bit of trouble with Trestle (and hence >FormsVBT) under Linux 0.99pl15. Trestle and FormsVBT compile without >errors, and so do programs which use them (I've tried 'Cards', >'Hello Trestle', 'Calculator' and 'Cube'). > > When I try to run these programs, however, they run successfully for > anything from a few to 20 or more seconds, and then crash with a 'Virtual >time alarm' error. This (or so gdb and the kernel source tell me) is due >to their being sent signal 26 (Virtual time alarm). If I insert a signal >handler for SIGVTALRM, Cards and Hello Trestle work fine - I haven't tried >with Calculator and Cube because they take ages to build on my platform. > > I assume that what's happening is that Trestle is setting a timeout, >expecting to recieve SIGALRM or something, and is instead recieving what >it expects, is getting SIGVTALRM. This is fixable by inserting a null >SIGVTALRM handler in whatever package under Trestle (ui, X11R4 or tcp) >sets the alarm, but since the alarm is probably supposed to be doing >something, that sounds like a silly thing to do. > > I'd be grateful if someone could tell me what incredibly obvious and >stupid thing I'm doing wrong, or provide some sort of bug fix ? Are you sure that there is no SIGVTALRM handlers in Trestle or M3 itself? The most common problem with uncaught signals under Linux is that signal() has SYSV semantics and NOT the BSD semantics most people expect. This reulst in the signal handler being reset to the default after it's first occurance. I've yet to see an offical release of M3 for Linux, so the chances are that problems like that may still appear. > > Thanks in advance, > > > >Richard. > -- --- Zeyd M. Ben-Halim zmbenhal@netcom.com 10479 1/4 Santa Monica Blvd, LA, CA, 90025 (310) 470-0281 ======================================================================= 2 === Date: Tue, 1 Mar 1994 08:49:30 GMT From: qs101@pelican.cl.cam.ac.uk (Quentin Stafford-Fraser) Subject: M3 3.1 on SPARCs - Questions I now have most of 3.1 up and running on my SPARC - many thanks to all who made it possible. Overall the system looks like a considerable improvement, though I have a few gripes (What happened to ParseParams? List.Push()? ). m3build is wonderful. * Is there a document anywhere that describes the changes to the interfaces? Or an index to them? * Was it a policy decision that you should have to explicitly import libm3 in m3makefiles? There can't really be many cases where you don't need it. Many thanks, Quentin -- ---------------------------------------------------------------------- Quentin Stafford-Fraser Cambridge University Computer Lab Rank Xerox Cambridge EuroPARC qs101@cl.cam.ac.uk fraser@europarc.xerox.com Tel: +44 223 334411 Tel: +44 223 341521 Fax: +44 223 334679 Fax: +44 223 341510 http://pelican.cl.cam.ac.uk/people/qs101/me.html ---------------------------------------------------------------------- ======================================================================= 3 === Date: Tue, 1 Mar 1994 15:05:18 GMT From: dagenais@froh.vlsi.polymtl.ca (Michel Dagenais) Subject: Re: M3 3.1 on SPARCs - Questions Overall the system looks like a considerable improvement, though I have a few gripes (What happened to ParseParams? List.Push()? ). The interfaces in libm3 have been cleaned up. IMHO the interface police has done a good job. Data structures (lists...) were converted to generic modules, OS neutral interfaces were provided to manipulate files and a few things have been made more coherent. List.Push is an example, this function destructively alters a list received as argument. It is easily replaced: List.Push(\([^,]*\),\([^,]*\)) by: \1 := List.Cons(\2,\1) . (i.e. List.Push(list,item) -> list := List.Cons(item,list) ). * Is there a document anywhere that describes the changes to the interfaces? Or an index to them? I have seen the draft of a nicely organized document. I presume it will come out fairly soon. The info is already stored in the interfaces and can be extracted with m3totex although you may be missing the glue that ties all this info together. * Was it a policy decision that you should have to explicitly import libm3 in m3makefiles? There can't really be many cases where you don't need it. I was under the impression that it is not intentional and would be fixed with some "quake goo", as they say, in the next release. -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 4 === Date: Tue, 1 Mar 1994 16:43:55 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: M3 3.1 on SPARCs - Questions In article qs101@pelican.cl.cam.ac.uk (Quentin Stafford-Fraser) writes: I now have most of 3.1 up and running on my SPARC - many thanks to all who made it possible. Overall the system looks like a considerable improvement, though I have a few gripes (What happened to ParseParams? List.Push()? ). m3build is wonderful. We also had some existing code that heavily used the old List interface and a lot of its features. One thing to note is that you can recompile the old List interface with the new compiler since it only counts on Text and Thread. Sx has also changed a lot. I believe many of changes were due to SRC's interface police inspecting the libm3 libraries. I agree that m3build is really nice to use. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 5 === Date: 1 Mar 94 21:05:04 GMT From: rro@CS.ColoState.EDU (Rod Oldehoeft) Subject: Re: Where to start In article , Joe.Klemmer@f370.n109.z1.fidone t.org (Joe Klemmer) writes: |> Ok, I got the DOS port of Modula-3, I got the Book, I got a computer, |> where to start? What is a good begining for learning M3? Any suggestions? |> Tips? Pitfalls? |> |> Thanks, |> Joe If by the book you mean Harbison's, I have these observations: 1. The text approaches language features like a Swiss army knife demo. This is OK for people experienced at learning new languages, but it is confusing for my data structures class. Every esoteric option of a language feature is explained right up front, instead of showing the commonly used options first, and saving the juicy bits for later. 2. The book is extremely loaded with errors of every kind. For quite awhile I've been collecting errata and maintaining a sheet for people who want to spend a couple of hours getting the book more nearly correct. You can get this via anonymous ftp from beethoven:~ftp/pub/M3/{harbison-errs.ps.Z,harbison-errs.tex}. Machine beethoven is also known as 129.82.102.183. If, after looking over my errata sheet, you have more, please send them to me (rro@cs.colostate.edu). Another book is Nelson's "Systems Programming in Modula-3." I haven't spent a lot of time in it, but it assumes you pretty much know M3 and want to know how to do system-like things. Now to the software. The PC version from DEC hasn't received very much good press as far as I can tell, and I'm anxiously awaiting the GNU version. Can anyone provide an update on the state of this project? ======================================================================= 6 === Date: Tue, 1 Mar 1994 23:24:49 GMT From: rrw1000@cus.cam.ac.uk (Richard Watts) Subject: Re: Trestle problems under Linux 0.99pl15 In article , zmbenhal@netcom.com (Zeyd M. Ben-Hal im) writes: |> In article <1994Feb26.125121.8435@infodev.cam.ac.uk> rrw1000@cus.cam.ac.uk ( Richard Watts) writes: |> [things deleted] |> Are you sure that there is no SIGVTALRM handlers in Trestle or M3 |> itself? The most common problem with uncaught signals under Linux |> is that signal() has SYSV semantics and NOT the BSD semantics most |> people expect. This reulst in the signal handler being reset to the |> default after it's first occurance. |> |> I've yet to see an offical release of M3 for Linux, so the chances are |> that problems like that may still appear. |> Yup. Spot on - SIGVTALRM is used by the thread scheduler to, um, schedule threads. The problem stems from src/unix/linux/Usignal.i3 - the sigaction structure members are defined in the wrong order (actually, the wrong members for Linux are defined too, but this doesn't make any difference in practice). The Modula-3 definition for the sigaction structure for 0.99pl15 should be something along the lines of : struct_sigaction = RECORD sa_handler : SignalActionHandler; sa_mask : sigset_t; sa_flags : int; (* signal action flags *) (* sa_resv : ARRAY [0..1] OF int; *) sa_restorer : SignalRestoreHandler; END; After which the scheduler kicks in very nicely, thank you :-). Unfortunately, even with this fix applied, my threaded programs die nastily when they try to reschedule threads - the program : MODULE Hello EXPORTS Main; IMPORT Wr, Stdio, Thread, Fmt, Rd; TYPE ch = Thread.Closure OBJECT ms : TEXT; OVERRIDES apply := SayCheese; END; PROCEDURE SayCheese(self : ch ): REFANY RAISES {} = VAR a: TEXT; BEGIN Wr.PutText(Stdio.stdout, self.ms ); Wr.Flush(Stdio.stdout); LOOP (* Wr.PutText(Stdio.stdout, self.ms); *) END; RETURN self; END SayCheese; VAR i : INTEGER; BEGIN Wr.PutText(Stdio.stdout," Type a line : \n"); Wr.Flush(Stdio.stdout); EVAL Rd.GetLine(Stdio.stdin); FOR i := 1 TO 23 DO EVAL Thread.Fork(NEW(ch, ms := "Cheese " & Fmt.Int(i))); END; LOOP END; END Hello. Falls over in the following rather nasty way : $ LINUX/thread Type a line : gfreg Cheese 4Cheese 3Cheese 2Cheese 1Cheese 9Cheese 8Cheese 7Cheese 6Cheese 5Cheese 14 Cheese 13Cheese 12Cheese 11Cheese 10Cheese 19Cheese 18Cheese 17Cheese 16Cheese 15 Cheese 23Cheese 22Cheese 21Cheese 20Illegal instruction $ [line breaks inserted for clarity] [If I uncomment the second Wr.PutText, it segfaults instead - the default handler does kick in properly ] On having a look at the Linux-specific thread scheduling code, I can't find anything at all wrong with the context switching itself, so my current suspect is the thread stacks. I've replied to Marc Brown at DEC (who very kindly emailed me about it), so I presume some knowledgable person who actually knows how it all works will toddle along and sort it out. Meanwhile, I'd be interested to hear from anyone who can get my program past the 'Illegal instruction' stage under Linux... there's a chance it's just my setup (though I don't really see how it could be), and I'll carry on trying to hack at it in my (by now fairly limited - things have been piling up !) spare time. Thanks for all the help ! Richard. |> [things deleted] ======================================================================= 7 === Date: Wed, 2 Mar 1994 08:28:33 GMT From: kirschnt@informatik.uni-muenchen.de (Torsten R. Kirschner) Subject: Re: Where to start rro@CS.ColoState.EDU (Rod Oldehoeft) writes: [...] >Another book is Nelson's "Systems Programming in Modula-3." I haven't >spent a lot of time in it, but it assumes you pretty much know M3 and >want to know how to do system-like things. SPwM3 is, I think, a very good introduction to what you can do with SRC M3 and how to do it. It also contains good chapters about Thread programming and how Trestle, a window system for SRC M3, can be used. As far as I can tell, the "System Programming" part refers to the design of software systems implemented in SRC M3 rather than to something like "UNIX system programming". SPwM3 doesn't tell about kernel structures or something like that, and that's good so, I think. Harbison's book is still the only language reference except for the (reveised) M3 Report, SRC-52 technical report, if I remember correctly. >Now to the software. The PC version from DEC hasn't received very much >good press as far as I can tell, and I'm anxiously awaiting the GNU >version. Can anyone provide an update on the state of this project? As far as I remember, neither the SRC nor DEC as such has ever put out a/the PC version of M3. What's stored on gatekeeper.dec.com as m3pc is merely a port of the M3 language, done by someone outside SRC. As far as its quality is concerned, I have no personal experience. However, one should take into account that (a) It's a language port, not a complete implementation of the SRC M3 package. It (b) does not have thread support since that would be an interesting task in itself, I guess. After all, it's (c) free, and if you have enough DOS disk space but cannot get a UNIX for some reason it might just be your only chance to work with the M3 language on a PC. So, lacking competition, it's per se "the best there is" on PC-DOS. As far as the original question - where to start with M3 (on PC) - is concerned, one might also consider Robert Sedgewick's "Algorithms in Modula-3", Addison-Wesley. Although all the traditional algorithms are merely re-written in Modula-3 and there are no algorithms or data structures that make inventive use of M3-specific features, it is all in all a good book. Torsten -- Torsten R. Kirschner Torsten.R.Kirschner@Informatik.TU-Muenchen.DE [PGP 2.1 1024-bit key, Key ID 8A3F97, created 1993/11/14 available via finger] Torsten ======================================================================= 8 === Date: 02 Mar 94 07:25:32 -0500 From: Joe.Klemmer@f370.n109.z1.fidonet.org (Joe Klemmer) Subject: Where to start rro@CS.ColoState.EDU wrote in a message to All: |> Ok, I got the DOS port of Modula-3, I got the Book, I got a computer, |> where to start? What is a good begining for learning M3? Any suggestions? |> Tips? Pitfalls? rCE> If by the book you mean Harbison's, I have these rCE> observations: That's the book. rCE> 1. The text approaches language features like a Swiss army rCE> knife demo. This is OK for people experienced at rCE> learning new languages, but it is confusing for my data rCE> structures class. Every esoteric option of a language rCE> feature is explained right up front, instead of showing rCE> the commonly used options first, and saving the juicy rCE> bits for later. This could be confusing, thanks for pointing it out. rCE> 2. The book is extremely loaded with errors of every kind. rCE> For quite awhile I've been collecting errata and rCE> maintaining a sheet for people who want to spend a rCE> couple of hours getting the book more nearly correct. rCE> You can get this via anonymous ftp from rCE> beethoven:~ftp/pub/M3/{harbison-errs.ps.Z,harbison-errs.te rCE> }. Machine beethoven is also known as 129.82.102.183. I'll definitely go and get this. Couldn't hurt, huh? rCE> If, after looking over my errata sheet, you have more, rCE> please send them to me (rro@cs.colostate.edu). Will do. rCE> Another book is Nelson's "Systems Programming in Modula-3." rCE> I haven't spent a lot of time in it, but it assumes you rCE> pretty much know M3 and want to know how to do system-like rCE> things. I saw it when I picked up the Harbison book. There was those two and one about algorithms in Modula-3. I figured I'd get the "basic" book and maybe work my way up from therem. rCE> Now to the software. The PC version from DEC hasn't rCE> received very much good press as far as I can tell, and I'm rCE> anxiously awaiting the GNU version. Can anyone provide an rCE> update on the state of this project? Ditto. Later, Joe ======================================================================= 9 === Date: Wed, 2 Mar 1994 15:06:24 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: Where to start In article kirschnt@inform atik.uni-muenchen.de (Torsten R. Kirschner) writes: rro@CS.ColoState.EDU (Rod Oldehoeft) writes: [...] >Another book is Nelson's "Systems Programming in Modula-3." I haven't >spent a lot of time in it, but it assumes you pretty much know M3 and >want to know how to do system-like things. SPwM3 is, I think, a very good introduction to what you can do with SRC M3 and how to do it. It also contains good chapters about Thread programming and how Trestle, a window system for SRC M3, can be used. As far as I can tell, the "System Programming" part refers to the design of software systems implemented in SRC M3 rather than to something like "UNIX system programming". SPwM3 doesn't tell about kernel structures or something like that, and that's good so, I think. Harbison's book is still the only language reference except for the (reveised) M3 Report, SRC-52 technical report, if I remember correctly. I love SPwM3, but I think it is not a good book for people who are looking for a tutorial that gets them started with Modula-3. You have had to have some exposure both to Modula-3 and system implementation to appeciate it, I think. Someone annouced a textbook for Modula-3 here a while ago, but I think it will be a while before it gets published. Any chance we can get pre-released copies for review? (Though I can't read German.) I never liked "Modula-3" that much (though some of the language comparisons are nice to someone who knows those languages), but I have seen people who like it better than SPwM3 and also use it as a reference. I think it is not a good reference (possibly due to its errors) and I hate its typesetting. (Sorry, I can't cover my prejudices. Typesetting Modula in Helvetica is a _big_ mistake. Too many uppercase characters. Follow ups to comp.fonts.) The SRC references are now outdated. I think the book needs a second revision, but I am not sure if Sam is doing Modula-3 related stuff anymore. Sam, are you out there? >Now to the software. The PC version from DEC hasn't received very much >good press as far as I can tell, and I'm anxiously awaiting the GNU >version. Can anyone provide an update on the state of this project? I think your best bet right now is to run Linux on your PC. I know this is not a good option for the people with _really_ tight space or ones that don't want to deal with running UNIX, but I think this way, you can use the full-fledged Modula-3 system including FormsVBT, Obliq, Network Objects. By the way the first two of these systems are very well-introduced as SRC TRs by Marc H. Brown and Luca Cardelli. Also, don't hesitate to ask questions if you have any. This is one of the best comp.lang newsgroups because of its load and the readership, I think. (Disclaimer: I haven't run Linux myself, but those who have say it is great. I am awaiting a port to the Mac...) --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 10 === Date: 2 Mar 1994 17:46:18 GMT From: Eric Muller Subject: Modula-3 Frequently Asked Questions (FAQ) Archive-name: Modula-3-faq Last-modified: Feb 3 1994 Modula-3 Frequently Asked Questions =================================== 1. The language 1.1 What is Modula-3? 1.2 Is Modula-3 a superset of Modula-2? 2. The documentation 2.1 Where can I get a description of Modula-3? 2.2 Where can I get other information on Modula-3? 3. The implementations 3.1 Where can I get an implementation? 3.2 What is SRC Modula-3? 3.3 What is m3pc? 3.4 What is GNU Modula-3? 4. Some specific questions 4.1 Why is "Hello World" so large? 4.2 Why objects and interfaces? 4.3 What is the story with Trestle and OpenWindows? 4.4 When is the next release of SRC Modula-3 ? 5. FTP 5.1 What if I don't have ftp access? 6. Contributing 6.1 Can I contribute Modula-3 software? 1.1. What is Modula-3? Modula-3 is a systems programming language that descends from Mesa, Modula-2, Cedar, and Modula-2+. It also resembles its cousins Object Pascal, Oberon, and Euclid. The goal of Modula-3 is to be as simple and safe as it can be while meeting the needs of modern systems programmers. Instead of exploring new features, we studied the features of the Modula family of languages that have proven themselves in practice and tried to simplify them into a harmonious language. We found that most of the successful features were aimed at one of two main goals: greater robustness, and a simpler, more systematic type system. Modula-3 retains one of Modula-2's most successful features, the provision for explicit interfaces between modules. It adds objects and classes, exception handling, garbage collection, lightweight processes (or threads), and the isolation of unsafe features. 1.2. Is Modula-3 a superset of Modula-2? No; valid Modula-2 programs are not valid Modula-3 programs. However, there is a tool to help convert Modula-2 programs to Modula-3. 2.1. Where can I get a description of Modula-3? The definition of Modula-3 is contained in: System Programming with Modula-3 Edited by Greg Nelson Prentice Hall Series in Innovative Technology ISBN 0-13-590464-1 L.C. QA76.66.S87 1991 also known as SPwM3. Here is the table of contents: 1. Introduction 2. Language Definition 3. Standard Interfaces 4. An Introduction to Programming with Threads 5. Thread Synchronization: A Formal Specification 6. I/O Streams: Abstract Types, Real Programs 7. Trestle Window System Tutorial 8. How the Language Got its Spots Chapters 2 and 3 have been reprinted in Sigplan Notices, Volume 27, Number 8, August 1992, pp 15-42. Sam Harbison has written a more tutorial book about Modula3: Modula-3 Samuel P. Harbison Prentice Hall, 1992 ISBN 0-13-596396-6 The errata sheet is available via anonymous ftp from gatekeeper.dec.com in pub/DEC/Modula-3/errata. 2.2. Where can I get other information on Modula-3? There is a Usenet newsgroup, comp.lang.modula3. The archives of that group are available via anonymous ftp from gatekeeper.dec.com in pub/DEC/Modula-3/comp.lang.modula3. If you do not have access to Usenet, there is a relay mailing list; send a message to m3-request@src.dec.com to be added to it. There are a couple high-level overview articles available: "Modula-3", Sam Harbison, Byte, Vol. 15, No. 12, November 1990, pp 385+. "Safe Programming with Modula-3", Sam Harbison, Dr. Dobb's Journal, Vol. 17, No. 10, October 1992, pp 88+. A description of the Modula-3 type system is in "The Modula-3 Type System", Luca Cardelli, Jim Donahue, Mick Jordan, Bill Kalsow, Greg Nelson, Conference Record of the Sixteenth Annual ACM Symposium on Principles of Programming Languages (POPL), Austin Texas, January 11-13 1989, pp 202-212. The Trestle window system toolkit, higher-level FormsVBT toolkit, and Zeus animation system available with Modula-3, are documented in the following reports: "Trestle Reference Manual", Mark S. Manasse and Greg Nelson, SRC Research Report 68, December 1991. "Trestle Tutorial", Mark S. Manasse and Greg Nelson, SRC Research Report 69, May 1, 1992. "VBTkit Reference Manual: A toolkit for Trestle", edited by Marc H. Brown and James R. Meehan. (soon to be a SRC Research Report) A draft version is available via anonymous FTP from gatekeeper.dec.com in pub/DEC/Modula-3/contrib/vbtkit.25Mar93.ps.Z. "The FormsVBT Reference Manual", Marc H. Brown and James R. Meehan, (soon to be a SRC Research Report). A draft version is available via anonymous FTP from gatekeeper.dec.com in pub/DEC/Modula-3/contrib/formsvbt.25Mar93.ps.Z and pub/DEC/Modula-3/contrib/formsvbt.AppC.26Mar93.ps.Z. "Zeus: A System for Algorithm Animation and Multi-View Editing", Marc H. Brown, SRC Research Report 75, February 28, 1992. Available via anonymous FTP from gatekeeper.dec.com in pub/DEC/SRC/research-reports/SRC-075*.ps.Z. "Color and Sound in Algorithm Animation", Marc H. Brown and John Hershberger, SRC Research Report 76a, August 30, 1991. Available via anonymous FTP from gatekeeper.dec.com in pub/DEC/SRC/research-reports/SRC-076a*.ps.Z. "The 1992 SRC Algorithm Animation Festival", Marc H. Brown, SRC Research Report 98, March 27, 1993. Available via anonymous ftp from gatekeeper.dec.com in pub/DEC/SRC/research-reports/SRC-098*.ps.Z. Hardcopy versions of these reports can be ordered by e-mail; send your request including a postal mail address to src-reports@src.dec.com. Sedgewick's classic text on computer algorithms is presented in Modula-3 in: Alogorithms in Modula-3 Robert Sedgewick Addison-Wesley, 1993 ISBN 0-201-53351-0 3.1. Where can I get an implementation? Two implementations are available, SRC Modula-3 and a PC version of it (m3pc). Work is also progressing on GNU Modula-3. As far as we know, implementations are not available for VMS, Macintosh, or Alpha AXP/OSF. 3.2. What is SRC Modula-3? SRC Modula-3 was built by the DEC Systems Reseach Center and is available via anonymous ftp from gatekeeper.dec.com in pub/DEC/Modula-3/release. The current version, 2.11, implements the language defined in SPwM3. There are versions for the following machines: AIX386 IBM PC running AIX/PS2, AP3000 Apollo DN4500 running Domain/OS ARM Acorn R260 running RISC iX 1.21 DS3100 DECstation 3100 and 5000 running Ultrix 4.0 and 4.2 HP300 HP 9000/300 running HP-UX 8.0 HPPA HP 700/800 running HP-UX 8.0 IBMR2 IBM R6000 running AIX 3.1, IBMRT IBM RT running IBM/4.3, NEXT NeXT running ?? OKI Okidata 7300 (i860) running AT&T SVR4.0 SPARC SPARCstation running SunOS 4.1.x SUN3 SUN3 running SunOS SUN386 Sun 386i running SunOS 4.0.1 UMAX Encore Multimax running UMAX 4.3 (R4.1.1) VAX VAX running Ultrix 3.1 SRC Modula-3 includes a user manual, compiler, runtime library, some libraries and a few other goodies (see below). The compiler generates C as an intermediate language and should be fairly easy to port. Except for the very lowest levels of the thread implementation, the entire system is written in Modula-3. 3.3. What is m3pc? m3pc is available via anonymous ftp from gatekeeper.dec.com in pub/DEC/Modula-3/contrib/m3pc. From: laszlo@post.ifi.uni-klu.ac.at (Prof.Dr.Laszlo BOESZOERMENYI) Subject: M3 pn PC The Modula-3 system ported by us on the PC and available on the gatekeeper, runs with MSDOS, gnu c compiler and djgpp memory manager (detailed description in the read me file). The first version was ported by Klaus Preschern. Carsten Weich rewrote the handling of long filenames for the second version. You may compile, link and run Modula-3 programs, without threads. From the library modules only those are tested which are used by the compiler. We are using the actual version for our beginners-course and we are working on "student-friendly" programming environmnents and a simple windowing graphic-library. From: carsten@post.ifi.uni-klu.ac.at (Carsten WEICH) Subject: New version of Modula-3/PC Date: 28 Jan 1994 13:25:28 GMT We have built a new version of Modula-3 running on a PC. You will need a PC: * 80486 or 80386 with floatingpoint-coprocessor, * at least 4 but preferable 8 MByte of RAM, * 20 MByte discspace free at least. You can run a modified Modula-3 version 2.09 without thread- support. Our version features: * support for Unix-like long filenames * automatic compilation (to substitute the lack of m3make) * library to access DOS-files and directories. We have built a little shell which translates the long filenames you type in into DOS-filenames. It has a unix-lookalike "ls"-command. There are commands to teach the system new long filenames. You can savely move such files around through DOS-directories (which was not possible with the first version). This version accesses DOS-files significantly faster than the first version. Especially linking large Modula-3 programs is much faster now. On a 50 MHz 80486 with 16 MByte of RAM we measured a turn-around time of 2 minutes for a 5 module, 500 lines program. If you want to compile only one module you will have to wait about one minute. The necessary files are available at "gatekeeper.dec.com" in /pub/DEC/Modula-3/contrib/m3pc: Install.txt please read this first! djgpp.tar the DJGPP-gnu-C-compiler m3.tar compiler binaries m3src.tar sources of the compiler (optional) tar.exe program to unpack these files 3.4. What is GNU Modula-3? From: moss@cs.cmu.edu (Eliot Moss) Subject: GNU Modula-3 alpha release status Date: 25 Mar 93 17:53:12 GMT We said we'd try to get the initial (alpha) release of GNU Modula-3 out some time this month (March), and we're in the process of figuring out what to release and how to package it up. We expect to have something in roughly two weeks (watch this space for a notice). What would this be? First, it is a compiler for the VAX (only) under Ultrix (only), definitely without threads, and probably still failing a small number of the tests in the SRC test suite (which implies that not all of libm3 is likely to work either). The actual release information will detail more about what's working and what's not at that time. We DO currently pass all the compiler tests that the SRC compiler did when it was first released (i.e., the ones we fail are more obscure things that users uncovered over time). Second, the release itself will be a compressed tar file including sources and executables. The executables will probably work only if you put libraries, etc., in the expected places (otherwise, you'll need to rebuild from sources).The compiler is written in C and should be compiled with a recent version of gcc (so you'll need gcc installed). The system also uses gld (the GNU linker). This release should be most useful to people undertaking ports to other machines and operating systems, since it will give them a head start on understanding the compiler and getting the full system ready. It may be of some use for ordinary programming, but it really depends on whether you tend to use features that tickle the remaining bugs. We are indeed interested in alpha test reports, but only if they tell us something new (i.e., we'll provide a list of known deficiencies). When the release is made, we'll also start email discussions with the various parties who have indicated they might undertake ports, to help sort out who will do what. Regards, and thanks for your continued interest and encouragement -- EM From: moss@cs.cmu.edu (Eliot Moss) Subject: GNU Modula-3 pre-release Date: Wed, 5 May 1993 23:49:33 GMT At long last, the GNU Modula-3 project has a pre-release ready, for the VAX/Ultrix platform ONLY. Various folks had notified us of their interest in doing ports or alpha testing, and they have already been sent email with particulars on how to obtain the tar file, etc. There are a number of known bugs; I'll see about making a list available by ftp or something, for interested parties. It is our opinion that the prerelease is not mature enough for general use, but we wished to give a head start to those folks attempting ports, and we will make periodic patches available. If you want to use this compiler for serious program development or need something solid with debugging support for classroom use, you should wait until we've fixed more of the problems. (But to give a sense of what we HAVE accomplished, as I recall, all but 3 of the SRC compiler tests compile (there are 137 of them).) We hope to do a more general release, and support more platforms, in the summer. If you're interested in helping and have not previously contacted us, please send email to me and/or Rick Hudson (hudson@cs.umass.edu). Thanks to Digital and SRC for supporting us, and to Rick Hudson, Amer Diwan, and Norm Walsh, the guys who do all the hard work! 4.1. Why is "Hello World" so large? Modula-3 programs are larger than C programs for the following reasons: 1) The fixed runtime is substantially larger. It contains a garbage collector, a thread runtime, and exception support. Note that "Hello World" is virtually all runtime. For larger programs the runtime is not an issue. 2) The generated code includes runtime checks for out-of-bound array references and NIL pointer. Many of these checks could be removed by a better compiler. The current compiler is only a research prototype. 3) The compiler generates C code as its intermediate language consequently the final machine code suffers. For example, the compiler is constantly storing single-precision floating point values into memory to get around C's predisposition for double precision. 4.2. Why objects and interfaces? Allan Heydon on comp.lang.modula3, May 4th 1993: Modula-3 provides two separate mechanisms for data-hiding: one for hiding details about how interfaces are implemented, and the other for hiding details about how objects are implemented. The first data-hiding mechanism is realized by the distinction between interfaces and modules. Clients can only import interfaces, so the names declared in the modules implementing those interfaces are hidden from clients. Note that this mechanism has only two levels; a name is either declared in an interface, or it isn't. If a name is only declared in a module, it can't be used by a client. The second data-hiding mechanism is realized by opaque types and revelations. A Modula-3 interface may declare an object type to be opaque, in which case only a subset of the fields and methods of that object are revealed to clients importing the interface. Furthermore, the Modula-3 revelation mechanism allows a designer to reveal successively more fields and methods of an object in a series of interfaces. The fields and methods visible to a client then depends on which interfaces the client imports. The latter mechanism is quite flexible. As opposed to the interface/module data-hiding mechanism, opaque types allow you to define an arbitrary number of levels at which more and more information about the implementation of your object is revealed. See Sections 2.2.10, 2.4.6, and 2.4.7 of "Systems Programming with Modula-3" for more information about opaque types and about partial and complete revelations. 4.3. What is the story with Trestle and OpenWindows? Mark Manasse says: I think that the OpenWindows release should be enough (no need to get the MIT X release], although there are a few things in Trestle that trigger devastating bugs in OpenWindows. But the only library we depend on is Xlib, R4 or later. The main thing I know that crashes OW 2.0 is the code where we call GrabKey specifying AnyKey. You can either loop over all of the keys, or you can just comment out the call; programs won't run exactly the same, but you probably won't notice the difference. 4.4 When is the next release of SRC Modula-3 ? The next release will be 3.1. Here are some of the new things you will find in it: 1. the compiler has a new internal interface between the front-end and the back-end, M3CG. This interface is supposed to be easy to implement. 2. the front-end can compute in the target arithmetic system; in particular it is possible to cross-compile to machines with larger integers than the host. 3. one back-end has been implemented on top of gcc. The implementation of M3CG interface generates the tree representation used internally in gcc. From the gcc point of view, this back-end looks like a new front-end. Using this back-end, we have cross-compiled solitaire for mips, alpha and 386 processors; there is no reason to believe that there would be a problem for the other architectures supported by gcc. 4. Dave Hanson wrote another implementation of the M3CG that is self-contained. He is currently working on the 386 code generation (he has done the mips code generation already). 5. gdb has been modified to understand Modula-3 debugging information produced by the back-ends. gdb can now parse Modula-3 expressions, print Modula-3 values and evaluate some of the Modula-3 built-in operations. There is also a little bit of support for multi-threaded programs (you can look at the stacks of other threads). 6. there is a replacement for m3make, m3build, that does not rely on cpp/awk/sed/make and what not, and removes some of the limitations of m3make. m3makefiles are very similar. 7. libm3 has been significantly changed by the Interface Police, mostly in the area of OS interfaces and data structures. 8. for the OS interfaces, we still have the U* interfaces, but applications are not supposed to use those. Instead they should use a new set of interfaces that are os-independent; for example, there is a Pathname interface that manipulates file names; there is a Process interface that manipulate child processes. These interfaces enabled a prototype port of the C based version to Windows NT machines. 9. for the data structures, generics have been introduced and the various data structures are more consistent. 10. because of 6 and 8, we can think about going to different os than Unix. In particular a Windows NT port will be available at some point (may not be in 3.0). 11. the runtime has been improved quite a bit. 12. new platforms: Alpha running OSF/1, 386 running Linux. We will pay more attention to the porting instructions and support. 13. I am not sure about all the changes in the libraries other than libm3. I suspect that there will be few changes in trestle, but that mentor changed quite a bit. 14. The Windows NT port uses native threads. This should be a good model for other implementations of Thread using native threads. The current status is: . the front-end is very stable . the gcc-based back-end has been stable for 4 months . the gdb extensions are brand new and need some test . the interface police work is very stable . we are working on bringing the system up on the machines we have in the building, and building the export machinery. We don't have a date for the 3.1 release. Given the amount of changes introduced by 3.1, I suspect that the first few releases will not work out of the box for any machine but the ones for which we can test (decstations [mips and alpha], linux). Consequently, I expect a high rate of releases for a while. We will try to post accurate information about the status of each machine, but we can only rely what you tell us. At this point, I would not encourage anybody to start a new port. If you have a new port, or are close to complete one, you can send us your bits, we will try to put them in 3.1. 5.1. What if I don't have ftp access? Unfortunately, we cannot deliver Modula-3 other than by anonymous ftp. Fortunately, Prime Time Freeware (PTF) includes Modula-3. PTF is a set of two ISO-9660 CDroms filled with 3GB of freeware, issued semi-annually. The latest issue, Volume 1, Number 2, July 1992, contains SRC Modula-3 2.07. PTF is distributed via bookstores and mail. You can reach PTF using: Email: ptf@cfcl.com Fax: [1] (408) 738 2050 Voice: [1] (408) 738 4832 Mail: Prime Time Freeware 415-112 N. Mary Ave., Suite 50 Sunnyvale, CA 94086 USA 6.1. Can I contribute Modula-3 software? Certainly. Send us what you are willing to share, be it programs, libraries or other things. We'll put them in the distribution. Right now, the pub/DEC/Modula-3/contrib directory contains: m3rpc an rpc system from Xerox Parc M2toM3 a translator from Modula-2 to Modula-3 m3pc an implementation of Modula-3 for PCs. ---- Eric. ======================================================================= 11 === Date: 02 Mar 1994 20:53:10 GMT From: hudson@cs.umass.edu (Rick Hudson) Subject: Re: Where to start - GNU M3 status. >>>>> "Rod" == Rod Oldehoeft writes: > I'm anxiously awaiting the GNU > version. Can anyone provide an update on the state of this project? I put in as much time as anyone on the GM3 project and have it wired up and sort of running on my desk well enough to do run experiments. Unfortunately, it is in no shape for prime time. No threads and 20% of the SRC libraries don't even compile. Work slowed in July due to me switching projects and pretty much halted when it became obvious that SRC was doing such a good job on their compiler. We are porting pieces of our backend and runtime into the SRC stuff but it is just for internal research use since that is all we really have time for. Cheers, -- Richard L. Hudson - Hudson@cs.umass.edu; (413) 545-1220; Advanced Languages Project - University Computing Services Lederle Graduate Research Center University of Massachusetts Amherst, MA 01003 ======================================================================= 12 === Date: Thu, 03 Mar 94 08:21:28 -0800 From: mcjones@src.dec.com Subject: Re: Document on M3 3.1 interface changes Quentin Stafford-Fraser asks: Is there a document anywhere that describes the changes to the interfaces? Or an index to them? Michel comments: I have seen the draft of a nicely organized document. I presume it will come out fairly soon. It is in fact now available, as: Some Useful Modula-3 Interfaces Jim Horning, Bill Kalsow, Paul McJones, Greg Nelson SRC Research Report 113 You can retrieve a PostScript version via anonymous FTP from gatekeeper.dec.com in pub/DEC/SRC/research-reports/SRC-113.ps.Z The document includes an index. Please send any feedback on the document to me: Paul McJones mcjones@src.dec.com ======================================================================= 13 === Date: Thu, 3 Mar 1994 21:14:12 GMT From: murphy@sbu.edu (Michael Murphy) Subject: SRC Modula-3 FTP sites *************************************************************** * * * Saint Bonaventure University * * Department of Computer Science * * * *************************************************************** * * * Michael Josephy Murphy III * * * *************************************************************** I WAS UNABLE TO FTP THE SRC MODULA-3 FROM GATEKEEPER.DEC.COM BECAUSE I COULD NOT MATCH THE DIRECTORIES I FOUND IN A FAQLIST WITH XARCHIE OR MANUAL FTP. DOES ANYBODY KNOW HOW I CAN REALLY GET IT... I AM RUNNING DeskTop SPARC w/X11 olwm mounted on SunOS 4.1.x *************************************************************** * * * * mmurphy@sbu.edu * NO LIBERTY WITHOUT DISCRETION WHILE * * * LAWS OF CAUSE & EFFECT STILL APPLY * * * * *************************************************************** -- ***************************************************** * * * Michael Murphy mmurphy@sbu.edu * * * ======================================================================= 14 === Date: Thu, 3 Mar 1994 20:34:43 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Annoucement: Modula-3 on a SPARC Mailing List I have created the mailing list "m3-sparc@gte.com" for the people running Modula-3 on SPARCs. If you would like to be on this mailing list, please send me e-mail at "nayeri@gte.com". The purpose of this mailing list is to keep everybody who runs Modula-3 on SPARCs updated on the changes to new releases to Modula-3 or anything that is specific to running Modula-3 on SPARCs, SunOS or Solaris. I will keep an archive of messages in this in: ftp.gte.com:/pub/m3/m3-sparc.tar To post to the mailing list, send your messages to "m3-sparc@gte.com". Please let me know if you have any ideas or suggestions about the operations, as this is the first time I have made a public mailing list. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 15 === Date: 02 Mar 94 23:11:59 -0500 From: Joe.Klemmer@f370.n109.z1.fidonet.org (Joe Klemmer) Subject: Where to start Torsten R. Kirschner wrote in a message to All: >Now to the software. The PC version from DEC hasn't received very much >good press as far as I can tell, and I'm anxiously awaiting the GNU >version. Can anyone provide an update on the state of this project? TRK> As far as I remember, neither the SRC nor DEC as such has TRK> ever put out a/the PC version of M3. What's stored on TRK> gatekeeper.dec.com as m3pc is merely a port of the M3 TRK> language, done by someone outside SRC. As far as its quality TRK> is concerned, I have no personal experience. TRK> However, one should take into account that (a) It's a TRK> language port, not a complete implementation of the SRC M3 TRK> package. TRK> It (b) does not have thread support since that would be an TRK> interesting task in itself, I guess. After all, it's (c) TRK> free, and if you have enough DOS disk space but cannot get a TRK> UNIX for some reason it might just be your only chance to TRK> work with the M3 language on a PC. So, lacking competition, TRK> it's per se "the best there is" on PC-DOS. Well, since that is the situation I'm in it definitely looks like the "best there is" all right. I know enough programming in other languages to get by, I think. I'm interested in picking up some different points-of-view. M3 looks like a bet to learn some new things. TRK> As far as the original question - where to start with M3 (on TRK> PC) - is concerned, one might also consider Robert TRK> Sedgewick's "Algorithms in Modula-3", Addison-Wesley. TRK> Although all the traditional algorithms are merely TRK> re-written in Modula-3 and there are no algorithms or data TRK> structures that make inventive use of M3-specific features, TRK> it is all in all a good book. I saw that book when I picked up the Harrison one. Nothing like spending $109 on programming books in one shot. Guess I could go back and get the Algorithms book, too. Thanks, Joe ======================================================================= 16 === Date: 04 Mar 94 05:51:44 -0500 From: Joe.Klemmer@f370.n109.z1.fidonet.org (Joe Klemmer) Subject: Where to start fn00@gte.com wrote in a message to All: fc> I think your best bet right now is to run Linux on your PC. fc> I know this is not a good option for the people with fc> _really_ tight space or ones that don't want to deal with fc> running UNIX, but I think this way, you can use the fc> full-fledged Modula-3 system including FormsVBT, Obliq, fc> Network Objects. By the way the first two of these systems fc> are very well-introduced as SRC TRs by Marc H. Brown and fc> Luca Cardelli. I'd be ecstatic if I could run Linux. But to do so would bean my purchasing another PC. All my systems are currently tied up running my BBS. I have to share one for all the stuff I want to do. Thus I'm limited to DOS only. :-( Later, Joe ======================================================================= 17 === Date: 4 Mar 94 16:45:37 GMT From: rro@CS.ColoState.EDU (Rod Oldehoeft) Subject: Revised Harbison errata sheet available Peter Klein has supplied me with many new errata for Harbison's book. The following are available from beethoven.cs.colostate.edu via anonymous ftp, in ~ftp/pub/M3/*: harbison-errs.ps.Z Newest errata sheet harbison-errs.tex LaTeX form of above previous-version.ps.Z The previous errata sheet latest-errs Just the info I got from Peter I provide the previous sheet, and the new errata, in case your book has the previous updates, and you want to conveniently apply the new changes without going through the whole new thing. ======================================================================= 18 === Date: Mon, 7 Mar 94 13:58:59 EST From: gwyant@cloyd.East.Sun.COM (Geoffrey Wyant - Sun Microsystems Labs BOS) Subject: problems using netobj rule in m3makefile I have a directory that looks like: pkgname/SPARC pkgname/src/ pkgname/src/interfaces where the pkgname/interfaces directory contains some network object interfaces. The src directory makefile contains the line: include_dir(interfaces) and the interfaces directory m3makefile contains import(netobj) netobj(Coordinator,T) When I try to build this, I get the following error: % m3build --- building in SPARC --- stubgen -sno Coordinator.T -T.M3IMPTAB m3cfe: (Error) failed to find source for interface 'Coordinator' stubgen: NetObj.T not known in this scope. *** error code 1 It is executing it in the SPARC subdirectory, and it seems unable to find the Coordinator.i3 source file in the interfaces directory (../src/interfaces). Can anyone sshed any clues as to what is going on with 'stubgen' or how to track this farther ? thanks ! --geoff ======================================================================= 19 === Date: 7 Mar 1994 15:41:03 GMT From: laszlo@post.ifi.uni-klu.ac.at (Laszlo BOESZOERMENYI) Subject: Re: Where to start A new version of M3-PC has been avaiable for a few weeks. (By the way: the port wasn't made by DEC, rather by Klaus Preschern und Carsten Weich at the University of Klagenfurt.) Next enhancements (e.g. adding graphics) are to be hoped still before summer. An introduction into programming with Modula-3 is also written by us. It should appear in autumn in German and not too much later in English (at Springer). Best regards Laszlo Boszormenyi ********************************* * Prof. Laszlo Boeszoermenyi * * Institut fuer Informatik * * * * Universitaet Klagenfurt * * Universitaetsstr. 65-67 * * A-9022 Klagenfurt / Austria * * * * e-mail: * * laszlo@ifi.uni-klu.ac.at * * Tel.: * * 00-43-463-2700-509 * * 00-43-463-2700-506 (secr.) * * Fax.: * * 00-43-463-2700-505 * ********************************* ======================================================================= 20 === Date: 7 Mar 1994 19:35:31 GMT From: patl@bodacia.Eng.Sun.COM (Pat Lashley [MtV NeWStech Eng]) Subject: :? SRC m3-3.1 on SPARC/Solaris2 Has anyone successfully built SRC Modula-3 v3.1 on a SPARC running Solaris2? I'm trying to build on Solaris 2.3 using the SC2.0.1 SPARCware compilers. To build quake, I had to replace random() with rand() in boot-SOL2/quake/src/hash.c, add an explicit -D__sparc to the DEFINES macro in boot-SOL2/quake/src/makefile, and modify a conditional in boot-SOL2/quake/src/parser.c. (The test for whether to include was failing; apparently because of the check for ANSI c.) I've included a context diff below, along with a change to the SOL2 config file to correct the location of as and ar. With these changes, everything proceeds smoothly until it attempts to link the complier. At this point a few undefined symbols are found - all BSD routines that aren't supported in Solaris2 native mode. I suspect that either the bootstrap libm3 wasn't built with the USG flag set, or that a few more tests for it need to be inserted... Undefined first referenced symbol in file getwd ../../libm3/SOL2/libm3.a(ProcessPosix_m.o) getdtablesize ../../libm3/SOL2/libm3.a(ProcessPosix_m.o) sigsetmask ../../libm3/SOL2/libm3.a(FloatMode_m.o) _longjmp ../../libm3/SOL2/libm3.a(RTThreadC.o) _setjmp Host_m.o ld: fatal: Symbol referencing errors. No output written to m3c -Pat *** boot-SOL2/m3build/templates/SOL2 Mon Feb 14 10:43:41 1994 --- boot-SOL2-local/m3build/templates/SOL2 Fri Mar 4 14:03:17 1994 *************** *** 174,181 **** if defined("_bootstrap") BOOT_CC = "/opt/SUNWspro/bin/cc" ! BOOT_ASM = "/bin/as" ! BOOT_AR = "ar cru" BOOT_RAN = "touch" BOOT_LINK = "/opt/SUNWspro/bin/cc" include("COMMON.BOOT") --- 174,181 ---- if defined("_bootstrap") BOOT_CC = "/opt/SUNWspro/bin/cc" ! BOOT_ASM = "/usr/ccs/bin/as" ! BOOT_AR = "/usr/ccs/bin/ar cru" BOOT_RAN = "touch" BOOT_LINK = "/opt/SUNWspro/bin/cc" include("COMMON.BOOT") *** boot-SOL2/quake/src/parser.c Tue Jan 25 16:46:22 1994 --- boot-SOL2-local/quake/src/parser.c Fri Mar 4 13:53:27 1994 *************** *** 317,328 **** along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ - #ifndef alloca #ifdef __GNUC__ #define alloca __builtin_alloca #else /* not GNU C. */ ! #if (!defined (__STDC__) && defined (sparc)) || defined (__sparc__) || define d (__sparc) || defined (__sgi) #include #else /* not sparc */ #if defined (MSDOS) && !defined (__TURBOC__) --- 317,328 ---- along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #ifndef alloca #ifdef __GNUC__ #define alloca __builtin_alloca #else /* not GNU C. */ ! /*#if (!defined (__STDC__) && defined (sparc)) || defined (__sparc__) || defi ned (__sparc) || defined (__sgi)*/ ! #if (defined (sparc)) || defined (__sparc__) || defined (__sparc) || defined (__sgi) #include #else /* not sparc */ #if defined (MSDOS) && !defined (__TURBOC__) --- PMLashley plashley@Sun.COM SunSoft "Those who do not understand NeWS are doomed to recreate it - badly" ======================================================================= 21 === Date: Tue, 8 Mar 94 13:27:24 EST From: gwyant@cloyd.East.Sun.COM (Geoffrey Wyant - Sun Microsystems Labs BOS) Subject: debugging m3build files/quake scripts ? Is there any way to get quake to print out the actions/procedures it is executing when it executes an 'm3makefile' ? In general, what is a good way of debugging m3makefiles and quake scripts ? thanks ! --geoff ======================================================================= 22 === Date: 8 Mar 1994 16:02:23 GMT From: will@marvin.or.uni-bonn.de (Hans-Martin Will) Subject: Installation of M3 3.1 on AIX 3.2.5 Hi, has anybody out there succeeded in installing boot-IBMR2... from gatekeeper.dec.com upon an IBM AIX 3.2.5 system? On my system m3boot and m3ship seem to terminate successfully, but the created executable m3c loops endlessly in sbrk(). [That's what dbx is telling after ^C] Does anybody have working executables? Thank you very much, Martin Will email: will@marvin.or.uni-bonn.de ======================================================================= 23 === Date: 8 Mar 1994 12:43:54 -0800 From: omalley@porte-de-st-ouen.ics.uci.edu (T. Owen O'Malley) Subject: Comparing the types of object I would like to compare the types of two objects. For example, I would like to be able to do something along these lines: PROCEDURE Subtype(Left,Right : ROOT) : boolean = BEGIN TYPECASE Left OF Right(Something) => RETURN TRUE; ELSE RETURN FALSE; END; END Subtype; This _is_not_ legal because the selectors of a typecase must be types. Alternately, if I used: BEGIN RETURN TYPECODE(Left) = TYPECODE(Right); END Subtype; I'll only see if the types are equal and not subtypes. Is there a way to do this? Thanks, Owen O'Malley Department of ICS | omalley@ics.uci.edu (ARPA) UC Irvine | "http://www.ics.uci.edu/~omalley/" (WWW) Irvine, CA 92717 | ucbvax!ucivax!omalley (UUCP) ======================================================================= 24 === Date: Wed, 9 Mar 1994 00:10:59 GMT From: fandrich@malibu.sfu.ca (Daniel Theodore Fandrich) Subject: Class library for standard I/O? Has anyone written a class library for doing standard I/O in M-3 in a better fashion than the default libraries? There must be an easier way than Rd, Wr, Fmt and Scan! >>> Dan ======================================================================= 25 === Date: 9 Mar 1994 05:31:04 GMT From: lucee@ryoohki.its.rpi.edu (Scanner) Subject: Re: M3 3.1 on SPARCs - Questions fn00@gte.com (Farshad Nayeri) writes: >We also had some existing code that heavily used the old List >interface and a lot of its features. One thing to note is that you can >recompile the old List interface with the new compiler since it only >counts on Text and Thread. Sx has also changed a lot. Oh no! I just wrote about 10,000 lines of an imbeddable scheme interpreter that used Sx... oh well.. -- - You ever have one of those days where you just want to teleport in - somewhere, destroy a couple of monsters, do some property damage, - then teleport out? ** internet: scanner@rpi.edu ** bitnet: userf0f3@rpitsmts.bitnet ======================================================================= 26 === Date: 9 Mar 1994 12:12:32 GMT From: laszlo@post.ifi.uni-klu.ac.at (Laszlo BOESZOERMENYI) Subject: Re: Class library for standard I/O? > Has anyone written a class library for doing standard I/O in M-3 in a better > fashion than the default libraries? There must be an easier way than Rd, > Wr, Fmt and Scan! > >>> Dan SRC has an "IO" interface for simple IO. We have made another, a bit more powerful standard I/O. (Modules SIO and SF.) If some people are interested, we can send the sources, resp. to put the whole stuff on the gatekeeper. I append the interfaces. You don't need SF if you don't use disk files, just standard i/o. SF is used to open files, naming is quite flexible. Usage of SIO is quite simple, such as SIO.PutInt(i); SIO.PutChar(ch), i:= SIO.GetInt(); ch:= SIO.GetChar() etc. With "Gets" you can read a stream of typed data. The interfaces are intended mainly for use in education. Greetings Laszlo Boszormenyi ********************************* * Prof. Laszlo Boeszoermenyi * * Institut fuer Informatik * * * * Universitaet Klagenfurt * * Universitaetsstr. 65-67 * * A-9022 Klagenfurt / Austria * * * * e-mail: * * laszlo@ifi.uni-klu.ac.at * * Tel.: * * 00-43-463-2700-509 * * 00-43-463-2700-506 (secr.) * * Fax.: * * 00-43-463-2700-505 * ********************************* ---------------------------------------------------------------- INTERFACE SIO; (*Simple Input/Output 12.01.94. LB*) (*SIO provides a Reader and a Writer type. A reader or a writer is a stream of typed data. Data can be read from a reader by "Get" procedures, and can be written by "Put" procedures onto a writer. The "Get" and "Put" procedures advance a (hidden) position over the reader resp. writer stream. The procedure "LookAhead" returns the next character, without advancing the reader position. All "Put" procedures flush automtically on stdout. GetText terminates with new line (other white space characters may occur insi de). GetInt, GetReal, GetBool terminate with any white space character, leading whitespaces are ignored. (White spaces are: new line, tab, space, form feed and carriage return.) "Error" exception is raised only for readers, not connected to Stdio.stdin. For Stdio.stdin, the user is prompted to type in a new value. SIO provides some additional functions, such as positioning (seeking) in readers, length and end of readers, flushing of writers etc. Writers are strictly sequential, positioning is not supported. The default value of the reader or writer parameter is always NIL, with the effect, selecting the appropriate standard device. Standard reader is Stdio.stdin, which is normally the keyboard. Standard writer is Stdio.stdout, which is normally the screen.*) IMPORT Rd, Wr; EXCEPTION Error; TYPE Reader = Rd.T; Writer = Wr.T; (*-------- Basic procedures --------*) PROCEDURE GetChar(rd: Reader := NIL): CHAR RAISES {Error}; (*Returns the next char.*) PROCEDURE PutChar(ch: CHAR; wr: Writer := NIL); (*Outputs a single character.*) PROCEDURE GetText(rd: Reader := NIL): TEXT RAISES {Error}; (*Reads a text. The terminating character is not appended.*) PROCEDURE PutText(t: TEXT; wr: Writer := NIL); (*Outputs all characters in t.*) PROCEDURE GetInt(rd: Reader := NIL): INTEGER RAISES {Error}; (*Reads a decimal integer.*) PROCEDURE PutInt(i: INTEGER; length := 3; wr: Writer := NIL); (*Outputs an integer number. *) PROCEDURE GetReal(rd: Reader := NIL): REAL RAISES {Error}; (*Reads a real number.*) PROCEDURE PutReal(r: REAL; wr: Writer := NIL); (*Outputs a real number. *) PROCEDURE GetBool(rd: Reader := NIL): BOOLEAN RAISES {Error}; (*Reads a Boolean constant, terminated by any character in "terminate". Legal values are: TRUE, True, true, T, t resp. FALSE, False, false, F, f.*) (*!!! Currently works only with T, t, F, f due to an error in "Lex"*) PROCEDURE PutBool(b: BOOLEAN; wr: Writer := NIL); (*Outputs a Boolean value.*) (*-------- Additional procedures --------*) PROCEDURE LookAhead(rd: Reader := NIL): CHAR RAISES {Error}; (*Returns the next character, without removing it from the reader.*) PROCEDURE Nl(wr: Writer := NIL); (*Outputs a new line character*) PROCEDURE PutUnsigned(i: INTEGER; length := 6; base: [2..16] := 16; wr: Writer := NIL); (*Outputs an unsigned number with given base.*) PROCEDURE End(rd: Reader := NIL; skip:= TRUE): BOOLEAN; (*Returns TRUE iff end of reader reached. On the keyborad CTRL-Z on the PC, and CTRL-D in Unix. Skips whitespaces if "skip" is true.*) PROCEDURE Flush(wr: Writer := NIL); (*Flushes the writer on the file. Not necessary for Stdio.stdout.*) PROCEDURE Available(rd: Reader := NIL): BOOLEAN; (*Returns TRUE iff some characters are available.*) (*!!! Currently doesn't work on the PC !!! - it blocks*) PROCEDURE Length(rd: Reader := NIL): CARDINAL; (*Returns the length of the reader. Returns 0 if the length cannot be computed.*) (*!!! Currently doesn't work on the PC !!!*) PROCEDURE Seek(rd: Reader := NIL; position: CARDINAL := 0); (*Sets the reader on "position" if it is seekable. Default corresponds to reset a reader.*) (*!!! Currently doesn't work on the PC !!!*) END SIO. ---------------------------------------------------------------- INTERFACE SF; (*SimpleFiles 7.01.94. LB*) (*SimpleFiles provides simple procedures to connect readers and writers with files ("Open" procedures), and to decouple them ("Close" procedures). "OpenRead" connects a reader to an existing file, "OpenWrite" connects a writer to a new file. If the file with the given name already exists, the user may type in a new name or let it overwrite. "OpenAppend" positions the writer at the end of a file. The text in parameter "prompt" is displayed on standardout (screen). The file connected to a writer is made permanent by a "CloseWrite". SimpleFiles provides a flexible mechanism for file name specification. If the "name" parameter is omitted (or NIL) at opening, then the opening procedures ask the user for a file name. If the user enters an empty line, the file defaults to standard i/o, i.e. normally to keyboard (standardin) and screen (standardout). "GetFileName" provides more explicit control over file name specification. *) IMPORT Rd, Wr; CONST PromptStart = "Type file name "; PromptEnd = " or NL for standard = "; InPrompt = PromptStart & "for input" & PromptEnd; OutPrompt = PromptStart & "for output" & PromptEnd; AppPrompt = PromptStart & "for append" & PromptEnd; TYPE Reader = Rd.T; Writer = Wr.T; PROCEDURE OpenRead(name: TEXT := NIL; prompt:= InPrompt): Reader; (*Connects the file, called "name" to a reader. If "name" is NIL, or the file does not exists, it asks the user for a file name until the file can be opened. It returns Stdio.stdin if the user enters "return".*) PROCEDURE OpenWrite(name: TEXT := NIL; prompt:= OutPrompt; overwrite:= FALSE): Writer; (*Connects a writer to the file, called "name". If "name" is NIL it asks the user for a file name. It returns Stdio.stdout if the user enters "return". If the file (given in "name" or typed in by the user) already exists, and "overwrite" is false, it prompts the user to give another file name or to type "End" (CTRL-Z on PC, CTRL-D in Unix) for overwriting. If parameter "overwrite" is true, an existing file with same name will be overwritten, without asking the interactive user!*) PROCEDURE OpenAppend(name: TEXT := NIL; prompt:= AppPrompt): Writer; (*Connects a writer to the file, called "name". If "name" is NIL it asks the user for a file name. It returns Stdio.stdout if the user enters "return". Creates a new file, if the specifeid file doesn't exists yet. If the file already exists, it positions the writer to the end of it.*) (*!!! Currently doesn't work on the PC !!! - it doesn't append*) PROCEDURE FileExists(name: TEXT): BOOLEAN; (*Returns TRUE iff the file exists.*) PROCEDURE CloseRead(VAR rd: Reader); (*Closes the file. Assigns NIL to rd - except rd is stdin.*) PROCEDURE CloseWrite(VAR wr: Writer); (*Flushes the writer and closes the file. Assigns NIL to wr - except wr is stdout. Close MUST be called if the content of the writer should be made permanent!*) PROCEDURE GetFileName(prompt:= PromptStart & PromptEnd): TEXT; (*Asks the user for a file name - actually a simple text. It is the very same procedure used by OpenRead and OpenWrite. Returns NIL if the user enters "return".*) END SF. ======================================================================= 27 === Date: Wed, 9 Mar 1994 03:27:24 GMT From: 890930g@dragon.acadiau.ca (D. Aaron Gallagher) Subject: installation problems on a SPARC Hi, I'm trying to install release 3.1, and it keeps hanging, or appearing to hang. It makes it through the basic installation ok, although it appears to be using cc sometimes, although gcc is specified, and it doesn't seem to be putting in debugging code, even with the ccflag set in the build file in m3cc. But the real problem seems to occur when it try to compile the libm3 package. I run m3build and it builds all the files, then hits: m3 @M3novm -w1 -why -g -times -a libm3.a -F/usr/tmp/aaaa10336 and stays there. I've let this go for atleast half an hour. I guess what I'm asking is, is this getting stuck there, or am I just being impatient? Any help is gratly appreciated, -- ------------------------------------------------------------------------------- Aaron Gallagher, aka 890930g@dragon.acadiau.ca Go ahead and flame. Just remember, I'm on a Dragon. +++++++++++++++++++++++++E-mail at 890930g@acadiau.ca++++++++++++++++++++++++++ ======================================================================= 28 === Date: Wed, 9 Mar 1994 17:25:23 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: M3 3.1 on SPARCs - Questions In article <2ljmuo$2ja@usenet.rpi.edu> lucee@ryoohki.its.rpi.edu (Scanner) writ es: fn00@gte.com (Farshad Nayeri) writes: >We also had some existing code that heavily used the old List >interface and a lot of its features. One thing to note is that you can >recompile the old List interface with the new compiler since it only >counts on Text and Thread. Sx has also changed a lot. Oh no! I just wrote about 10,000 lines of an imbeddable scheme interpreter that used Sx... oh well.. That's pretty much what we had as well. And the new Sx is pickier about identifiers (only alphanumeric characters) since it is now using Atom.T instead of Sx.Symbol. Though they did add ";" as a comment character. :-) You can hack the new Sx to allow for other identifiers, but it sure ain't a pretty hack. How complete is your Scheme implementation? Is it R4Rs? What Scheme datatypes do you support? Is your system going to be available for public use? --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 29 === Date: Wed, 9 Mar 1994 17:16:09 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: installation problems on a SPARC In article <1994Mar9.032724.8342@relay.acadiau.ca> 890930g@dragon.acadiau.ca (D . Aaron Gallagher) writes: [...] I run m3build and it builds all the files, then hits: m3 @M3novm -w1 -why -g -times -a libm3.a -F/usr/tmp/aaaa10336 and stays there. I've let this go for atleast half an hour. Release 3.1 included infinite loops in place of some SPARC dependencies. Checkout early messages in: ftp.gte.com:/pub/m3/m3-sparc directory to solve some of these problems as well as some others that you may face if you continue working on the port to SPARC. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 30 === Date: Wed, 9 Mar 94 08:53:17 EST From: gwyant@cloyd.East.Sun.COM (Geoffrey Wyant - Sun Microsystems Labs BOS) Subject: using stubgen 'Stubgen' seems to have some magic knowledge about the way packages are structured. In particular, it seems to work only when the interfaces to be processed are located directly in the 'src' subdirectory of the package. If they are located in a subdirectory of 'src', it fails to find them. This can be manually worked around by invoking 'stubgen' manually and passing it the '-D' flag to give indicate the directory in which they live. There doesn't seem to be anyway to do this through th 'netobj' quake template, though. Is there anyway of specifying additional flags for 'stubgen' through the 'netobj' quake templates ? --geoff ======================================================================= 31 === Date: Wed, 9 Mar 94 10:20:16 EST From: gwyant@cloyd.East.Sun.COM (Geoffrey Wyant - Sun Microsystems Labs BOS) Subject: disregard posting on stubgen and flags I recently posted a question about 'stubgen' not finding the source for interfaces. My problem turned out to be an incomplete 'm3makefile'. It turns out the the netobj(Foo, T) rule in an 'm3makefile' needs to really have the form interface(Foo) netobj(Foo, T) and then stubgen can find the interface source. --geoff Geoff Wyant Geoff.Wyant@east.sun.com Sun Microsystems Laboratories, Inc. 2 Elizabeth Drive Chelmsford, Ma. 01824 ======================================================================= 32 === Date: 9 Mar 1994 19:40:03 GMT From: jbpn1@hermes.cam.ac.uk (J.B.P. Naylon) Subject: Documentation for netobjs in 3.1? Is there any documentation for the new NetObj library in release 3.1 available anywhere? I'm writing a distributed system using Modula-3 and the ANSA system at the moment, but I'd like to evaluate the netobj library just for fun. I downloaded the netobj.tar from gatekeeper but there did not seem to be any documentation in there, nor in DOC.ps from the same source. Any pointers? Thanks. -- John Naylon ======================================================================= 33 === Date: 09 Mar 94 06:02:59 -0500 From: Joe.Klemmer@f370.n109.z1.fidonet.org (Joe Klemmer) Subject: Revised Harbison errata sheet available rro@CS.ColoState.EDU wrote in a message to All: rCE> Peter Klein has supplied me with many new errata for rCE> Harbison's book. The following are available from rCE> beethoven.cs.colostate.edu via anonymous ftp, in rCE> ~ftp/pub/M3/*: rCE> harbison-errs.ps.Z Newest errata sheet rCE> harbison-errs.tex LaTeX form of above rCE> previous-version.ps.Z The previous errata sheet rCE> latest-errs Just the info I got from Peter rCE> I provide the previous sheet, and the new errata, in case rCE> your book has the previous updates, and you want to rCE> conveniently apply the new changes without going through the rCE> whole new thing. Any chance of getting this in plane ascii format for those of us who can't to tex or ps? Thanks. Later, Joe ======================================================================= 34 === Date: Wed, 9 Mar 1994 18:31:23 GMT From: mhcoffin@tolstoy.uwaterloo.ca (Michael Coffin) Subject: m3build I think that m3build, quake, etc. are quite nice. However, from the standpoint of education, it represents a problem. First, it covers up quite a bit of complexity because it is so automatic. Second, part of my mandate is to teach university students how to use standard tools such as "make". There seems to be no easy way to use modula3 version 3 with make. Would it be possible to provide a more "Unix-like" wrapper for m3? Something that acts roughly like the m3 of version 2? Michael Coffin mhcoffin@tolstoy.uwaterloo.ca Dept. of Computer Science office: (519) 888-4820 University of Waterloo home: (519) 884-6877 Waterloo, Ontario, Canada N2L 3G1 ======================================================================= 35 === Date: 9 Mar 1994 20:19:36 GMT From: sims@gehrig.ucr.edu (david sims) Subject: Re: Documentation for netobjs in 3.1? jbpn1@hermes.cam.ac.uk writes: john> Is there any documentation for the new NetObj library in release john> 3.1 available anywhere? It's not documentation, but there's a paper in the proceedings of the December 93 SOSP (acm symposium on operating system principles) about network objects. I don't have mine here, but I do know that the proceedings were mailed out to members of the ACM SIG on operating systems. cheers! -- David L. Sims Department of Computer Science sims@cs.ucr.edu University of California +1 (909) 787-6437 Riverside CA 92521-0304 PGP encryption key available on request. USA ======================================================================= 36 === Date: 9 Mar 1994 23:15:13 GMT From: lucee@ryoohki.its.rpi.edu (Scanner) Subject: Re: M3 3.1 on SPARCs - Questions fn00@gte.com (Farshad Nayeri) writes: >In article <2ljmuo$2ja@usenet.rpi.edu> lucee@ryoohki.its.rpi.edu (Scanner) wri tes: > Oh no! I just wrote about 10,000 lines of an imbeddable scheme interpreter > that used Sx... oh well.. >That's pretty much what we had as well. And the new Sx is pickier >about identifiers (only alphanumeric characters) since it is now using >Atom.T instead of Sx.Symbol. Though they did add ";" as a comment >character. :-) You can hack the new Sx to allow for other identifiers, >but it sure ain't a pretty hack. Hm. Well, one has to do what one has to do.. but these modules I am making are getting larger as time goes on! (I was in the middle of dumping List.T because I need full cons cells to be more compliant) Maybe I should just ditch using Sx, and write my own stuff. >How complete is your Scheme implementation? Is it R4Rs? What Scheme >datatypes do you support? It is R4R. I support integers, reals, strings, Time.T, Boolean, Char, Array, SxSymbol.T, List.T, Address (Uin.struct_sockaddr_in), NetAuth.AuthEnt, MyUID.T (their UID.T was not unique enough), and Glom.T I didn't have plans for rationals and imaginaries, but if I were to be motivated that would be easy to add. All of those other weird types are needed by the 'greater' system that this is being written for. The Glom is an associative table the supports the above types as keys and values. The Glom has a self-like inheritance model. (It is what I use to define regions inside the scheme interpreter.) > Is your system going to be available for public use? Yes. I had not said anything yet because I hate promising something without having a more or less completed working module. The scheme interpreter is used as the interest language for a mildly distributed database I am writing called Scatter. Scatter uses a networking library called NetJack that provides authenticated network connections (using bsd sockets & TCP), with the ability to have a NetJack connection declared "stale" after some idle time, which causes the network layer to be dismantled. It would be reconnected transparently whenever more data is pushed across the NetJack.T. Basically, I was going to release all of these when they were finished. They are (were) quite near being done. Now it will take a little longer since I have to regroup with 3.1 and see how that affects what I have. -- - You ever have one of those days where you just want to teleport in - somewhere, destroy a couple of monsters, do some property damage, - then teleport out? ** internet: scanner@rpi.edu ** bitnet: userf0f3@rpitsmts.bitnet ======================================================================= 37 === Date: 10 Mar 94 03:27:21 GMT From: jeffo@uiuc.edu (J.B. Nicholson-Owens) Subject: Re: Revised Harbison errata sheet available Joe Klemmer writes: >Any chance of getting this in plane ascii format for those of us who can't to >tex or ps? Thanks. Or for those of us who can't read the PS file generated by TeX because the font used in the output is so incredibly illegible? -- *NO* NeXTmail please ======================================================================= 38 === Date: Thu, 10 Mar 1994 13:55:33 GMT From: rroupski@borg.bell.ca (Roumen Roupski NTMS ) Subject: Modula-3 on Solaris >> Has anyone successfully built SRC Modula-3 v3.1 on a SPARC >> running Solaris2? I'm trying to build on Solaris 2.3 using >> the SC2.0.1 SPARCware compilers. Hi Pat. I am trying to install SRC Modula-3 on Solaris 2.3 as well. I had the same problems as you have now. Here is what I have done so far: 1. I used /usr/ucb/cc C compiler insted of the SC2.0.1 SPARCware compilers.I was able to build quake as it is, and to compile m3 and m3c. If you use /usr/ucb/cc, you will not have these undefined symbols: getwd ../../libm3/SOL2/libm3.a(ProcessPosix_m.o) getdtablesize ../../libm3/SOL2/libm3.a(ProcessPosix_m.o) sigsetmask ../../libm3/SOL2/libm3.a(FloatMode_m.o) _longjmp ../../libm3/SOL2/libm3.a(RTThreadC.o) _setjmp Host_m.o BUT, you will probably have the following undefined symbol: __cg92_used in /opt/SUNWspro/SC2.0.1/cg92/libm.a 2. In order to solve this I linked m3 and m3c with: /opt/SUNWspro/SC2.0.1/cg89/libm.a So far, so good. Everything compiles and links just fine. Then I ran m3ship and I thought that I have got working Modula-3 compiler. But I was wrong. Both m3 and m3c are not working. When I run either one them consumes a lot CPU time, but does nothing. It seems to be hanging somewhere. I used gdb to find out what happend, and here is where I am now: gdb m3 GDB is free software and you are welcome to distribute copies of it under certain conditions; type "show copying" to see the conditions. There is absolutely no warranty for GDB; type "show warranty" for details. GDB 4.11 (sparc-sun-solaris2.3), Copyright 1993 Free Software Foundation, Inc...(no debugging symbols found)... (gdb) run Starting program: /ntms/users/rroupski/modula3/SOL2/m3 warning: stopped at unknown breakpoint while handling shared libraries Program received signal SIGTRAP (5), Trace/breakpoint trap 0xef6a9040 in _alloc_cnts () (gdb) where #0 0xef6a9040 in _alloc_cnts () #1 0xef6ae948 in _alloc_cnts () #2 0xef6ae88c in _alloc_cnts () #3 0x55160 in Scan__Int () #4 0x54d6c in TextList () #5 0x57ff8 in IntRefTbl__Iterate () #6 0x44654 in FloatMode__BuildConversionArrays () #7 0x24ee4 in RTHeapMap__Walk () #8 0x2507c in RTHeapMap__Walk () #9 0x24d10 in RTHeapMap__BuildGlobalMap () (gdb) That's all. Roumen ======================================================================= 39 === Date: 11 Mar 1994 09:35:19 GMT From: will@unitas.or.uni-bonn.de (Hans-Martin Will) Subject: Re: Modula-3 on Solaris/AIX 3.2.5 This hang-up seems to be independent (more or less) of the underlying architecture as it ressembles the problems I encountered when installing on AIX 3.2.5 on RS/6000 systems. (Dbx tells me m3c the program is frickling around with sbrk ()). Martin will@marvin.or.uni-bonn.de ======================================================================= 40 === Date: Fri, 11 Mar 94 09:45:17 -0800 From: heydon@src.dec.com Subject: Re: Comparing the types of object In message <2lio2a$ajh@porte-de-st-ouen.ics.uci.edu>, Owen O'Malley asks if there is a way to dynamically decide if one object "a" is a subtype of another object "b". Of course, if "b"'s type is known statically to be of some type "T", then you can use the built-in "ISTYPE" operation: ISTYPE(a, T) which is true iff the object "a" is a member of the object or traced reference type "T". If "b"'s type is not static, then you can test the subtype relation at run-time using the "IsSubtype" procedure in the "RTType" interface: RTType.IsSubtype(TYPECODE(a), TYPECODE(b)) This expression is true iff the dynamic type corresponding to "a" is a subtype of the dynamic type corresponding to "b". - Allan Heydon ------------------------------------------------------------------------------- Allan Heydon heydon@src.dec.com DEC Systems Research Center (415) 853-2142 130 Lytton Ave. (415) 853-2104 (FAX) Palo Alto, CA 94301 ======================================================================= 41 === Date: Fri, 11 Mar 94 10:09:23 -0800 From: heydon@src.dec.com Subject: Re: Documentation for netobjs in 3.1? jbpn1@hermes.cam.ac.uk writes: > Is there any documentation for the new NetObj library in release 3.1 > available anywhere? Documentation on network objects is now available in the following SRC Research Report: SRC Research Report 115 "Network Objects" Andrew Birrell, Greg Nelson, Susan Owicki, Edward Wobber The network objects GC algorithm is described in: SRC Research Report 116 "Distributed Garbage Collection for Network Objects" Andrew Birrell, David Evers, Greg Nelson, Susan Owicki, Edward Wobber These reports are available for anonymous FTP from gatekeeper.dec.com in the directory "pub/DEC/SRC/research-reports"; they are named "SRC-115.ps.Z" and "SRC-116.ps.Z", respectively. Hardcopy versions of SRC Research Reports can also be ordered by e-mail; send your request including a postal mail address to "src-reports@src.dec.com". SRC 115 is still being printed, so there may be a slight delay in processing requests for it. - Allan Heydon ------------------------------------------------------------------------------- Allan Heydon heydon@src.dec.com DEC Systems Research Center (415) 853-2142 130 Lytton Ave. (415) 853-2104 (FAX) Palo Alto, CA 94301 ======================================================================= 42 === Date: Sat, 12 Mar 94 12:15:12 EST From: dagenais@vlsi.polymtl.ca (Michel Dagenais) Subject: shared libraries on SUN David Goldberg got all the answers for shared libraries on SUN: From: David Goldberg I got shared libraries to work. Just add m3_option("-X1@-PIC@") m3_option("-X6@-fPIC@") m3_option("-Y3@/tmp/ld-so@") m3_option("-Y4@touch@") m3_option("-X7@-k@") to libm3/src/m3makefile. The results for the Hello World test program compiled with shared vs non-shared libraries are text data bss dec hex 32768 24576 0 57344 e000 SPARC/foo.shared 286720 81920 7816 376456 5be88 SPARC/foo.nonshared The /tmp/ld-so file should contain: #!/bin/csh -f # ld-so : based on version from Olly Stephens (os102@cl.cam.ac.uk) - 20/7/92 set argv[1]=$1:r.so.1.1 # change .a to .so.1.1 /bin/ld -assert pure-text -o $* -david P.S. This is with libraries compiled -g. Compiling -O will perhaps result in even better size savings. ======================================================================= 43 === Date: Sat, 12 Mar 94 11:52:43 EST From: dagenais@vlsi.polymtl.ca (Michel Dagenais) Subject: shared libraries Farshad Nayeri asked about shared libraries: I was finally able to find the relevant documentation on shared libraries under LINUX. It looks a bit painful to build but the resulting executables should link very fast, be tiny and startup rapidly as well. Basically you fix a priori the address of the libraries in virtual space. Someone on the internet must ensure that libraries that may be linked together in a program use different addresses. When you link, you get your own copy of global variables and a jump table that points to the functions in the library (this is much smaller than the library). When the program starts, the dynamic loader insures that the needed libraries are in core and simply maps then in the process virtual space at the predefined addresses; no symbol linking is needed per see. When 3.2 comes out i will build the full LINUX distribution with shared libraries. ======================================================================= 44 === Date: Sat, 12 Mar 1994 16:57:19 GMT From: dagenais@froh.vlsi.polymtl.ca (Michel Dagenais) Subject: Re: Modula-3 on Solaris/AIX 3.2.5 This hang-up seems to be independent (more or less) of the underlying architecture as it ressembles the problems I encountered when installing on AIX 3.2.5 on RS/6000 systems. (Dbx tells me m3c the program is frickling around with sbrk ()). This is probably the big/small endian problem encountered on SPARC as well. Check on ftp.gte.com:pub/m3 for a fix. -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 45 === Date: 14 Mar 1994 08:23:38 GMT From: serrano@savigny.inria.fr (Manuel Serrano) Subject: compilation ? I'm wondering (excuse me if it is a FAQ) if their have been some thesis about the modula3 compilation. In fact, i'm looking for any references on this subject. Thanks to anyone who could help me. ----------------------------------------------------------------------------- Manuel SERRANO (equipe ICSLA, Bat 8) _/_/ _/_/ _/_/_/_/ INRIA-Rocquencourt, BP 105, 78153 _/ _/ _/ _/ _/ Le Chesnay CEDEX, FRANCE _/ _/ _/ _/_/_/ tel: (work) 39-63-57-32 _/ _/ _/ email: Manuel.Serrano@inria.fr _/ _/ _/ _/_/_/_/ ----------------------------------------------------------------------------- ======================================================================= 46 === Date: Mon, 14 Mar 1994 12:27:24 GMT From: klaus@tfl.dk (Klaus Elmquist Nielsen) Subject: Re: Documentation for netobjs in 3.1? jbpn1@hermes.cam.ac.uk (J.B.P. Naylon) writes: >Is there any documentation for the new NetObj library in release 3.1 >available anywhere? gatekeeper.dec.com:/pub/DEC/SRC/research-reports/SRC-115.ps.Z Cheers! Klaus -- Klaus Elmquist Nielsen | Big blues: Tele Danmark Research | Q: Am I uniX? Email: klaus@tdr.dk | A: No! ``Mass production means chemestry, not cookery'', The Chocolate Book (Penguin) ``Hvis Fakta er fakta, s{\aa} er Brugsen fup!'' ======================================================================= 47 === Date: 15 Mar 1994 08:40:25 -0800 From: omalley@porte-de-st-ouen.ics.uci.edu (T. Owen O'Malley) Subject: Trestle question Hi, I've been working on a Trestle application under 2.11 and I've hit a problem that I can't see how to fix (and I've tried). The layout is a ZSplit with a RigidVBT over a TextureVBT as background. Above the background is a leaf painted with PaintOp.TransparentFg. When I originally draw the leaf, it looks fine but when I iconify the window (or strech it) it picks up the pixels underneath the window as the background. So if I've got an xterm where the application's window will uniconify, I'll pickup text in the background of the Leaf. I've tried forcing a repaint of the background (including trying to overridding the ZSplits repaint method), but nothing seems to help. Is this a known bug or is there a fix? Thanks, Owen O'Malley Department of ICS | omalley@ics.uci.edu (ARPA) UC Irvine | "http://www.ics.uci.edu/~omalley/" (WWW) Irvine, CA 92717 | ucbvax!ucivax!omalley (UUCP) ======================================================================= 48 === Date: 15 Mar 1994 21:53:46 GMT From: jsmith@red-branch.MIT.EDU () Subject: Questions, suggestions on porting Mod3 to linux Does anyone out there have a list of reasonable referances to look at, for who is intersted in getting modula3 running on a pc(486)under linux or one of freebsd's?. And or, is anyone out there willing to help with such a project? Jonathan Smith ======================================================================= 49 === Date: 15 Mar 1994 23:29:36 GMT From: jgoldst@jaguar.cs.wisc.edu (Jonathan Goldstein) Subject: Can't compile Modula-3 I am trying to compile Modula-3 on a DEC 3100 running Ultrix I follow the instructions exactly, and after I produce the make for Modula-3, and I try to actually compile Modula-3, I get stuck. (that is, I jump out of the make with Error Code 1). I think it has something to do with all the @ symbols. What are these things supposed to do anyway? Thanks for the help, Jonathan Goldstein (jgoldst@cs.wisc.edu) ======================================================================= 50 === Date: Tue, 15 Mar 94 16:04:38 -0800 From: msm@src.dec.com Subject: Re: Trestle question Not a bug. A ZSplit isn't like a canvas widget; it implements overlapping windows with full clipping. Once you've installed the leaf VBT, it's the only VBT that's allowed to paint anywhere inside its domain. Painting with TransparentFg doesn't make the VBT transparent in the spots where you paint, it just doesn't change what's displayed on the screen at those points where the transparent operation is used. I've been considering implementing a new split which would be more like a canvas, but haven't done so yet. The hard part is that you have to live with the flicker you get from originally painting everything from back-to-front, and that you have to implement a paintbatch method that causes VBTs that might obscure you to repaint themselves when you paint. If you're willing to have a painting model where you don't worry about defining which VBTs paint wins when multiple VBTs paint, it's a lot easier, and you could probably write that yourself in an hour or two. Mark ======================================================================= 51 === Date: 17 Mar 1994 07:02:13 GMT From: jcm@octavia (James McPherson) Subject: Macintosh Port Hi all. I'm really hankering after a Mac port of M3, so I can get some real work done on my machine at home. I've got Symantec C++ 6.0.1, and a fair bit of free space on my hd, but no idea about how to get started. Is __anyone__ working on a mac port? Is there a place where I can get some reasonably idiot-proof instructions/pointers/tips on where/how to start otherwise? ta muchly, jcm -- jcm@octavia.anu.edu.au The Mathematician beats the Chemist and Engineer hands down - "I started with the assumption that all the cans were open, and got the desired result by imbibing." -- other email addresses: jcm@cairo.anu.edu.au ht325689@dingo.cc.uq.oz.au j.mcpherson@mailbox.uq.oz.au -- jcm@octavia.anu.edu.au The Mathematician beats the Chemist and Engineer hands down - "I started with the assumption that all the cans were open, and got the desired result by imbibing." ======================================================================= 52 === Date: Thu, 17 Mar 1994 08:29:24 GMT From: skj@oasis.icl.co.uk (Simon Johnston) Subject: FOR SRC: can we have OS2 please? For the attention of the SRC Modula-3 team: Can we have an OS2 version please, released with WIN32? I am using the Linux version of the 3.1 compiler now and it is great, I have had copies of SPwM3 and the Harbison book for ages and have read them over and over, this is the language for me. I have converted a good few people here onto M3, but our major platform is IBM OS2 2.1, a 32bit OS which is (IMHO) better than the WIN32 offerings (specifically NT). GCC exists for OS2 as a port (GCC 2.5.7 called GCC/2) and as a more specific OS2/DOS version (called EMX based on 2.5.7). Along with a complete set of GNU tools, shells etc OS2 can look very Unix like if you want. Another thing is rather distasteful, can we have, or look at the posibility of *commercial* versions of M3. Our design authorities here like the language and concede that for the work we are doing it would be ideal. However they cannot commit large scale, long term development (on OS2) to a freeware compiler. Any ideas ?? Sorry if all this sounds petty, its just that the language is good, and I want to see it prosper. Thanks. MODULE Sig; FROM ICL IMPORT StdDisclaimer; FROM Interests IMPORT Modula2, Modula3, Linux, OS2; BEGIN (* ------------------------------------------------------------------------. |Simon K. Johnston - Development Engineer |ICL Retail Systems | |------------------------------------------------------|3/4 Willoughby Road| |Unix Mail : S.K.Johnston.bra0801@oasis.icl.co.uk |Bracknell, Berks | |Telephone : +44 (0)344 476320 Fax: +44 (0)344 476084|United Kingdom | |Internal : 7261 6320 OP Mail: S.K.Johnston@BRA0801|RG12 8TJ | `------------------------------------------------------------------------ *) END Sig. ======================================================================= 53 === Date: Thu, 17 Mar 1994 14:13:15 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: Macintosh Port In article <2m8v9l$lq6@manuel.anu.edu.au> jcm@octavia (James McPherson) writes: Hi all. I'm really hankering after a Mac port of M3, so I can get some real work done on my machine at home. I've got Symantec C++ 6.0.1, and a fair bit of free space on my hd, but no idea about how to get started. Is __anyone__ working on a mac port? Is there a place where I can get some reasonably idiot-proof instructions/pointers/tips on where/how to start otherwise? ta muchly, jcm I have thought about a Mac port for a long time, and I will consider helping others who want to try. (I have a Sym. C++ 6.0 as well and a Centris 650 sitting idle at home.) I remember someone else (from UK?) said that they had a partial port of 2.11 working, but I can't seem to find the message. Maybe that someone could post a progress report. Some of the challenges of the Mac port may be: - Threads. Mac's thread support is very picky about preemptive threads. - Changing the "batch-oriented" model or modifying it in someway to work on the Mac. - Mac's IO is very slow. Modula-3 compiling is IO intensive. - Window interaction. - Toolbox calls. A while ago someone posted an article about the need for a tool that takes C headers and maps them to Modula interfaces (partially of course). Such a tool will be very useful for the Mac port. And I think it will help with popularizing Modula-3 since it will make it easier to link with C code quickly. Any other things that I am leaving off? However, one of the greatest things about the Mac port may be that Modula-3 programs will be highly portable between PowerPC and 68k systems. The main question is that is it a good idea to port the current implementation or wait until someone does an implementation that was meant for lower-level, more popular systems. In contrast, and relating to an earlier port, I think an OS/2 port is not nearly as hard. Boy, Modula-3 seems to be taking off a little bit. Two posts to port on two popular platforms... This makes my day. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 54 === Date: Thu, 17 Mar 1994 15:38:02 GMT From: dagenais@froh.vlsi.polymtl.ca (Michel Dagenais) Subject: Re: Questions, suggestions on porting Mod3 to linux Well, there are already binaries for LINUX on ftp.vlsi.polymtl.ca lude/modula3-3.1/run/linux.tar.Z; however, not much else than the compiler and libm3 is there. The next release (soon :-) should contain all the libraries and come with Linux shared libraries. -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 55 === Date: Thu, 17 Mar 1994 15:55:43 GMT From: dagenais@froh.vlsi.polymtl.ca (Michel Dagenais) Subject: Re: FOR SRC: can we have OS2 please? I am using the Linux version of the 3.1 compiler now and it is great, I have had copies of SPwM3 and the Harbison book for ages and have read them over and over, this is the language for me. I have converted a good few people I agree, and once you get to use network objects and the "very cool" Obliq embedded OO language (you can call it from m3 and you can call pretty much all the m3 libraries from it), Modula-3 becomes irresistible! here onto M3, but our major platform is IBM OS2 2.1, a 32bit OS which is (IMHO) better than the WIN32 offerings (specifically NT). GCC exists for OS2 as a port (GCC 2.5.7 called GCC/2) and as a more specific OS2/DOS version (called EMX based on 2.5.7). Along with a complete set of GNU tools, shells etc OS2 can look very Unix like if you want. If you have all of this available, the port should be fairly simple. Another thing is rather distasteful, can we have, or look at the posibility of *commercial* versions of M3. Our design authorities here like the languag e and concede that for the work we are doing it would be ideal. However they cannot commit large scale, long term development (on OS2) to a freeware compiler. Any ideas ?? This is a good point. Frode Odegard (you should be able to find his e-mail in the comp.lang.modula3 archives) was offering commercial support for Modula-3 at some point. Some of the companies specialized in "freeware" support may be willing to offer support contracts. -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 56 === Date: Thu, 17 Mar 1994 14:31:57 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: FOR SRC: can we have OS2 please? In article skj@oasis.icl.co.uk (Simon Johnston) wr ites: For the attention of the SRC Modula-3 team: Can we have an OS2 version please, released with WIN32? I am using the Linux version of the 3.1 compiler now and it is great, I have had copies of SPwM3 and the Harbison book for ages and have read them over and over, this is the language for me. I have converted a good few people here onto M3, but our major platform is IBM OS2 2.1, a 32bit OS which is (IMHO) better than the WIN32 offerings (specifically NT). GCC exists for OS2 as a port (GCC 2.5.7 called GCC/2) and as a more specific OS2/DOS version (called EMX based on 2.5.7). Along with a complete set of GNU tools, shells etc OS2 can look very Unix like if you want. It is probably not that hard since OS/2 is 32 bits and I believe it has thread support, etc. (On thing that will be great and interesting is to see SRC Modula-3 working on top of IBM System Object Model!) Unfortunately, I don't have a PC, so I can't be of much help here. I am not sure if SRC itself needs to do a port, or just someone out there who knows OS/2 well should try do this. I have heard people talking about an OS/2 port before. So far, most of the ports of the 2.11 version have been made by people outside SRC. Would you be able to do such a port? The m3-request folks are very good in helping people with porting and other problems. With release 3.1 it is even easier to port because most/all of the UNIX/Make dependencies have been removed. Another thing is rather distasteful, can we have, or look at the posibility of *commercial* versions of M3. Our design authorities here like the language and concede that for the work we are doing it would be ideal. However they cannot commit large scale, long term development (on OS2) to a freeware compiler. Any ideas ?? Although I've found that the level of quality and support of SRC Modula-3 is better than most commercial packages I've ever used, I agree with you that a second commercial source will be excellent for popularizing Modula-3. The question is that who's going to do it... People have talked about it, but I've never seen one. I even talked with a vendor who was doing an Object Modula-2 on the Macintosh who did not feel that they want to get into doing Modula-3. Even if someone is willing to do an implementation of Modula-3 (though the market share does not seem large) how would they be able to compete with SRC's system? Would you be willing to pay $1000+ for a system that you can get for free. Then again, if there is enough demand, a cygnus-like system may work. They can just use SRC's code, but sell support for it. I heard that SRC is revising their commercial licesnsing agreement for Modula-3 to allow more flexibility. Can someone from SRC elaborate on the intent behind the changes? Sorry if all this sounds petty, its just that the language is good, and I want to see it prosper. I think you are making a good point; I think much of the readership here would want to make Modula prosper. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 57 === Date: 17 Mar 1994 21:47:22 GMT From: dstengle@max (Renata Wentzcovitch) Subject: Re: Questions, suggestions on porting Mod3 to linux jsmith@red-branch.MIT.EDU wrote: : Does anyone out there have a list of reasonable referances to look at, for : who is intersted in getting modula3 running on a pc(486)under linux or one of : freebsd's?. And or, is anyone out there willing to help with such a project? : Jonathan Smith Look on gatekeeper.dec.com for the new version (3.1). This is a native code compiler. Linux was one of the first three platforms to be ported to and has fewer bugs than some others at the moment, having been one of the first. For anyone where they have decided to use Modula-3 for teaching, this is a definite boon, as compared to using sometimes (or always) overloaded facilitie s. David Stenglein ======================================================================= 58 === Date: 17 Mar 1994 21:58:22 GMT From: dstengle@max (Renata Wentzcovitch) Subject: installing 3.1 on IBM RS/6000 Simply put, bas anyone acomplished this? There are a few errors, including looking for a lib file that isn't there. The make process hangs and says it doesn't know how to make m3.o. It would be great just to that it was possible to do. David Stenglein ======================================================================= 59 === Date: Thu, 17 Mar 94 13:57:38 PST From: frode@toaster.SFSU.EDU (Frode Odegard) Subject: commercial support Michel Dagenais writes: > This is a good point. Frode Odegard (you should be able to find his e-mail > in the comp.lang.modula3 archives) was offering commercial support for > Modula-3 at some point. Heh, okay, it's frode@odegard.com. :) Odegard Labs is still in contract negotiations with DEC SRC. If things work out you should be able to buy a commercially supported version from us for SPARC later this year (there is no set release date yet.). Various PC OS'es are being looked at also, since we have two clients that have massive amounts of Modula-2 code and want to move to Modula-3 (one is under OS/2, the other under DOS.) A problem with this is that if you support some platform FOO and only two people are interested, you're not going to make any money. So the plan was to get a solid legal arrangement with DEC SRC (this has taken months and months because the DEC SRC legal staff was laid off) and then see how the "Modula-3 marketplace" had developed at that point. (SPARC is the easiest platform for us initially because that's the hardware we and lots of other people have. :->) Now that I'm writing this note anyway, perhaps those who are interested in some form of commercial support can send a note to me at the above- mentioned address with the subject line "M-3 Poll" and give me an idea of what they're interested in (platform, features, bin vs src, pricing, levels of support, porting, etc.). I'll then summarize and post the result here. If nobody's interested then I hope we won't see any more "whining" about how Modula-3 doesn't have any commercial backing. :-) Frode Odegard ======================================================================= 60 === Date: Fri, 18 Mar 1994 04:10:34 GMT From: torki@mpipf-muenchen.mpg.de (Torsten R. Kirschner) Subject: Q: M3 3.1 formsvbt package / videovbt? Hello, when I tried to install the formsvbt package from the M3 3.1 release, I ran into this problem. Here are some lines from the "m3makefile.formsvbt": import (ui) import (videovbt) import (vbtkit) import (formsvbtpixmaps) What is videovbt and where do I get this library from? Either I have over- looked it in the distribution or it is just not there. I have all the other libs installed. TIA for any help, Torsten -- Torsten R. Kirschner (EDV) torki@mpipf-muenchen.mpg.de MPI f. Psychologische Forschung tel: ++49 89 38602258 Leopoldstr. 24 fax: ++49 89 342473 80802 Muenchen Germany X.400: no longer available [PGP 2.3A 35AABD 1993/12/15 Torsten R. Kirschner ] Torsten ======================================================================= 61 === Date: Fri, 18 Mar 1994 13:14:25 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: Q: M3 3.1 formsvbt package / videovbt? In article torki@mpipf-muenchen.mpg.de (Torsten R. Kirschner) writes: Hello, when I tried to install the formsvbt package from the M3 3.1 release, I ran into this problem. Here are some lines from the "m3makefile.formsvbt": import (ui) > import (videovbt) import (vbtkit) import (formsvbtpixmaps) What is videovbt and where do I get this library from? Either I have over- looked it in the distribution or it is just not there. I have all the other libs installed. VideoVBT is an extension for multimedia VBTs. Since it relies on SRC hardware, it was not included in the release. I have created a "fake" version which you can find ftp.gte.com:/pub/m3/m3-sparc/19. I believe version 3.2 will fix the dependencies so that you can compile without problems. Note that if you use my "fake" videovbt and you (mistakenly) access a videovbt you will raise either either a nil dereference or a NonImplemented execption. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 62 === Date: 18 Mar 1994 11:45:57 -0600 From: 0006378709@mcimail.com (Ward Harold) Subject: NT version? An NT version of Modula-3 was rumored to be under development by the SRC people. Does anyone care to confirm or deny? For what it's worth I'd be happy to beta test. Thanks ... WkH Ward K. Harold wkh@ACM.org MCI Consumer Markets (512) 416-4356 4115 Freidrich Lane Austin, TX 78744 ======================================================================= 63 === Date: Fri, 18 Mar 1994 13:15:29 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: Q: M3 3.1 formsvbt package / videovbt? [sorry if this was posted twice.] In article torki@mpipf-muenchen.mpg.de (Torsten R. Kirschner) writes: Hello, when I tried to install the formsvbt package from the M3 3.1 release, I ran into this problem. Here are some lines from the "m3makefile.formsvbt": import (ui) > import (videovbt) import (vbtkit) import (formsvbtpixmaps) What is videovbt and where do I get this library from? Either I have over- looked it in the distribution or it is just not there. I have all the other libs installed. VideoVBT is an extension for multimedia VBTs. Since it relies on SRC hardware, it was not included in the release. I have created a "fake" version which you can find ftp.gte.com:/pub/m3/m3-sparc/19. I believe version 3.2 will fix the dependencies so that you can compile without problems. Note that if you use my "fake" videovbt and you (mistakenly) access a videovbt you will raise either either a nil dereference or a NotImplemented exception. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 64 === Date: 18 Mar 1994 15:14:12 GMT From: hendrik@CAM.ORG (Hendrik Boom) Subject: Re: FOR SRC: can we have OS2 please? Farshad Nayeri (fn00@gte.com) wrote: : In article skj@oasis.icl.co.uk (Simon Johnston) writes: : For the attention of the SRC Modula-3 team: : Can we have an OS2 version please, released with WIN32? I'm not on the SRC team, but have been planning to start an OS/2 port for a while now. I managed to get a version 2.somethine port for DOS, and its source, and was getting ready to port from that, using C as cross-platform bootstrapping tool and thereby avoiding much of the hassle with object file formats, etc. Now that 3.1 is out, it's not as clear what I should be doing. Probably downloading the lot and thinking a bit longer before I start. One question I'd like to have answered: Will the 3.1 code compile using any of the released 2.xx compilers? (my guess: it will almost compile, and is probably patchable) Well, two questions: Is there a configuration of 3.1 that will generate C code? (my guess: no) I will be setting up an OS/2 modula 3 porters' mailing list as soon as my home site's net feed is regularized. I currently use hendrik@cam.org and potential porters should write me here. I'l post any relevant mail on the mailing list when it is ready unless you ask me not to. : I am using the Linux version of the 3.1 compiler now and it is : great, I have had copies of SPwM3 and the Harbison book for ages : and have read them over and over, this is the language for me. I The Linux-to-os/2 route seems to be the way to go, but I do not have Linix installed. Does anyone have any suggestions where the best place is to get the parts of Linux I need for a Modula 3 port wiht the minimum of installation effort? : have converted a good few people here onto M3, but our major : platform is IBM OS2 2.1, a 32bit OS which is (IMHO) better than the : WIN32 offerings (specifically NT). GCC exists for OS2 as a port : (GCC 2.5.7 called GCC/2) and as a more specific OS2/DOS version : (called EMX based on 2.5.7). Along with a complete set of GNU : tools, shells etc OS2 can look very Unix like if you want. : It is probably not that hard since OS/2 is 32 bits and I believe it : has thread support, etc. (On thing that will be great and interesting : is to see SRC Modula-3 working on top of IBM System Object Model!) : Unfortunately, I don't have a PC, so I can't be of much help here. : I am not sure if SRC itself needs to do a port, or just someone out : there who knows OS/2 well should try do this. I have heard people Well I'm a someone out there. : talking about an OS/2 port before. So far, most of the ports of the : 2.11 version have been made by people outside SRC. Would you be able : to do such a port? The m3-request folks are very good in helping : people with porting and other problems. With release 3.1 it is even : easier to port because most/all of the UNIX/Make dependencies have : been removed. : Another thing is rather distasteful, can we have, or look at the : posibility of *commercial* versions of M3. Our design authorities : here like the language and concede that for the work we are doing : it would be ideal. However they cannot commit large scale, long : term development (on OS2) to a freeware compiler. Any ideas ?? : Although I've found that the level of quality and support of SRC : Modula-3 is better than most commercial packages I've ever used, I : agree with you that a second commercial source will be excellent for : popularizing Modula-3. The question is that who's going to do it... : People have talked about it, but I've never seen one. I even talked : with a vendor who was doing an Object Modula-2 on the Macintosh who : did not feel that they want to get into doing Modula-3. Even if : someone is willing to do an implementation of Modula-3 (though the : market share does not seem large) how would they be able to compete : with SRC's system? Would you be willing to pay $1000+ for a system : that you can get for free. Ah! but is it *really* the same system? I've found the CSET++ to be much better than gcc++, and well worth the introductory price. I could see doing a p.d.port just to make the language available, and then elaborately doing a non-p.d. port to pay the bills. The p.d. port would sell the language, and the non-p.d. port would sell quality-of-implentation and OS integration. : Then again, if there is enough demand, a cygnus-like system may work. : They can just use SRC's code, but sell support for it. I heard that : SRC is revising their commercial licesnsing agreement for Modula-3 to : allow more flexibility. Can someone from SRC elaborate on the intent : behind the changes? : Sorry if all this sounds petty, its just that the language is good, : and I want to see it prosper. : I think you are making a good point; I think much of the readership : here would want to make Modula prosper. : --farshad : -- : /// | : {@,@} | "He's so unhip that when you : (...) Farshad Nayeri | say Dylan, he thinks you're : " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 65 === Date: 19 Mar 1994 00:32:27 GMT From: rrw1000@cus.cam.ac.uk (Richard Watts) Subject: Re: Questions, suggestions on porting Mod3 to linux Does the version on vlsi.polymtl.ca have working threads ? I've just tried to get it to check, but the network transfer speed from here to vlsi is fairly low, and I didn't want to clog up the net all night. If anyone's got threads working properly (if the program appended below doesn't segfault, that is), could they please email me ? I've been trying to get the thread support working under Linux for a while now. It's also probably worth mentioning that the unix interfaces for Linux disagree with the Linux header files - I'm reworking them for Linux 1.0. Is anyone out there also doing either of these ? It'd be nice to get together and pool our efforts. Richard. (test program : MODULE Hello EXPORTS Main; IMPORT Wr, Stdio, Thread, Fmt, Rd; TYPE ch = Thread.Closure OBJECT ms : TEXT; OVERRIDES apply := SayCheese; END; PROCEDURE SayCheese(self : ch ): REFANY RAISES {} = VAR a: TEXT; BEGIN Wr.PutText(Stdio.stdout, self.ms ); Wr.Flush(Stdio.stdout); LOOP (* Wr.PutText(Stdio.stdout, self.ms); *) END; RETURN self; END SayCheese; VAR i : INTEGER; BEGIN (* Wr.PutText(Stdio.stdout," Type a line : \n"); Wr.Flush(Stdio.stdout); EVAL Rd.GetLine(Stdio.stdin); *) Wr.PutText(Stdio.stdout, " Integer byte size is " & Fmt.Int(BYTESIZE(INTEGER)) &"\n"); Wr.Flush(Stdio.stdout); FOR i := 1 TO 23 DO EVAL Thread.Fork(NEW(ch, ms := "Cheese " & Fmt.Int(i))); Wr.PutText(Stdio.stdout, "Just forked " & Fmt.Int(i) & "\n"); Wr.Flush(Stdio.stdout); END; LOOP END; END Hello. ) ======================================================================= 66 === Date: 19 Mar 1994 03:11:08 GMT From: mako@FNALF.FNAL.GOV Subject: Re: FOR SRC: can we have OS2 please? > I managed to get a version 2.somethine port for DOS, > and its source, and was getting ready to port from that, using C as > cross-platform bootstrapping tool and thereby avoiding much of the > hassle with object file formats, etc. > > Now that 3.1 is out, it's not as clear what I should be doing. Probably > downloading the lot and thinking a bit longer before I start. > > One question I'd like to have answered: > > Will the 3.1 code compile using any of the released 2.xx compilers? > (my guess: it will almost compile, and is probably patchable) > Well, two questions: > Is there a configuration of 3.1 that will generate C code? > (my guess: no) Wouldn't it be easier to use IL (those generated in .ic/.mc files) + m3cg (the gcc backend) rather than trying to use 2.xx + a C compiler as a cross- platform bootstrapping tool? You can keep these files by uncommenting the line m3_option("-keep") in the make file (m3makefile). mako ======================================================================= 67 === Date: Sat, 19 Mar 1994 05:54:09 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: Q: M3 3.1 formsvbt package / videovbt? (uuencode, 98 lines) Due to popular demand, here is a fake videovbt package for those of you who can't wait for 3.2 (for good reasons, I am sure. :-) This and other patches to bring up version 3.1 can be found in ftp.gte.com:/pub/m3/m3-sparc. Some patches are sparc-specific while others should be made in other configurations as well. --farshad begin 644 videovbt.tar.gz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end -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 68 === Date: Sun, 20 Mar 1994 05:58:40 GMT From: k2cheung@undergrad.math.uwaterloo.ca (Cheung Kai Chung) Subject: source code? Hello. Can someone point out to me where I can find a modula 3 compiler source code? i.e. the source code of the c (or c++?) program which generate c++ code from modula 3 programs. Any help is greatly appreciated. Thanks in advance. Best regards, William. ======================================================================= 69 === Date: 21 Mar 1994 11:18:06 GMT From: toshok@cs.uidaho.edu (Chris Toshok) Subject: Problem with TextPort filters Hello all, I am having a problem with SRC Modula-3 release 3.1. I try to do the following: RETURN NEW(FVTypes.FVTextEdit, tp:= NEW(TextPort.T, filter:= MyFilter)) where MyFilter is defined as: PROCEDURE MyFilter(VAR cd : VBT.KeyRec) = BEGIN < code deleted > END MyFilter; and modula-3 gives me the following errors: new source -> compiling ../src/Main.m3 "../src/Main.m3", line 988: default is incompatible with method type (filter ) "../src/Main.m3", line 988: procedure is not compatible with method (filter) 2 errors encountered Why? I really need this code to compile. It used to, with release 2.11 (whichever was the last release), but it doesn't anymore. Help. -- ------------------------------+----------------------------------- Chris Toshok | Assistant Systems Administrator a.k.a. Toshok the Functional | Computer Services | University of Idaho | email: toshok@cs.uidaho.edu | Research Assistant phone: (208)885-5562 | Laboratory for Applied Logic smail: P.O. Box 3900 | University of Idaho Moscow, Id 83843-1919 | | Vice Pres. Student Chapter, ACM ------------------------------+----------------------------------- www: http://www.cs.uidaho.edu/lal/students/toshok.dir/toshok.html ------------------------------------------------------------------ ======================================================================= 70 === Date: 21 Mar 1994 06:23:26 GMT From: luca@src.dec.com (Luca Cardelli) Subject: Announcement: Obliq paper Obliq is a distributed scripting language. It is based on Modula-3 Network Objects, and was designed to be embedded in Modula-3 applications. It has now evolved into a self-contained language, whose direct ties to Modula-3 are limited to the network objects run-time, and to a rich collection of libraries. The following paper can be obtained via ftp; instructions are at the bottom. The Obliq software distribution (matching the paper) will be part Modula-3 v3.2. The current distribution of Modula-3 v3.1 includes a slightly older version of Obliq. All these are or will be available from gatekeeper.pa.dec.com Obliq: A Language with Distributed Scope Luca Cardelli Digital, Systems Research Center Abstract Obliq is a lexically-scoped untyped language that supports distributed object-oriented computation. An Obliq computation may involve multiple threads of control within an address space, multiple address spaces on a machine, heterogeneous machines over a local network, and multiple networks over the Internet. Objects are local to a site, while computations can roam over the network. The most novel feature of Obliq is its combination of distributed computation and lexical scoping. In lexically scoped languages, the binding location of every identifier is determined by simple analysis of the program text. Therefore, one can be sure of the meaning of program identifiers, and can much more easily reason about the behavior of programs. In a distributed language like Obliq, lexical scoping assumes a further role. It ensures that computations have a precise meaning even when they migrate over the network: a meaning that is determined by the original binding location and network site of identifiers, and not by their execution sites. Network-wide scoping becomes an issue in presence of higher-order distributed computation, as in the situation of remote sites accepting procedures for execution; the typical case being compute servers. The question is what happens to the free identifiers of such network-transmitted procedures. Obliq takes the view that such identifiers are bound to their original locations, as prescribed by lexical scoping, even when these locations belong to different network sites. Distributed computation interacts well with the notion of objects. For example, network services normally accept a variety of messages; it is then natural to see each service as a network object (or, more neutrally, as a network interface). Obliq supports objects in this spirit, relying for its implementation on Modula-3's network objects. The Obliq object primitives are designed to be simple and powerful; a main trait of the language is the relation between local and distributed semantics of these primitives. Obliq objects are collections of named fields, with four basic operations: selection/invocation, update/override, cloning, and delegation. Every object is potentially and transparently a network object. An object may become accessible over the network either by the mediation of a name server, or simply by being used as the argument or result of a remote method. In addition to the distribution of objects, the distribution of computations must be designed carefully. It is clearly desirable to be able transmit agents for remote execution. However, in general, one should not be satisfied with transmitting just the program text of such agents. Program text cannot carry with it live connections to its originating site, nor to any data or service at any other site. Hence the process of transmitting program text over the network implies a complete network disconnect from the currently running distributed computation. In addition, unpredictable dynamic scoping results from transmitting and then running program text containing free identifiers. Obliq computations, in the form of procedures or methods, can be freely transmitted over the network. Actual computations (i.e. closures, not source text) are transmitted; lexically scoped free identifiers retain their bindings to the originating sites. Through these free identifiers, migrating computations can maintain connections to objects and locations residing at various sites. Disconnected agents can still be represented as procedures with no free identifiers. >From the point of view of a server executing a foreign agent, distributed lexical scoping and strong run-time typing together provide a safety and security guarantee. Namely, the foreign agent has access only to the data that it carries with it via lexical bindings, or that it explicitly receives in the form of parameters. Contents 1. Introduction 1.1 Language Overview 1.2 Distributed Semantics 2. Local Objects 2.1 Objects and their Fields 2.2 Object Operations 2.3 Simple Examples 3. Remote Objects 3.1 State 3.2 Transmission 3.3 Distributed Computation 3.4 Self-inflicted Operations 3.5 Protected Objects 3.6 Serialized Objects 3.7 Name Servers 3.8 Execution Engines 4. Local Techniques 4.1 Recursion and Iteration 4.2 The Object-Oriented Numerals 4.3 The Prime Numbers Sieve 4.4 A Calculator 4.5 Surrogates 5. Distributed Techniques 5.1 A Serialized Queue 5.2 A Compute Server 5.3 Remote Agents 5.4 Agent Migration 5.5 Object Migration 6. Conclusions 6.1 Future Work 6.2 Acknowledgments 7. Syntax overview A. Language Reference A.1 Syntactic Structures A.1.1 Identifiers A.1.2 Definitions A.1.3 Terms A.1.4 Term Sequences A.1.5 Built-In Operators A.1.6 Operator Precedence A.2 Data Structures A.2.1 Value Identity A.2.2 Constants A.2.3 Operators A.2.4 Arrays A.2.5 Options A.2.6 Objects A.2.7 Protection and Serialization A.2.8 Object and Engine Servers A.3 Control Structures A.3.1 Definitions A.3.2 Assignment A.3.3 Sequencing A.3.4 Procedures A.3.5 Conditionals A.3.6 Case A.3.7 Iteration A.3.8 Concurrency A.3.9 Exceptions A.3.10 Errors A.4 Methodology A.4.1 Type Comments A.5 Lexicon A.6 Syntax B. System Reference B.1 The Executables B.2 The Top-Level B.3 Program Files B.4 Modules B.5 The Network Objects Daemon B.6 Built-in Modules B.6.1 Sys B.6.2 Bool B.6.3 Int B.6.4 Real B.6.5 Math B.6.6 Ascii B.6.7 Text B.6.8 Array B.6.9 Net B.6.10 Thread B.6.11 Rd B.6.12 Wr B.6.13 Pickle B.6.14 Lex B.6.15 Fmt B.6.16 Process B.6.17 Color B.6.18 Form C. Programming Reference C.1 The Package Hierarchy C.2 The Interfaces C.3 The Libraries C.4 Embedding Obliq C.5 Extending Obliq C.6 Extending Obliq with new Packages C.7 Building a Customized Obliq Interpreter References Index ------------------------------------------------ ftp gatekeeper.pa.dec.com anonymous cd pub/DEC/SRC/personal/luca binary get ObliqPaper.ps.Z quit uncompress ObliqPaper.ps.Z Print Obliq.ps on Apple or HP postscript printers, for best results. Online postscript viewers do not necessarily work properly on this file (although it is legal postscript). Luca ======================================================================= 71 === Date: 21 Mar 1994 15:29:59 GMT From: jason@convex.com (Jason L. Eckhardt) Subject: M3 for i860 problems I have just recently picked up the boot-OKI tar for the i860 version of SRC M3. I actually have a Stardent platform rather than an OKI platform, but they both are i860's running SVR4. When m3boot tries to build the assembly files, many of them fail with assembler errors: "error: st.l misaligned: 0xfff9." The lines that produce the errors are always stores like this: st.l %r0, [-7]@l(%r31). The -7 is sometimes -1, -3, or -9 as well. Can anyone tell me how to assemble these properly? It does seem that -7 displaced from %r31 would be unaligned. Is this possible a gcc compiler error? (The assembly files have gcc2_compiled: in them). Or if anyone knows of an already built binary for the i860, that would be great. thanks, jason. jason@wagner.convex.com ======================================================================= 72 === Date: Mon, 21 Mar 94 08:46:55 -0800 From: mhb@src.dec.com ("Marc H. Brown") Subject: Re: Problem with TextPort filters In FVTypes.i3, the FVTextEdit is defined to take a Port. A Port is a subtype of TextPort.T with the following comment: (* If you create a subtype of "FVTextEdit", its ".tp" field must be "NIL" or a subtype of "FVTypes.Port". *) Thus, if you change your code from RETURN NEW(FVTypes.FVTextEdit, tp:= NEW(TextPort.T, filter:= MyFilter)) to RETURN NEW(FVTypes.FVTextEdit, tp:= NEW(FVTypes.Port, filter:= MyFilter)) I think you'll be in business. You should also check TextPort's filter methd to be sure that your procedure MyFilter has the correct signature. Sorry for the confusion. There are probably a host of other incompatible changes that were made between 2.11 and 3.0 in various libraries. Hopefully, you'll find the changes make life in M3land sufficiently better that you'll endure the conversion hassles with a smile! :-) Marc ======================================================================= 73 === Date: Mon, 21 Mar 1994 14:03:44 GMT From: dagenais@gutrune.vlsi.polymtl.ca (Michel Dagenais) Subject: Re: source code? Hello. Can someone point out to me where I can find a modula 3 compiler source code? i.e. the source code of the c (or c++?) program which generate c++ code from modula 3 programs. The latest release is a Modula-3 program that generates an intermediate language passed to a special version of gcc, from Modula-3 programs. The source code is on gatekeeper.dec.com:pub/DEC/Modula-3. -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 74 === Date: 21 Mar 1994 13:32:37 GMT From: hendrik@CAM.ORG (Hendrik Boom) Subject: Re: FOR SRC: can we have OS2 please? mako@FNALF.FNAL.GOV wrote: : Wouldn't it be easier to use IL (those generated in .ic/.mc files) + m3cg : (the gcc backend) rather than trying to use 2.xx + a C compiler as a cross- : platform bootstrapping tool? You can keep these files by uncommenting the : line m3_option("-keep") in the make file (m3makefile). It would be if I had a working Modula 3 version 3.1 on a system I had access to. Without that, it's a bit hard to generate IL. Or is there already a 3.1 DOS port to work from? : mako ======================================================================= 75 === Date: Mon, 21 Mar 1994 20:15:20 GMT From: jmeehan@mendoza.adobe.com (Jim Meehan) Subject: Re: Problem with TextPort filters PROCEDURE MyFilter(VAR cd : VBT.KeyRec) = BEGIN < code deleted > END MyFilter; MyFilter should be a procedure of two arguments. 'cd' is the second. As Marc points out, the first should be a subtype of FVTypes.Port. ======================================================================= 76 === Date: 21 Mar 1994 20:51:42 GMT From: gwyant@cloyd.East.Sun.COM (Geoffrey Wyant - Sun Microsystems Labs BOS) Subject: Modula-3 User's group - Initial Call for Participation Modula-3 User's Group '94 - Initial Call for Participation ----------------------------------------------------------- Now that the much anticipated 3.x release is out, it seems like a good time to hold another Modula-3 User's Group Meeting. As many of you know, the last one was in 1992 and I think all of the attendee's found it worthwhile to attend. I'm willing to organize this if there are enough people interested in attending. Of course, willingness to attend is always predicated on date and location. Here are some possible dates and locations for it: Summer Usenix - Boston, Ma June 6-10 ASPLOS VI - San Jose October 4-7 OOPSLA '94 - Portland, Or October 23-28 ACM MultiMedia San Francisco October ??? The format of the meeting would be much the same as the previous one; more or less a workshop style with presentations and informal discussion groups. If you are interested in attending please send mail directly to me with your preference of date/location. I'll tally the interest level and make another posting. If you are interested in presenting something informally, please let me know as well ! Geoff Wyant Geoff.Wyant@east.sun.com Sun Microsystems Laboratories, Inc. 2 Elizabeth Drive Chelmsford, Ma. 01824 -- Geoff Wyant Geoff.Wyant@east.sun.com Sun Microsystems Laboratories, Inc. 2 Elizabeth Drive Chelmsford, Ma. 01824 ======================================================================= 77 === Date: 21 Mar 94 21:50:49 GMT From: harbison@tartan.com (Sam Harbison) Subject: Harbison's "Modula-3": errata. And, new edition? The errata list "beethoven.cs.colorado.edu:pub/M3/harbison-errs.ps.Z" appears to be the most complete listing. I noticed only the following omissions: p. 28:-5: Change "before that of A" to "before that of a". Also, interface declarations must have constant initializers, which is not true of blocks. p. 32:22: Change "A, B, C, and D" to "A, B, C, D, and E". p. 100:9: Add to end of first paragraph: "Indexing expressions must be assignable to the index type of the array. Using an index that is outside the range of the index type is a checked run-time error." p. 116: In the railroad diagram, change "ALL" to "ANY". Also, eliminate the semicolon at the end of ProcDecl; the semicolon is shown on p. 20. --------- This seems to be a good time to raise the issue of doing a new edition of "Modula-3." The outstanding errors can be fixed, of course, but it would be good to think of new material to add. So, I'd like to solicit readers of this newsgroup to tell me: - What new material would you like to see in a 2nd edition? A more complete description of SRC Modula-3? Descriptions of new libraries? (Basic ones, or also more advanced libraries like net objects?) - What existing material do you think could be improved or expanded upon? - How are you using the book now? Teaching, learning, referencing? And, finally, if you have a big wish list, would you be willing to help with the revision? If you respond directly to me, I'll collect and repost the suggestions. Sam Harbison, Director of Commercial Products Tartan, Inc.; 300 Oxford Drive; Monroeville, PA 15146; USA Phone +1 (412) 856-3600 x147. Fax +1 (412)856-3636. Email: harbison@tartan.com -- Sam Harbison, Director of Commercial Products Tartan, Inc.; 300 Oxford Drive; Monroeville, PA 15146; USA Phone +1 (412) 856-3600 x147. Fax +1 (412)856-3636. Email: harbison@tartan.com ======================================================================= 78 === Date: Wed, 23 Mar 1994 13:28:35 GMT From: shang@corp.mot.com (David L. Shang) Subject: Who Win the War of Co. vs. Contra? Alas, I have to give up in trying to let my pony be an animal. Why cann't I prohibit somebody to feed my pony with flesh just because pony is an animal and flesh is kind of animal food? I have to say: my pony is not an animal! I won't reuse animal definition for my pony! What's wrong? Let's have an example: class Animal { function eat (food: AnimalFood); }; Class Animal has a method eat, which takes animal food as its input. In a herbivore animal class, the animal food should be specialized to plant food: class HerbivoreAnimal is an Animal { function eat (food:PlantFood); }; Now we have an animal variable: var anAnimal: Animal; When the varaible refers to some animal, we can feed it: anAnimal.eat(flesh); Is it right? Sure, it is. The variable anAnimal is declared as an Animal, and the eat interface in class Animal says that any animal food is okay for its input. Again, is it right? No, it is not. If the variable anAnimal happens to take a herbivore animal, oops, it's going to choke. Solution 1: Let Herbivore Animal Eat Flesh Some people argue that it is not right to specialize the food type for a herbivore animal to eat. If you do not specialize the food, there is no problem in your program then. Languages like C++ and Modula-3 prohibit such specialization. Why no problem now? Because you let your herbivore animal be able to eat flesh! Yes, there is no problem with my program. But there is a problem with my herbivore animal. My program does not reflect the real world. Solution 2: Let Herbivore Animal Eat Anything Some languages even go to an opposite extreme. You cannot specialize the food type, but you can generalize the food type. In class HerbivoreAnimal, you can specify its food as any thing, but you cannot limit it to specific animal food. It doesn't matter even if it is not food at all. Languages like Sather, POOL-I, and Trellis let you do this way. This approach leads to a long debate between covariance and contravariance. Do you feel that contravariance is better than novariance in languages like C++ or Modula-3? I don't. By contravariance, now I can let my herbivore animal be able to eat any thing: stone, knife, cannon, needless to say flesh. Solution 3: Animal Cannot Eat Some people said that you should not declare eat method in animal. It should be declared in subclasses respectively. Then, eat is not a property of an animal, that is, an animal cannot eat! Obviously, this is not a solution. It eliminates many opportunities of code reuse. Solution 4: Herbivore Animal Is Not an Animal Then, some other people said that a herbivore animal is not an animal because it does not eat something that some other animals eat. What? You may not accept this argument. But wait, there is a compromise: a herbivore animal is not a subtype of animal but they can still be a subclass of animal. To be more specific, the approach tells you that in the case of a covariantly typed methods, the subclass is not a subtype. Therefore, a herbivore animal is not a subtype of animal, but it can still reuse the animal code by being a subclass. The F-bounded polymorphism is established on this belief. A number of languages are designed on this thoery. ABEL and TOOPL are examples. The latest version of Eiffel will tell you that subclasses which are subtypes by declaration but not in terms of the interface which is supported. Being not a subtype of animal, a herbivore animal cannot be assigned to a animal variable. An animal cannot represents a herbivore animal. After all, a herbivore animal is not an animal. You cannot have a polymorphic all-animal interface that can be shared and reused by other software components which are supposed independent to a concrete animal. This certainly affects the software reusability. Separating the subtype concept from subclass signaficantly complicates the concept. When the compiler tells you that a herbivore animal is not a subtype of an animal, you'll definitely have a hard time to figure it out why not. This approach also complicates the language compiler. It requires system-level type checking, which gives up on local checkability. It has been proved that to determine whether a type is a subtype of another is undecidable in general. It is an open question that whether this approach is able to work on the open world assumption, the essential assumption for developing large-scale software. Eifel's system-level type-checking works only on the closed world assumption, an assumption usually works for small toy programs. The Right Solution: Let the Food Type Depend on the Animal Type A herbivore animal should eat pland food only. A herbivore animal should be an animal and a animal variable should be able to represent a herbivore animal. It is not the problem that we specialize the food in herbivore animal, nor the problem that we consider a herbivore animal being an animal. The problem is the eat method interface defined in Animal class. It is not a precise interface. Given a class C and a protocol (method interface) P defined in C, for each object O that belongs to C, P should be applicable to O. Obviously, the method eat defined in class Animal is not applicable to each individual animal because it can accept any type of food. The problem comes from the type intra-dependency, which has not been fully recognized so far. Oftentimes, the type of a component or the protocol of a method interface depends on the type of the enclosing object. I call this phenomenon as type intra-dependency (refer to my paper: Type-Safe Redefinition in Proc. of 3rd Intl. Conf. on Software Engineering and Knowledge Engineering for detail). The food class in method eat should be specified as a variable to the class of Animal: class Animal { function eat (food: thisclass'Food); }; that is, the food for the animal to eat should be of the type edible to the animal, instead of any food. When subclasses are derived, the class of food is automatically redefined in covariance with animal classes. Since the interface is accurate, local type checkability is always ensured. A language named Cluster can reflect this dependency very well. The language was first introduced in SIGPLAN Notices, Jan. 1991. At the time I wrote this paper, I did not realized the covariance problem yet. Therefore, when you read the paper, you may not find the direct anwser to covariance, but the answer is there. I have recently finished a paper: Covariance Spsecification. If anybody is interested, contact me. David Shang ======================================================================= 79 === Date: Wed, 23 Mar 94 16:40:42 EST From: gwyant@cloyd.East.Sun.COM (Geoffrey Wyant - Sun Microsystems Labs BOS) Subject: Converting Capability.T to text ? Before I implement this myself, I thought I'd check to see if anyone has already done this and might be willing to share.. --geoff Geoff Wyant Geoff.Wyant@east.sun.com Sun Microsystems Laboratories, Inc. 2 Elizabeth Drive Chelmsford, Ma. 01824 ======================================================================= 80 === Date: 23 Mar 94 18:16:39 From: jeff@karazm.math.uh.edu (Jeff M Younker) Subject: Re: Who Win the War of Co. vs. Contra? This discussion seems to be ignoring how things are actually done in Sather. We have paramaterized classes. We also have a difference between inheritence and implementation. I think this code should do everything that people are arguing about. It should also be legal in Sather. The first irst thing we need to do is to define what foods are. Its a pretty broad concept, but we can break them all down to nutrients. Now we need to define what a nutrient is. Since everything breaks down to the same amino acids and protiens, we'll just use a single nutrient class. No comments about silicon based life please. And we'll assume water and not ammonia for the primary liquid. ***** class NUTRIENT is ... amount:INT is ... end; create(INT):SAME is ... end; ... end; -- class NUTRIENT ******* Now that we have nutrients, lets make foods. ******* abstract class $FOODS is breakdown:NUTRIENT is end; end; -- abstract class $FOODS ****** Now lets do some organization. Most of what we are talking about are living things, so we'll make a generic class to organize the whole thing. Something very important is that all living things can be eaten. They are all foods. But notice that all foods are not living things. Think about twinkies and velveta cheese slices if you have any doubt of this. ****** abstract class $LIVING_THINGS < $FOODS is include $FOODS; -- we get all the definitions of what food is. -- living things have some other characteristics also... -- metabolize(NUTRIENT) is end; reproduce(SAME):ARRAY{SAME} is end; die:NUTRIENT is end; end; ****** Plants are living things. By the very nature of living things plants are also foods. Were also going to ignore plant foods for now. Carnivorous plants are also ignored. We can add them later. What we want to talk about today are animals. ****** abstract class $PLANTS < $LIVING_THINGS is include $LIVING_THINGS; -- get default actions for living things $FOODS -- now we implement some additional plant actions... -- create(INT):SAME is ... end; pick:SAME is end; ferment:ALCOHOL is end; end; ***** Dung isn't very interesting to me, but some beetles like it, so they get to eat it too. And I don't want to model decay, so lets just leave one kind of turd in the world. ***** class EXCREMENT < $FOODS is ... breakdown:NUTRIENT is ... end; turd_size:INT is ... end; create(size:NUTRIENT):SAME is .. end; ... end; -- class EXCREMENT ***** Now we seem to have enough things that we can really start making sense with animals. Note that all animals are also foods. ***** abstract class $ANIMAL{FOOD<$FOODS} < $LIVING_THINGS is include $LIVING_THINGS; ... create(size:INT):SAME is ... end; make_sound is end; eat(FOOD):EXCREMENT is end; vomit(dinner:$FOODS):VOMIT{$FOODS} -- all animals can get the same behavior for vomiting is return #(dinner); end; is_this_digestable_food($FOODS}:BOOL is end; ... end; -- abstract class $ANIMAL ***** Now lets make a general class of animals. Lets call them herbivores. The special thing about herbivores is that they can eat only plants, so we build in this restriction. However we need to notice that all herbivores can not eat all plants. If we feed grass to a koala it's going to starve to death. By the same token, flossy the cow is going to be poisoned by eucalyptis leaves. So it makes a great deal of sense that we are not allowed to instantiate a generic herbivore. The herbivore class just provides a generic framework for a group of animals. ***** abstract class $HERBIVORE{PLANT<$PLANTS} < $ANIMAL{PLANT} is include $ANIMAL{PLANT}; ... eat(dinner:PLANT):EXCREMENT -- -- they all eat plants the same way -- is return #EXCREMENT(dinner.pick.breakdown); -- this is a cow of perpetual motion; end; is_this_digestable_food(questionable_food:$FOOD):BOOL -- -- all herbivores know that they can *at least* eat plants... -- is typecase questionable_food when PLANT then return #(true); end else return #(false); end; end; ... end; -- abstract class $HERBIVORE ***** With these classes, now we can make specific herbivores. Notice: it makes no sense to make a generic animal ( $ANIMAL{$FOOD} ). It makes no sense to make a generic herbivore ( $HERBIVORE{$PLANTS} ). But now we can easily make specific herbivores. Let's start with a cow. Cow's eat specific kinds of plants. They eat grass (big simplification). So we need to create grass. ***** class GRASS < $PLANTS is breakdown:NUTRIENT is ... end; pick:SAME is ... end; ferment:ALCOHOL is ... end; end; -- end class GRASS ***** Now that we have a kind of food for our cows, we can actually create the cows. ***** class BOVINE < $HERBIVORE{GRASS} is include $HERBIVORE{GRASS}; -- we get the default behaviors for herbivores make_sound -- -- cows go mmooooooo -- is #SPEAKER::output("mmoooooo"); end; come_home:SAME is ... end; -- something that only cows can do... end; -- end class BOVINE ****** And now we'll make another herbivore. Lets make the koalas we mentioned earlier. First we make koala chow. ****** class EUCALYPTUS < $PLANTS is include $PLANTS ... produce_tannin -- something characteristic of eucalyptus I believe -- is ... end; end; -- end class EUCALYPTUS ****** And now we can create the koalas. In addition to herbivore things, koalas can look cuddly. They also shred your arms. ****** class KOALA < $HERBIVORES{EUCALYPTUS} is include $HERBIVORE{EUCALYPTUS}; -- -- and now for koala specific actions -- look_cute is ... end; act_vicious is ... end; climb is ... end; end; -- end class KOALA ******** and now we could do things like... knoll::=#GRASS(10); -- a small knoll tree::=#EUCALYPTUS(20); -- a big tree flossy::=#BOVINE(30); -- flossy's a fat cow ned::=#KOALA(5); -- a tubby koala ned.is_this_digestable_food(tree); -- should be true ned.is_this_digestable_food(knoll); -- should be false flossy.is_this_digestable_food(knoll); -- should be true flossy.is_this_digestable_food(tree); -- should be false flossy.is_this_digestable_food(ned); -- should be false also big_turd::=flossy.eat(knoll); ned.is_this_digestable_food(big_turd); -- should be false or we've got one wierd koala. ****** and from here on we could create deer,antelopes, giraffes, elephants, etc. We could also specialize $ANIMALS to make carnivores and omnivores. Then we could make specific wolves, bears and humans. Does this example do everything people need it to? -- jeff younker jeff@math.uh.edu - These are my opinions. If they resemble those of the - - University of Houston it is possible that I am very sick. - GS/M/O d--- -p+ c++ l m*/--- s/- !g w+ t r x++ ======================================================================= 81 === Date: 24 Mar 1994 00:43:25 GMT From: toshok@cs.uidaho.edu (Chris Toshok) Subject: Proxac users? Hello all, I am in the process of trying to get Proxac (a text editor for program transformation and calculation) to build with the new release of SRC Modula-3. Unfortunately, it is riddled with errors because of changes in the libraries. I was wondering if anyone had taken the time to update the source code for this utility. Also, we are running proxac on a DecStation (DS3100) built with the previous version of SRC Modula-3 and it crashes at various times, dumping a horrendously big core image. Has anyone out there built a stable version of this utility (meaning one that doesn't die intermittently)? If so, mail me please, telling me which platform, which version of gcc, which version of Modula-3, etc., etc.... Thanks in advance, -- ------------------------------+----------------------------------- Chris Toshok | Assistant Systems Administrator a.k.a. Toshok the Functional | Computer Services | University of Idaho | email: toshok@cs.uidaho.edu | Research Assistant phone: (208)885-5562 | Laboratory for Applied Logic smail: P.O. Box 3900 | University of Idaho Moscow, Id 83843-1919 | | Vice Pres. Student Chapter, ACM ------------------------------+----------------------------------- www: http://www.cs.uidaho.edu/lal/students/toshok.dir/toshok.html ------------------------------------------------------------------ ======================================================================= 82 === Date: 24 Mar 1994 09:43:49 GMT From: jsmith@red-branch.MIT.EDU () Subject: Any suggestions for these errors? When attempting to compile Modula3 for Linux (486).. linux kernel pl15 i get the following errors: _m3main.o: Undefined symbol _M3__I_OSErrorPosix referenced from data segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): More undefined symbol _M3__I_OSErrorPos ix refs follow *** error code 1 "/usr/src/Modula3/boot-LINUX/m3build/templates/COMMON.BOOT", line 46: command e xecute failed *** call stack *** "/usr/src/Modula3/boot-LINUX/m3build/templates/COMMON.BOOT", line 46: call to b uilt-in exec "./make.boot", line 363: call to procedure boot_prog _m3main.o: Undefined symbol _M3__I_OSErrorPosix referenced from data segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): Undefined symbol _M3__I_OSErrorPosix re ferenced from text segment ../../libm3/LINUX/libm3.a(FSPosix_m.o): More undefined symbol _M3__I_OSErrorPos ix refs follow *** error code 1 "/usr/src/Modula3/boot-LINUX/m3build/templates/COMMON.BOOT", line 46: command e xecute failed *** call stack *** "/usr/src/Modula3/boot-LINUX/m3build/templates/COMMON.BOOT", line 46: call to b uilt-in exec "./make.boot", line 13: call to procedure boot_prog /usr/lib/crt0.o: Undefined symbol ___libc_init referenced from text segment ../gcc/genattr.c:220 (genattr.o): Undefined symbol __IO_stderr_ referenced from text segment ../gcc/genattr.c:221 (genattr.o): Undefined symbol __IO_stderr_ referenced from text segment ../gcc/genattr.c:222 (genattr.o): Undefined symbol __IO_stderr_ referenced from text segment ../gcc/genattr.c:429 (genattr.o): Undefined symbol __IO_stdout_ referenced from text segment ../gcc/genattr.c:430 (genattr.o): Undefined symbol __IO_stdout_ referenced from text segment ../gcc/rtl.c:454 (rtl.o): Undefined symbol __IO_stderr_ referenced from text se gment ../gcc/rtl.c:457 (rtl.o): Undefined symbol __IO_stderr_ referenced from text se gment ../gcc/rtl.c:458 (rtl.o): Undefined symbol __IO_stderr_ referenced from text se gment ../gcc/rtl.c:463 (rtl.o): Undefined symbol __IO_stderr_ referenced from text se gment ../gcc/rtl.c:463 (rtl.o): Undefined symbol __IO_stderr_ referenced from text se gment ../gcc/rtl.c:463 (rtl.o): Undefined symbol __IO_stderr_ referenced from text se gment ../gcc/rtl.c:0 (rtl.o): More undefined symbol __IO_stderr_ refs follow make: *** [genattr] Error 1 So... anyone have any helpful suggestions on what my problem is? or even better HOW to fix it? Or what stupid thing I've done or not done properly :0 Any help is gladly accepted. Jonathan Smith ======================================================================= 83 === Date: Thu, 24 Mar 1994 13:55:33 GMT From: dagenais@gutrune.vlsi.polymtl.ca (Michel Dagenais) Subject: Re: Any suggestions for these errors? The boot-LINUX compiled right out of the box for me (Slackware 1.1.0, which i beleive is 99pl15). Perhaps your compilation died because you needed more space and when restarted a module was half compiled... My binaries are on ftp.vlsi.polymtl.ca:lude/modula3-3.1/run/linux.tar.Z. -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 84 === Date: Thu, 24 Mar 1994 18:28:41 GMT From: shang@corp.mot.com (David L. Shang) Subject: Re: Who Win the War of Co. vs. Contra? This thread has been directed to comp.object David Shang ======================================================================= 85 === Date: 24 Mar 1994 21:48:47 GMT From: fraser@hare.europarc.xerox.com (Quentin Stafford-Fraser) Subject: Source packages The 3.1 dosumentation talks about 'source packages'. A source package "contains a collection of Modula-3 sources; it builds nothing". I want to create one of these because, for a complex reason, I want to include the same source in two different libraries, but don't want to keep a compiled version in a separate library. Am I right in thinking that I can't install a source package using m3ship? If an m3makefile doesn't contain Library() or Program() then no .M3SHIP file is created, so m3ship won't work. Is there an 'official' solution, or do I install by hand? Thanks, Quentin -- ---------------------------------------------------------------------- Quentin Stafford-Fraser Cambridge University Computer Lab Rank Xerox Cambridge EuroPARC qs101@cl.cam.ac.uk fraser@europarc.xerox.com Tel: +44 223 334411 Tel: +44 223 341521 Fax: +44 223 334679 Fax: +44 223 341510 http://pelican.cl.cam.ac.uk/people/qs101/me.html ---------------------------------------------------------------------- ======================================================================= 86 === Date: Thu, 24 Mar 1994 20:54:54 +0000 From: neil@aldur.demon.co.uk (Neil Wilson) Subject: Re: Who Win the War of Co. vs. Contra? David L. Shang (shang@corp.mot.com) wrote: : It is an open question that whether this approach is able to work on the open : world assumption, the essential assumption for developing large-scale softwar e. : Eifel's system-level type-checking works only on the closed world assumption , : an assumption usually works for small toy programs. It is clear then that you don't fully understand Eiffel type checking, read on. : The Right Solution: Let the Food Type Depend on the Animal Type : The food class in method eat should be specified as a variable to the class o f : Animal: : class Animal : { function eat (food: thisclass'Food); : }; : that is, the food for the animal to eat should be of the type edible to the : animal, instead of any food. : When subclasses are derived, the class of food is automatically redefined in : covariance with animal classes. Since the interface is accurate, local type : checkability is always ensured. In Eiffel this is defined as follows: class ANIMAL feature food: ANIMAL_FOOD eat (something: like food) is do ... end; -- eat ... end -- class ANIMAL class HERBIVORE inherit ANIMAL redefine food end; feature food: PLANT_FOOD end -- class HERBIVORE class CARNIVORE inherit ANIMAL redefine food end feature food: FLESH_FOOD end -- class CARNIVORE Nothing needs changing in the subclass except the type of food. All the feature s are automagically redefined to conform to the new interface. System level type checking prevents the feeding of flesh to a herbivore and you can use herbivore s and carnivores in generalised animal code. Parallel co-variant redefinition is a prominent part of the Eiffel method. Seems straightforward to me, did I miss something? Regards -- Neil Wilson (neil@aldur.demon.co.uk) +44 532 832000 x2242 6 Spring View, Ossett, Yorkshire, WF5 0QB, UK ...Arrive without travelling, see all without looking... ======================================================================= 87 === Date: 25 Mar 1994 13:37:42 GMT From: matthew@cpdapo.tele.nokia.fi (Matthew Faupel) Subject: Re: Who Win the War of Co. vs. Contra? JMY= Jeff M Younker in article JMY: and now we could do things like... JMY: knoll::=#GRASS(10); -- a small knoll JMY: tree::=#EUCALYPTUS(20); -- a big tree JMY: flossy::=#BOVINE(30); -- flossy's a fat cow JMY: ned::=#KOALA(5); -- a tubby koala JMY: ned.is_this_digestable_food(tree); -- should be true JMY: ned.is_this_digestable_food(knoll); -- should be false JMY: flossy.is_this_digestable_food(knoll); -- should be true JMY: flossy.is_this_digestable_food(tree); -- should be false JMY: flossy.is_this_digestable_food(ned); -- should be false also JMY: big_turd::=flossy.eat(knoll); JMY: ned.is_this_digestable_food(big_turd); -- should be false or we've got JMY: one wierd koala. JMY: Does this example do everything people need it to? I think (if I understand the problem correctly) is that while this all works as you've described it, it delays validity-checking to runtime where ideally we want it done at compile time. You could write a program with the above system that attempts to feed ned to flossy and it wouldn't complain until you ran it. The "perfect" system would reject this inhuman treatment at compile time :-) Matthew ======================================================================= 88 === Date: 25 Mar 94 18:05:23 GMT From: partha@paul.rutgers.edu (partha) Subject: Re: Who Win the War of Co. vs. Contra? I have a question about Neil's proposed eiffel solution for David's Animal and Animal_Food situation. For brevity I am not quoting Neil's proposed class hierarchy. Assuming his classes, consider the following declarations: a: ANIMAL; c: CARNIVORE; h: HERBIVORE; af: ANIMAL_FOOD; pf: PLANT_FOOD; ff: FLESH_FOOD; ... and the following cases: 1. h.eat(ff) --> static type error, right? 2. a.eat(af) --> this statement is ok as far as static type is concerned. but this is potentially dangerous because of obvious reasons: what if af is bound to an entity of FLESH_FOOD and a is bound to an entity of HERBIVORE? (both are possible according to subtype rules) thus the need of system-level validity and if the above kind of assignments happen this call will be system-level invalid. How, then we are going to write useful generalised animal code? Aren't calls like this in the animal code going to be rejected as system-level invalid? --partha ======================================================================= 89 === Date: 25 Mar 94 23:59:04 GMT From: rro@CS.ColoState.EDU (Rod Oldehoeft) Subject: Re: Harbison's "Modula-3": errata. And, new edition? The errata sheet is actually available via anonymous ftp from beethoven@cs.colostate.edu:~ftp/pub/M3/...... and not from the site Harbison mentioned. If I had it to do again, I would structure the errata into dated sections, so a user who had done some of the errata would know which to look at for more fixes. Now a poor user has to scan each updated errata list (ordered only by page number) to see if there are new ones in there. Oh well. I'm glad to hear a new book version may happen. Aside from the nine (9) (!!!) pages of errata I've compiled, I'd like to think a minute about how the book could be improved. I'll try to enable brain before engaging keyboard. ======================================================================= 90 === Date: Fri, 25 Mar 94 17:36:11 -0800 From: luca@src.dec.com Subject: Re: Who Win the War of Co. vs. Contra? I'll reformulate the (typecase-free) Sather solution and the (typecased) Cluster solution in type-theoretical terms, as I understand them. Why? Because this way we can discuss more quickly and rigorously. And if I'm wrong, you can tell me exactly where I am wrong. The problem types are A(nimal) and H(erbivore), with P(lant) < F(ood) A = {eat:F->R} H = {eat:P->R} where R is some result type, and {...} are class types, with the subclass rule {l1:T1...ln+m:Tn+m} < {l1:S1...ln:Sn} if T1 in the domain, we need F < P, which is the opposite of what we have. Trying to believe in H < A leads to unsoundness or, to be sound, to Eiffel-style global analysis. There is no other way out of this. The Eiffel solution is outside of the domain of typing, so I won't comment on it any further. ------------ The Sather solution is to change the types into type operators (i.e. parameterized classes): A = Lambda(X<:F) {eat:X->R} H = Lambda(X<:P) {eat:X->R} where G = Lambda(XR} < {eat:X->R} = A(X) Therefore, for any Plant food X, an Herbivore(X) is an Animal(X), while an Herbivore(Food) is not an Animal(Food) because H(F) is not even a legal application. The general technique is to abstract every contravariant occurrence into a parameter. (Doesn't this get a bit heavy?) The variations with GenericAnimal follow from the subtype rules given above. ------------ The Cluster solution is a bit less clear to me and, type-theoretically, quite different. It seems based on existential types (this is the only way I can make sense of the fact that A and H are again real types): A = Exists(X<:F) {eat:X->R} H = Exists(X<:P) {eat:X->R} (It seems that Cluster confuses "Lambda" and "Exists", allowing A(X), so that definitions can be reused. Probably there is no harm to that.) The type rule for Exists is covariant in the bound (unlike Forall), so we obtain H < A, as desired. The problem (if one wants to consider this a problem) is that animals cannot be fed without dynamic type tests. Given an animal, we know from the type specification that there exists a food that it will eat, but we have no way to know which one, statically (this is a consequence of the type rules for operations on existentials, akin to the rules for abstract types.) (N.B. Contrary to what David Shang seems to claim, I believe a run-time test is implicit in the constraints of feed, at the moment that feed is invoked: animal:Animal; food:typeof(animal).FoodType because typeof(animal) may not be knowable statically. So this is just fancy notation for a typecase on the type of food against the hidden food type of animal.) Luca ======================================================================= 91 === Date: Mon, 28 Mar 94 18:44:04 -0600 From: loplas@cs.wisc.edu (Loplas Account) Subject: LOPLAS/TOPLAS Merger As an individual who has been involved with ACM LOPLAS, I thought you'd be interested in the following editorial which will appear in the next issue of TOPLAS: Editorial In 1991 ACM Letters on Programming Languages and Systems (LOPLAS) was created to provide rapid publication of short papers describing research and development efforts in the area of programming languages and related systems. Its mandate was to reduce publication delays to months rather than years, while maintaining the highest editorial standards. Starting with this issue, LOPLAS is merged into TOPLAS. Each issue of TOPLAS will contain two or three LOPLAS-style "short communications," reviewed and published on an expedited basis. Each short communication should be a little "gem" in which the author explains an interesting new idea with a maximum of clarity and a minimum of verbiage. Charles Fischer, Editor-in-Chief of LOPLAS, is now Associate Editor of TOPLAS with responsibility for managing the short communications department. Submissions to the new section, not exceeding 5000 words, should be sent (in Postscript) to toplas@cs.princeton.edu; the cover letter should indicate that they are intended as short communications. Short communications will not be part of the normal TOPLAS backlog; they will be reviewed and published as rapidly as is practicable. We wish to acknowledge the outstanding contributions of Robert Cartwright, Ron Cytron, Saumya Debray, Susan Eggers, Robert Halstead, Jr., Susan Horwitz, Bernard Lang, Richard J. LeBlanc, Jr., Peter Lee, Uwe Pleban, Bill Pugh, Barbara Ryder, and Mayer D. Schwartz, who have served on the LOPLAS Editorial Board. These individuals will continue to handle previous submissions to LOPLAS which will soon appear as TOPLAS short communications. We believe the readership of TOPLAS and LOPLAS will find this merger beneficial. They will continue to receive the comprehensive, in-depth papers TOPLAS is known for, with the added dividend of concise and timely reports on new ideas and innovations. Andrew Appel, Editor-in-Chief, ACM TOPLAS Charles Fischer, Editor-in-Chief, ACM LOPLAS ======================================================================= 92 === Date: Mon, 28 Mar 1994 18:26:41 GMT From: strzelin@bnlku5.phy.bnl.gov (Robert J Strzelinski) Subject: m3 and gcc Will we soon see a m3 front end bundled with gcc the way c++ and obj-c front ends are now? Is the m3 release 3.1 an intentional step in that direction with its dependence on gcc? I think this would be a good way to increase m3's visibility and exposure, as well as increasing its availability on many architectures which are not currently "officially" supported. /////////////////////////////////////////////////////////////////////////////// // Robert J Strzelinski // Brookhaven National Laboratory // // strzelin@bnl.gov // Upton, NY 11973-5000 // /////////////////////////////////////////////////////////////////////////////// ======================================================================= 93 === Date: Tue, 29 Mar 1994 10:52:38 GMT From: aet@munta.cs.mu.OZ.AU (Bert THOMPSON) Subject: ---Looking for QUEST--- This is probably not the right newgroup to post to, but... Does anyone know if there exists an ftpable implementation of Luca Cardelli's QUEST language? (QUEST is the language used in his ``Typeful Programming'' paper.) Many thanks, Bert. ----------------------------------------------------------------------- bert thompson aet@cs.mu.oz.au. ----------------------------------------------------------------------- ======================================================================= 94 === Date: Tue, 29 Mar 94 11:03:54 EST From: wollrath@zaphod.East.Sun.COM (Ann Wollrath - Sun Microsystems Labs BOS ) Subject: bug in date conversion The following program crashes (sparc, sunOS 4.1.3, m3 3.1): MODULE datetest EXPORTS Main; IMPORT Date; IMPORT Time; VAR now: Time.T; today: Date.T; BEGIN now := Time.Now(); today := Date.FromTime(now); END datetest. It gets a segv while converting the time zone in the tm struct to a TEXT (in the procedure M3toC.CopyStoT). Do you have a fix for this? thanks, -- Ann Ann Wollrath email: Ann.Wollrath@east.Sun.COM Sun Microsystems Laboratories, Inc. phone: (508) 442-0565 2 Elizabeth Drive fax: (508) 250-5067 Chelmsford, MA 01824-4195 ======================================================================= 95 === Date: Tue, 29 Mar 1994 14:19:22 GMT From: fn00@gte.com (Farshad Nayeri) Subject: Re: ---Looking for QUEST--- In article <9408820.1997@mulga.cs.mu.OZ.AU> aet@munta.cs.mu.OZ.AU (Bert THOMPSO N) writes: This is probably not the right newgroup to post to, but... Does anyone know if there exists an ftpable implementation of Luca Cardelli's QUEST language? (QUEST is the language used in his ``Typeful Programming'' paper.) You can find an implementation at gatekeeper.dec.com:/pub/DEC/Quest. I believe it was written in Modula-3, but I don't know if it will work with the new libm3 3.x. --farshad -- /// | {@,@} | "He's so unhip that when you (...) Farshad Nayeri | say Dylan, he thinks you're " " nayeri@gte.com | talking about Dylan Thomas." ======================================================================= 96 === Date: 30 Mar 1994 00:01:30 +0100 From: hrnjad@spock.isar.muc.de (Muharem Hrnjadovic) Subject: Q: Is there a public domain modula3 compiler available? Hello, I read an article Niklaus Wirth wrote for the february/1994 issue of the Informatik Spektrum in Germany and from that reading modula3 seems to be a very nice language. That's why I would like to know if there's a public domain modula3 compiler (perhaps with a debugger) available. I am using a i486 PC driven by BSD/386 from BSDI if it matters. thanx, Muharem P.S.: What about Oberon? Is it available for PC's? -- Muharem Hrnjadovic (hrnjad@Spock.isar.muc.de) ======================================================================= 97 === Date: 30 Mar 1994 15:46:24 +0200 From: laverman@cs.rug.nl (Bert Laverman) Subject: Modula-3 FAQ in WWW html format Hi! Some of our PhD students are setting up a WWW site at our CS dept. They pushed me into making my own entry, so I fiddled around. Since my research work has to do with Mpodula-3, I wanted a hypertext link to the FAQ sheet, so I converted it to HTML format. I would love to put some images in it, so if somebody (Sam?) can provide me with the Modula-3 logo as a gif (postscript is fine also), please! Anyways, below my first attempt at a WWW entry for Modula-3. Greetings, Bert ---snip---snip---snip---8<--- Modula-3 Frequently Asked Questions

Modula-3 Frequently Asked Questions

Original file header: 
Newsgroups: comp.lang.modula3,news.answers
Subject: Modula-3 Frequently Asked Questions (FAQ)
Followup-To: comp.lang.modula3
Expires: 15 Mar 1993 00:00:00 GMT
Approved: news-answers-request@MIT.Edu

Archive-name: Modula-3-faq
Last-modified: Feb 04 1993

What is Modula-3 ?

The goal of Modula-3 is to be as simple and safe as it can be while meeting the needs of modern systems programmers. Instead of exploring new features, we studied the features of the Modula family of languages that have proven themselves in practice and tried to simplify them into a harmonious language. We found that most of the successful features were aimed at one of two main goals: greater robustness, and a simpler, more systematic type system.

Modula-3 descends from Mesa, Modula-2, Cedar, and Modula-2+. It also resembles its cousins Object Pascal, Oberon, and Euclid.

Modula-3 retains one of Modula-2's most successful features, the provision for explicit interfaces between modules. It adds objects and classes, exception handling, garbage collection, lightweight processes (or threads), and the isolation of unsafe features.

Is Modula-3 a superset of Modula-2 ?

No; valid Modula-2 programs are not valid Modula-3 programs. However, there is a tool to help convert Modula-2 programs to Modula-3.

Where can I get a description of Modula-3 ?

  • The definition of Modula-3 is contained in:
    System Programming with Modula-3,
    Edited by Greg Nelson,
    Prentice Hall Series in Innovative Technology
    ISBN 0-13-590464-1
    L.C. QA76.66.S87 1991
    also known as SPwM3. Here is the table of contents:
    1. - Introduction
    2. - Language Definition
    3. - Standard Interfaces
    4. - An Introduction to Programming with Threads
    5. - Thread Synchronization: A Formal Specification
    6. - I/O Streams: Abstract Types, Real Programs
    7. - Trestle Window System Tutorial
    8. - How the Language Got its Spots
    Chapters 2 and 3 have been reprinted in Sigplan Notices, Volume 27, Number 8, August 1992, pp 15-42.

  • Sam Harbison has written a more tutorial book about Modula3:
    Modula-3
    Samuel P. Harbison
    Prentice Hall, 1992
    ISBN 0-13-596396-6

Where can I get other information on Modula-3 ?

There is a Usenet newsgroup, comp.lang.modula3. The archives of that group are available via anonymous ftp from gatekeeper.dec.com in pub/DEC/Modula-3/comp.lang.modula3. If you do not have access to Usenet, there is a relay mailing list; send a message to m3-request@src.dec.com to be added to it.

  • There are a couple high-level overview articles available:
    Modula-3
    Sam Harbison, Byte, Vol. 15, No. 12, October 1990, pp 385+.
    Safe Programming with Modula-3
    Sam Harbison, Dr. Dobb's Journal, Vol. 17, No. 10, October 1992, pp 8 8+.
  • A description of the Modula-3 type system is in
    The Modula-3 Type System
    Luca Cardelli, Jim Donahue, Mick Jordan, Bill Kalsow, Greg Nelson, Conference Record of the Sixteenth Annual ACM Symposium on Principles of Programming Languages (POPL), Austin Texas, January 11-13 1989, pp 202-212.
  • The Trestle window system toolkit and higher-level FormsVBT toolkit, available with Modula-3, are documented in the following reports:
    Trestle Reference Manual
    Mark S. Manasse and Greg Nelson, SRC Research Report 68, December 199 1.
    Trestle Tutorial
    Mark S. Manasse and Greg Nelson, SRC Research Report 69, May 1, 1992.
    The FormsVBT Reference Manual
    Marc H. Brown and James R. Meehan, (soon to be SRC Research Report ?? ).
    These reports can be ordered by e-mail; send your request to src-reports@src.dec.com.

Where can I get an implementation ?

There is only one implementation available today. It has been built by SRC and is available via anonymous ftp from gatekeeper.dec.com in pub/DEC/Modula-3/release. Contributed software can be found in pub/DEC/Modula-3/contrib.

The current version, 2.11, implements the language defined in SPwM3. There are versions for the following machines: 

	AIX386	IBM PC running AIX/PS2,
	AP3000	Apollo DN4500 running Domain/OS
	ARM	Acorn R260 running RISC iX 1.21
	DS3100	DECstation 3100 and 5000 running Ultrix 4.0 and 4.2
	HP300	HP 9000/300 running HP-UX 8.0
	HPPA	700/800 running HP-UX 8.0
	IBMR2	IBM R6000 running AIX 3.1, 
	IBMRT	IBM RT running IBM/4.3, 
	NEXT	NeXT running ?
	Okidata	7300 (i860) running AT&T SVR4.0
	SPARC	SPARCstation running SunOS 4.1.x
	SUN3	SUN3 running SunOS
	Sun	386i running SunOS 4.0.1
	UMAX	Encore Multimax running UMAX 4.3 (R4.1.1)
	VAX	VAX running Ultrix 3.1
SRC Modula-3 includes a user manual, compiler, runtime library, some libraries and a few other goodies (see below).

The compiler generates C as an intermediate language and should be fairly easy to port. Except for the very lowest levels of the thread implementation, the entire system is written in Modula-3.

What if I don't have ftp access ?

Unfortunately, we cannot deliver Modula-3 other than by anonymous ftp.

Fortunately, Prime Time Freeware (PTF) includes Modula-3. PTF is a set of two ISO-9660 CDroms filled with 3GB of freeware, issued semi-annually. The latest issue, Volume 1, Number 2, July 1992, contains SRC Modula-3 2.07. PTF is distributed via bookstores and mail. You can reach PTF using: 

	Email:  ptf@cfcl.com
        Fax:    [1] (408) 738 2050
        Voice:  [1] (408) 738 4832
        Mail:   Prime Time Freeware
 	        415-112 N. Mary Ave., Suite 50
		Sunnyvale, CA 94086
	        USA

Can I contribute Modula-3 software ?

Certainly. Send us what you are willing to share, be it programs, libraries or other things. We'll put them in the distribution.

Right now, the pub/DEC/Modula-3/contrib directory contains: 

   m3rpc   an rpc system from Xerox Parc
   M2toM3  a translator from Modula-2 to Modula-3
   m3pc    an implementation of Modula-3 for PCs.


   
-- 
Eric.
---snip---snip---snip---8<--- ------------------------------------------------------------------ Bert Laverman, Dept. of Computing Science, Groningen University Email: laverman@cs.rug.nl Phone: +31-50-633948 Home: bert@rakis.iaf.nl Fax: +31-50-633800 -- ------------------------------------------------------------------ Bert Laverman, Dept. of Computing Science, Groningen University Email: laverman@cs.rug.nl Phone: +31-50-633948 Home: bert@rakis.iaf.nl Fax: +31-50-633800 ======================================================================= 98 === Date: Wed, 30 Mar 1994 14:53:21 GMT From: dagenais@froh.vlsi.polymtl.ca (Michel Dagenais) Subject: M3 Shared libraries on LINUX With more efforts than anticipated i got them working. The most notable problems were, apart from the inherent complexity of building shared libs on LINUX: - Two fixed arrays somewhere in the shared libs tools did overflow on libm3. Made me realize how much time i save in M3 not having to chase bugs in corrupted binaries. - The tools silently ignored symbols with '.' in them. - The shared libs tools intercept global symbols as modules are compiled. Modules recompiled due to new opaque info created duplicate entries. - Libm3 references symbols from the main program. Dummy symbols had to be added as place holders. With shared libraries a sample application went from 75s link time and 2721K to 12s link time and 122K. The application startup time is not much affected. Here is a script that calls "buildShared" to build a number of Modula3 libraries, specifying dependencies on other libraries and setting the fixed address where the library should be loaded at run time in virtual memory. #! /bin/sh # # Where libraries get installed # PKG=/usr/local/soft/modula3-3.1/run/lib/m3/pkg # # Libm3 # ./buildShared -m /home/dagenais/disk/m3 -p libm3 -n libm3 -d $PKG/libm3/LINUX - a "0x80000000" -j "0x4000" -g "0x2000" -x "3" -y "1" -i "/usr/lib/libc /usr/lib /libm" -l "-lm -lc -lgcc -lc" -f "-g" # # Libm3X11R4 # ./buildShared -m /home/dagenais/disk/m3 -p trestle/X11R4 -n libm3X11R4 -d $PKG/ X11R4/LINUX -a "0x80100000" -j "0x4000" -g "0x2000" -x "3" -y "1" -i "$PKG/libm 3/LINUX/libm3 /usr/lib/libXext /usr/lib/libX11 /usr/lib/libc /usr/lib/libm" -l "-L$PKG/libm3/LINUX -lm3 -lXext -lX11 -lm -lc -lgcc -lc" -f "-g" # # Libm3tcp # ./buildShared -m /home/dagenais/disk/m3 -p trestle/tcp -n libm3tcp -d $PKG/tcp/ LINUX -a "0x80200000" -j "0x4000" -g "0x2000" -x "3" -y "1" -i "$PKG/libm3/LINU X/libm3 /usr/lib/libc /usr/lib/libm" -l "-L$PKG/libm3/LINUX -lm3 -lm -lc -lgcc -lc" -f "-g" # # Libm3ui # ./buildShared -m /home/dagenais/disk/m3 -p trestle/ui -n libm3ui -d $PKG/ui/LIN UX -a "0x80300000" -j "0x4000" -g "0x2000" -x "3" -y "1" -i "$PKG/X11R4/LINUX/l ibm3X11R4 $PKG/tcp/LINUX/libm3tcp $PKG/libm3/LINUX/libm3 /usr/lib/libXext /usr/ lib/libX11 /usr/lib/libc /usr/lib/libm" -l "-L$PKG/X11R4/LINUX -lm3X11R4 -L$PKG /tcp/LINUX -lm3tcp -L$PKG/libm3/LINUX -lm3 -lXext -lX11 -lm -lc -lgcc -lc" -f " -g" Now the buildShared script. One can say that building shared libraries on LINUX is non obvious. However, global variables are automatically seperated from the rest of the library, unlike on Suns where this has to be done by hand if you want to have shared libraries that can really be shared in core. #! /bin/sh # # Information that characterize a shared library to build: # M3DIR=/home/dagenais/disk/m3 # where the DEC SRC Modula-3 source tree is PKG=libm3 # package name (libm3, trestle/ui ...) LIBNAME=libm3 # library name (libm3, libui ...) LIBDEST=/usr/local/soft/modula3-3.1/run/lib/m3/pkg/$PKG/LINUX # where the library built should go ADDRESS=0x80000000 # a virtual address assigned to that library JUMPSIZE=0x8000 # size of the jump table with space for growth GOTSIZE=0x4000 # size of the got table with space for growth MAJOR=3 # major release number MINOR=1 # minor release number IMPORTLIBS="/usr/lib/libc /usr/lib/libm" # shared libs on which it depends LDLIBS="-lm -lc -lgcc -lc" # all libs on which it depends M3FLAGS="m3_option(\"-g\")" # # These values can be read from the command line # USAGE="buildShared \ -m m3_directory\ -p package_name\ -n library_name\ -d library_destination\ -a shared_lib_address\ -j jump_table_size\ -g got_table_size\ -x major_version_no\ -y minor_version_no\ -i imported_shared_libraries\ -l libraries_to_link\ -f m3_option\ -F M3OPTIONS\ " while getopts m:p:n:d:a:j:g:x:y:i:l:f:F: name do case $name in m) M3DIR=$OPTARG;; p) PKG=$OPTARG;; n) LIBNAME=$OPTARG;; d) LIBDEST=$OPTARG;; a) ADDRESS=$OPTARG;; j) JUMPSIZE=$OPTARG;; g) GOTSIZE=$OPTARG;; x) MAJOR=$OPTARG;; y) MINOR=$OPTARG;; i) IMPORTLIBS=$OPTARG;; l) LDLIBS=$OPTARG;; f) M3FLAGS="m3_option(\"$OPTARG -Y1@gcc@-B/usr/dll/jump/@ -Y7@/usr/dll/ju mp/as@\")";; F) M3FLAGS="M3OPTIONS=\"$OPTARG -Y1@gcc@-B/usr/dll/jump/@ -Y7@/usr/dll/ju mp/as@\"";; \?) echo $USAGE exit 2;; esac shift; shift; OPTIND=1; done shift `expr $OPTIND - 1` # # The shared lib toolkit needs some info to be passed by the JUMP_DIR # and JUMP_LIB environment variables. # cd $M3DIR/$PKG JUMP_DIR=$M3DIR/$PKG/jump export JUMP_DIR if test \! -d $JUMP_DIR; then mkdir $JUMP_DIR; fi rm jump/* JUMP_LIB=$LIBNAME export JUMP_LIB VERSION=$MAJOR.$MINOR IMAGE=$LIBDEST/$LIBNAME # # To build shared libraries, special options are required in the m3makefile # such that all assembly is performed by /usr/dll/jump/as. # if test \! -f src/m3makefile.orig; then cp -p src/m3makefile src/m3makefile.orig fi sed "s+\( *[lL]ibrary.*\)+$M3FLAGS\\ \1+" src/m3makefile.new mv src/m3makefile.new src/m3makefile # # Everything must be built. The outcome is the file jump/jump.log which lists # all the global symbols in the library # rm LINUX/* m3build # # For libm3 we must insert dummy definitions for the symbols to appear in # the file _m3main.c produced by m3build when a program is compiled. # if test $LIBNAME = libm3; then (cat >jump/tmpm3main.s <jump.log) (cd $JUMP_DIR; /usr/dll/bin/getvars; /usr/dll/bin/getfuncs) (cd $JUMP_DIR; rm jump.log) # # Global symbols imported from other shared libs must be in jump/jump.import # if test -f $JUMP_DIR/jump.import; then rm $JUMP_DIR/jump.import; fi for i in $IMPORTLIBS; do (cd $JUMP_DIR; nm --no-cplus -o $i.sa | grep __GOT_ | \ sed 's/ __GOT__/ _/' >>jump.import) done # # Now that we have the jump.vars, jump.funcs and jump.import files, we # rebuild everything a second time. All the calls to global variables # and functions will be intercepted and treated appropriately. # The original makefile is subsequently restored. # rm LINUX/* m3build cp -p src/m3makefile.orig src/m3makefile if test $LIBNAME = libm3; then (/usr/dll/jump/as -o LINUX/tmpm3main.o jump/tmpm3main.s) fi # # The size of each global variable is measured before building the jump table # (cd $JUMP_DIR; /usr/dll/bin/getsize >jump.vars.new; \ mv jump.vars jump.vars.old; mv jump.vars.new jump.vars) # # The shared library is finally constructed and verified. # cd LINUX /usr/dll/bin/mkimage -f -l $IMAGE -v $VERSION -a $ADDRESS -j $JUMPSIZE \ -g $GOTSIZE -- '*o' $LDLIBS cp -p $LIBNAME.so.$VERSION $LIBDEST /usr/dll/bin/mkstubs -l $IMAGE -v $VERSION -a $ADDRESS -j $JUMPSIZE \ -g $GOTSIZE -- $LIBNAME cp -p $LIBNAME.sa $LIBDEST /usr/dll/bin/verify-shlib -l $IMAGE.so.$VERSION $IMAGE.sa # # A symlink redirects all requests to the major version to this specific # major.minor version # if test -f $IMAGE.so.$MAJOR; then rm $IMAGE.so.$MAJOR; fi ln -s $IMAGE.so.$MAJOR.$MINOR $IMAGE.so.$MAJOR -- --------------------------------------------------------------------- Prof. Michel Dagenais dagenais@vlsi.polymtl.ca Dept of Electrical and Computer Eng. Ecole Polytechnique de Montreal tel: (514) 340-4029 --------------------------------------------------------------------- ======================================================================= 99 === Date: Wed, 30 Mar 1994 16:29:05 -0500 From: Dick Orgass Subject: Building V 3.1 on HPUX I've run into a few non-fatal problems and one fatal problem trying to bring up the 3.1 system on HPUX. I'm writing in the hope that someone will recognize them and be able to point me in the right direction. (1) quake fails to compile with the HP C compiler because of alloca. Solution: Compile with gcc. (2) The functions utimes and getdtablesize are not in the HP libraries. Solutions: rewrite utimes using utime which is provided (and in POSIX 1001.1). Once the compiler is up, this can be directly changed in the Modula-3 source. For getdtablesize, I simply wrote a short C function that returns 2000, the default maximum in AIX -- not a very good solution but should get me started. The fatal problem: (3) For all of the links in the boot system, I get: cs: error 1405: "/bin/ld" terminated abnormally with signal. *** error code 139 Haven't found a hint in the man pages or /usr/include yet. Many thanks for any help you might be willing to provide. Dick PS: When I try the links again with gcc, I get: /usr/local/bin/gcc -o m3 _m3main.o OS_i.o OS_m.o Main_m.o ../../libm3/HPPA/libm3.a ../../linker/HPPA/libm3linker.a -lm -lBSD collect2: ld terminated with signal 11 , core dumped *** error code 1 "/u/decsrc/boot-HPPA/m3build/templates/COMMON.BOOT", line 46: command execute failed Signal 11 is segmentation violation. ======================================================================= 100 === Date: 31 Mar 1994 13:19:58 +0200 From: laverman@cs.rug.nl (Bert Laverman) Subject: Re: Building V 3.1 on HPUX Dick Orgass writes: >(2) The functions utimes and getdtablesize are not in the HP > libraries. Solutions: rewrite utimes using utime which is provided > (and in POSIX 1001.1). Once the compiler is up, this can be > directly changed in the Modula-3 source. For getdtablesize, I > simply wrote a short C function that returns 2000, the default > maximum in AIX -- not a very good solution but should get me started. I am told gcc has a compatibility library which also provides these two. I solved this with two stub routines to call utime() and getrlimit(). >The fatal problem: >(3) For all of the links in the boot system, I get: > cs: error 1405: "/bin/ld" terminated abnormally with signal. > *** error code 139 >Haven't found a hint in the man pages or /usr/include yet. HP first suggested it may be a linker problem, and they sent us a patch to ld. This didn't solve the problem however. Wat did, was to use gcc-as instead of as all around. I suppose the backend that generated the boot archive was configured for use with gas instead of hp-as. When built this way, you will get into the (by now known) problem of the big-endians, which was also plaguing SPARCs. When this is solved by replacing libm3/src/text/UnsafeHash.ms, again the program seems to hang. This was traced to a word-order problem in double constants. All '.double' ops were made into comments, and two separate '.word' entries generated instead. gas however knows about '.double', so I manually edited all occurences back into life. I now get a runtime error. Since the bootstrap compiler was not compiled with '-g' I could not use gdb to find out more. (out gcc by the way complains that '-g' is not implemented. I wonder if we could use it at all...) >PS: When I try the links again with gcc, I get: > /usr/local/bin/gcc -o m3 _m3main.o OS_i.o OS_m.o Main_m.o > ../../libm3/HPPA/libm3.a ../../linker/HPPA/libm3linker.a -lm -lBSD > collect2: ld terminated with signal 11 , core dumped > *** error code 1 > "/u/decsrc/boot-HPPA/m3build/templates/COMMON.BOOT", line 46: > command execute failed >Signal 11 is segmentation violation. This is the same problem. Only when you use gas on all assembly files will you be able to link. -- ------------------------------------------------------------------ Bert Laverman, Dept. of Computing Science, Groningen University Email: laverman@cs.rug.nl Phone: +31-50-633948 Home: bert@rakis.iaf.nl Fax: +31-50-633800 ======================================================================= 101 === Date: Wed, 30 Mar 1994 17:20:06 GMT From: shang@corp.mot.com (David L. Shang) Subject: Re: Who Win the War of Co. vs. Contra? In article <9403260136.AA17864@tsktsk.pa.dec.com> luca@src.dec.com writes: > > I'll reformulate the (typecase-free) Sather solution and the (typecased) > Cluster solution in type-theoretical terms, as I understand them. > Why? Because this way we can discuss more quickly and rigorously. > And if I'm wrong, you can tell me exactly where I am wrong. > ... The followup to this message has been redirected to comp.object. David Shang ======================================================================= 102 === Date: Thu, 31 Mar 1994 20:37:24 GMT From: mikeg@psg.com (Mike Gallo) Subject: Re: Q: Is there a public domain modula3 compiler available? There is a freeware version available from Digital Equipment Corp. on gatekeeper.dec.com. Try the newsgroup Comp.lang.modula3 for details. > > P.S.: What about Oberon? Is it available for PC's? > Yup. Ports of the Oberon system are freely available from neptune.inf.ethz.ch in /Oberon. See the upcoming Oberon FAQ for more details. -- I meant what I said and said what I meant, An elephant is faithful one hundred percent. -- Horton ======================================================================= 103 === Date: Thu, 31 Mar 1994 15:14:10 -0500 From: Dick Orgass Subject: Re: Building V 3.1 on HPUX Thanks to Bert Laverman for the suggestions concerning building version 3.1 for HPUX. Here's a report on what happened so far. We're running gcc version 2.5.8 and -g is not supported by this port either. I compiled all C code using gcc, assembled all source using the assembler with gcc and then linked with gnu ld which is automatically invoked by gcc. This was all vanilla source from the SRC distribution. I changed my getdtablesize stub to match Bart's suggestion. (My C code is below.) After I assembled, compiled, created libraries, linked an installed the system, I got a different result than Bert: My build of m3 goes into a long (probably infinite) loop trying to compile Hello.m3. I once ran it for over 4.5 CPU minutes and it was still gong strong. (Assuming that the time in HPUX's ps is actual CPU time.) At the moment, I'm compiling version 2.11 in the hope that I can use it to get on with some implementation work. I'm retaining the 3.1 system in the hope that it will be possible to get it working sometime soon. Dick --------------------- utimes.c ------------------------- #include struct timeval { long tv_sec; long tv_msec; }; void utimes (file, tvp) char *file; struct timeval tvp[2]; { struct utimbuf tmp; tmp.actime = tvp[0].tv_sec; tmp.modtime = tvp[1].tv_sec; utime(file, &tmp); } ------------------------ getdtablesize.c ------------------ #include int getdtablesize () { struct rlimit rlp; getrlimit(RLIMIT_NOFILE, &rlp); return (rlp.rlim_cur); } -------------------------- end ---------------------------