[SCC_Active_Members] Agenda for 8/25/04 SCC

[Paul McJones]

Gio,

Yes, that material is of interest to me, for several reasons:

1. As part of my Fortran project, it's interesting to look at what must have
been one of the first student-oriented compile-and-go systems (WATFOR being
another). I would be happy to scan the documents; I expect Al Kossow would
be willing to archive the scans in bitsavers.org.

2. As Berkeley grad (BS Enginering Mathematics '71), I'm interested in
documenting more of Berkeley's computer history.  For example, I recently
spoke with my friend Vance Vaughan about one of his first jobs, programming
Berkeley's IBM 701.

3. While undergrads at Berkeley, Charles Simonyi and I implemented a Snobol4
system for the CDC 6400 that was widely used for instructional purposes,
because our version used substantially less memory than the standard Bell
Labs implementation.


Paul

> -----Original Message-----
> From: Gio Wiederhold [mailto:gio at DB.Stanford.EDU] 
> Sent: Wednesday, August 25, 2004 11:48 AM
> To: Paul McJones
> Cc: 'SCC active'; 'Bernard L. PEUTO'
> Subject: RE: [SCC_Active_Members] Agenda for 8/25/04 SCC
> 
> Software folks,
>     I found some documentation for the Student Fortran 
> compiler written and used at UC Berkeley from 1962 to about 
> 1967, when UC's Computer Cemter switched from IBM equipment 
> (7090) to a CDC 6400.
> That compiler was structually different from the optimizing 
> multi-pass compilers available then, which were too slow to 
> allow teaching of Fortran to larger classes.
>     By limiting program size, the compiler could stay in 
> memory and pass sequences of programs through without having 
> to be reloaded.
>     Two unusual system adaptations at UC Berkeley's Computer 
> Center supported the scheme: A double memory (64K), although 
> the memories were not directly addressable from each other, 
> because of the 15-bit address limitations in the IBM 
> architecture, and a front-end 1401 with a `large' 1405 disk 
> which allowed collecting the student programs into batches 
> for processing.  
> 
>      Is that material of interest to you? The software was 
> also submitted to public use as part of the Share library, 
> under the name $tudent, but I don't know what happened to that.
> 
> Gio
> 
> 
> /Gio Wiederhold/
> http://www-db.stanford.edu/people/gio.html
>