<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii"> <META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7036.0"> <TITLE>RE: [SCC_Active_Members] Digital Data Format problems</TITLE> </HEAD> <BODY>  Randall's general point is well taken, but I cannot endorse his specifics as close to adequate.  (Randall does not assert that they are!) Notice that in the June meeting I did not represent what's been done so far with GSDL to be preservation-worthy.  GSDL is an adequate content management base on top of which preservation methodology is feasible.  I have doing that in mind starting in 2008, following the only technical scheme that I am confident is workable in the face of all known risks (including the disappearance of institutions such as CHM). To achieve it once we have a conventional content management infrastructure reliably in place, there are at least 3 challenges. (1) Some software needs to be written for durable coding and trustworthy packaging of content that is reliably bound to adequate provenance metadata.  This is not much software from the point of view of any moderate size development shop, but I believe it to be more than I can write, test, and deploy all on my own with suitable quality.  However, I believe that SPG could accomplish it if two or three volunteers would join me in such a subproject. (2) SPG and CHM would have to work out the details of metadata schema with a target of eventual endorsement as "sufficient and official" for the purposes and mission of CHM.  This is not a matter of broad outlines (the METS standard provides that), but instead requires specification of very detailed design (and validation programs to test input), perhaps for several levels of certification as adequate.  (For instance, CHM might require that an archivally-trained employee provide or endorse the final packaging for a formal accession into a CHM collect, whereas packaging under less stringent rules would almost certainly be the most we could expect from the many volunteers that it will take to create a broadly representative software collection.) (3) The biggest challenge is that proper representation of digital objects for long-term value involves work that people find tedious.  For instance, creation of adequate metadata for preservation is not today an institutional practice in any archive that I know of.  (Instead, people fritter away time and resources with local variants of the Dublin Core metadata scheme.)  So the challenge for SPG is to figure out how significant numbers of contributors can be persuaded to do this work with sufficient quality. To some extent the task can be eased by software that makes it easy to package closely related information with reliable links and with adequate metadata (ease of use extensions to the work alluded to as (1) above).  However, I think it will take a great deal of persuasion even after we have such software in hand. References: (1) My slides from the February 2007 SPG meeting (<A HREF="http://www.hgladney.net/PDI.pdf">http://www.hgladney.net/PDI.pdf</A>; also <A HREF="http://www.hgladney.net/ARCHpres.htm">http://www.hgladney.net/ARCHpres.htm</A>) (2) My slides from the June 2007 meeting (<A HREF="http://www.hgladney.net/CHMpres.pdf">http://www.hgladney.net/CHMpres.pdf</A>) (3) H.M. Gladney, Preserving Digital Information, Springer Verlag, 2007.  ISBN 978-3-540-37886-0  (<A HREF="http://home.pacbell.net/hgladney/PDIf.pdf">http://home.pacbell.net/hgladney/PDIf.pdf</A>) My Message: (1) Tailoring GSDL to be an adequate CM base for SPG collecting on an expanded scale will take us at least until year-end 2007.  There is little point in tackling long-term preservation until we have a reliable CM infrastructure that people are comfortable with.  The slides (2) sketch some specifics of what's needed. (2) I'm seeking volunteers to work with me on creating the software tools and institutional methodology for long-term preservation of software.  This can be an amusing task. Cheerio, Henry -----Original Message----- From: scc_active-bounces@mail.computerhistory.org [<A HREF="mailto:scc_active-bounces@mail.computerhistory.org">mailto:scc_active-bounces@mail.computerhistory.org</A>] On Behalf Of Randall Neff Sent: Thursday, July 05, 2007 9:53 PM To: SCC_active@computerhistory.org Subject: [SCC_Active_Members] Digital Data Format problems At the last meeting, Henry was demonstrating saving PowerPoint files into his demo content managment system.  I was concerned about saving files in secret, proprietorial formats.  Also, the metadata did not record WHICH version PowerPoint format. The British National Archives has the same sort of problems, and Microsoft is going to save the day through virtualization technology. Of course, you need a copy of every Microsoft OS and application version to make it work. The article doesn't mention different floppy disk sizes, and ignores other operating systems and non Microsoft applications.  Warning of data ticking time bomb <A HREF="http://news.bbc.co.uk/1/hi/technology/6265976.stm">http://news.bbc.co.uk/1/hi/technology/6265976.stm</A> My suggestion for SPG content is we store it in multiple forms, while we have access to the proprietary software application. So we save the original proprietary binary format, a image sequence format such as .png, and a text format like .txt or .rtf (for searching). Another problem for future display is that some document formats assume specific fonts and font metrics are available in the computer excuting the application. Randall. </BODY> </HTML>