It's a little unfair, and yet I find I have to write up my feelings and conclusions from a little local discovery yesterday.
We're talking about the exciting techie information world of "retrospective conversion," the two-word term of art coined by librarians a couple decades ago when they faced the daunting task of getting all that paper-based catalog information online.
Yesterday's little discovery wasn't of a situation that daunting, but it also was a situation you'd think (in glorious hindsight, of course) wasn't as inescapable as the librarian's plight. For decades (centuries?), librarians had to use paper. But what excuse was there, in this other case, for adopting proprietary software solutions in the recent (dim?) history of 1998 web application development?
The nearby (unnamed) university situation was that since 1998 (data collection more in earnest since 2001) one school had devised an online course website build tool, using ColdFusion (writing to Oracle).
By spring of 2003 someone concluded that it was best to abandon that proprietary format approach, to migrate course software to the wider university-based Java offering, and to spend some money "retrospectively converting" those five or so years of course instance data sets into a nice, neutral, TEI-Lite XML vocabulary. This remedial effort restores that data to its original state of "free" potential once more, by means of rendering to a standards-based, open source, text-based, openly processable information format. And so this modest investment in Perl mongering salvages the larger investment in corporate memory, and effectively "banks" the content, against some yet to be foreseen purpose/application.
-----
So, my feelings and conclusions?
Well, I have to say it's really kind of interesting (amusing?) (perhaps a bit of schadenfreude there...?) to find this example of the need for "retrospective conversion" as applied to such a short stint of history (fewer than 5 years!), as opposed to that enormous legacy catalog data libraries faced.
Maybe what this discovery really ought to be is a bit of a humbling message to overzealous "solutions providers" (like yours truly?) evangelizing certain approaches... On the other hand, I also think it pretty well validates my avoidance (since ca. 1999) of anything short of Java, XML, Unicode, and other real standards-based, open source, text-based solutions. No more Microsoft
VB, Lotus Notes, Macromedia Flash/ColdFusion and the like for me.
Anyway, a bit of vindication, for one (me) who's been there ("done that"), but has since been pretty well able to renounce those kinds of dev. environments.
Wm.