Archive Re-namery
Here's something horrible I discovered when I was writing a 25 word review of my recent viewing of The Incredibles and decided to look for my original review: My Movable Type re-installation and the export/import required to move from BerkeleyDB to MySQL re-numbered my entries and broke the URL of every single entry in the jefftidball.com archives.
After a fair bit of research, there's no feasible solution to this problem, which, sadly, will cause a great deal of linkrot out there in webland. (Google, for example, has apparently crawled deep enough to have direct links to individual archives pages. They're all pointing to the wrong entries now. Google-search for "jefftidball.com incredibles," for example. The link now points to a shitfilter entry for an entirely unrelated site. Bah!)
When I get a chance, I'm going to universally change the archive URL structure to use truncated entry names, rather than arbitrary numbers. If I find a spare week of free time, I may even put auto-redirects at the old, numerical URLs. Especially if I can find a way to automate that process. Anyone have any ideas about how to do that, when all I have access to is a backup of the .html archive files themselves? By which I mean, "Paging Jameson Simmons... paging Jameson Simmons...."
Comments
Quite a challenge! Nothing's impossible, though.
Your only record of the old numerical entry IDs is the .html archive files? If that's all there is, it sounds like the best automated solution would be to crawl through those files and associate each ID number with the title and date of its entry. Then, use that association to build redirects to the new pages (which would be fairly trivial with mod_rewrite and a PHP/Perl script accessing the new MySQL database).
I would manually adjust the phase inducers so that when the warp core comes back on line, the Reitman effect will re-link the value chain.
But that's just my opinion.
Make sure you reverse the polarity first.