Benjy at Linkstew did me a huuuuuge favor, translating an HTML page full of weblog entries into SQL for me, effectively doing what my web server does in reverse.

65 posts from mid-June to mid-August have now been recovered and re-incorporated into the site! Sure, you probably don't go back and read what I wrote back in July just for the fun of it, but as my categorization by topic nears completion, I know these posts wil get read more often, especially as their content starts to get indexed. More to the point, it's a weight of my mind that eight weeks worth of blogwork have now been 'un-lost'.

For the 99% of you who weren't reading back then, the story was I wanted to back up my SQL database, because I'm a firm believer in backing up stuff before you lose it. The problem was threefold: that I wasn't as handy with tar as I should have been, that I didn't specify unique, date-identifying names to my tar archives, and I didn't actually copy my backups off of the host computer. Basically, I mismatched the source and destination arguments in the tar command so instead of backing up the database to an archive, I overwrote the database with the previously saved archive, wiping out all the blog entries since that archive I'd made two months before.

After ripping of hair, checking of archives, and frantic calls to friends, I managed to find all the posts in html form within the thousands of files in my own Netscape cache from viewing my site (gooood netscape cache. never will I doubt thee again).

Formatted nicely into HTML I saw it as a huge chore to input it back into the database, and shelved it. Benjy, being such a whiz with emacs, and feeling a little freer now that his midterms are over, helped me out with it, and now we're back up and running!

Cool!