Bah. Another reason why I distrust all this stuff-it-into-a-database business. On one of my last posts, I think I may have missed a closing quotation mark, or maybe a closing angle-bracket. Which will understandably make the rest of the post unreadable. Unfortunately, because I am using the built-in text-editor for Wordpress, the editor decided to url-escape everything after the mistake. While I was able to extract meaningful text from some of it, some of it simply fell into /dev/null, never to be seen again.

Maybe it wouldn’t have munged it like that if I had used an external editor—although I don’t know. I wish there was a way to have it not munge up my posts, even if I did make a mistake.

And since I’m in no mood to switch to another blog engine right now, and I don’t have the time or wherewithal to complete my own XSLT based blog engine written in perl, I think I’m going to try mucking around with XML-RPC, the Metablog API, and Perl or maybe even Ruby and see if I can’t just use good ol’ emacs to write my posts in XML, including all the metadata and what-not, and just have a script extract the metadata, transform my custom tags to XHTML with XSLT, and send everything to Wordpress. I know it sounds ridiculously complicated, but the way I rationalize it is that if I ever get my file-system based blog engine running, I can just drop it in. (Right.)

The other thing is that I can simply reverse the process—use XML-RPC and the Metablog API to extract all my Wordpress posts, then reconstruct XML post entries by combining the metadata and post text. Voilà, an instant migration tool to my blog engine.

Man, I’ve gotta stop blogging about blogging.