There it is! The blog uses the technique I mentioned earlier this week to embed the blog and snaplog thumbnails directly into the HTML. Magically, it just works, and a little bit of Javascript takes care of loading images the old way if you’re using Explorer.

I couldn’t tell whether it loads faster, but like I said it makes more sense and it’s more efficient, in my head, to include all the little images into the HTML rather than have dozens additional HTTP requests. I’m sure the idea of forcing image data onto blind, deaf and mute readers would irritate the standards people to no end, but Firefox tells me the modified home page now weighs 66KB (which is surprising, but maybe it’s the gzipped size), so I don’t consider myself guilty of any crime against accessibility.

I’ll have to think about adapting my system to the comment avatars (and, while I’m there, loading favicons when there’s no gravatar), but there’s no hurry.

P.S. Yes, when I reload the page to proof-read something I just posted, it’s definitely faster. No waiting for the new thumbnail to load after my server has thoroughly confirmed that every other one hasn’t been updated since the last time (yes, I configured my Firefox to check the cache’s validity every time a page is displayed — haven’t you?).