In the interests of there being less business all up in here, I have created a crazy blog system.
Related in spirit to: e, Hobix, Blosxom, Yurt.
The HTML is the framework. The cache and the store are one. Lines of code, 150ish, maybe less, all procedural. Managed via the filesystem. Also there is some kind of read-only API going on, for free.
Ok, so now a ridiculous benchmark. On the server, a dynamic request from Typo, complete with MySQL climbing painfully out of swap:
$ time curl --head localhost:4001 real 0m11.270s
$ time curl --head localhost:4001 real 0m2.825s
Now with page caching:
$ time curl --head localhost:4001 real 0m0.015s
But, how about a dynamic request from the all-new Bax?
$ time curl --head localhost:4040 real 0m0.017s
Yep. Now to get that feed to validate.
So the crazy html-filled RSS feed is on purpose? Or can I submit a bug report somewhere? :)
What’s your RSS client? Google Reader managed to cache a broken version for a short time but it seems ok now.
Oh, and, you’re famous.
Twitterfying your comments, cool. Perhaps email notifications are old fashioned. :)
Bloglines’ copy of your feed looks funky too. All your angle brackets are replaced with double-escaped entities. But maybe they also have an cached copy.
Yeah, I’m not sure how to force a refresh for the web clients. I’ll have to wait until I post a new article, and see what happens.