Criminy. If I’d known GreyStork’s interview and Brett’s reprinting of my unscientific benchmark numbers would have created as much interest in Poseidon as they have, I’d have been a lot more careful about pointing out the unfairness of comparing two sets of benchmark results gathered under different conditions.
The LightPress folks have apparently gotten wind of the comparison, which kinda makes me feel bad about being so flippant with my remarks about LightPress’s performance. Not that I feel bad about dissing WordPress. In terms of performance, WordPress is absolute crap. But LightPress is a huge improvement, and I really do admire those guys for the effort they’ve put forth. I also admire them for not taking my remarks personally (at least, it doesn’t seem like they did).
I’ll be the first to say it, because (surprisingly) nobody else has: LightPress has already had several releases and is perfectly usable. Poseidon 1.0 still isn’t anywhere near being usable. Whether or not Poseidon is technically faster, it’s still serving 0 useful requests per second because it isn’t finished.
Now, while we’re on the subject: a lot of people have been grabbing the Poseidon source from CVS. That’s awesome. I’m glad people are so excited. The problem is that a lot of these people are getting in way over their heads. The reason Poseidon 1.0 is only in CVS right now is because it’s not finished. If you’re checking out the CVS code because you can’t wait to start using Poseidon, you’re in for a disappointment. If it were usable, I’d be using it on wonko.com.
If you’re checking out the CVS code because you’re curious or because you’d like to learn something or even, God forbid, because you’d like to contribute to Poseidon’s development, more power to you. Otherwise, you could probably save yourself some time and effort by waiting for an official release. And you can definitely save me some time and effort by not emailing me asking me how to use CVS. Especially when there are instructions right here.