Thursday, November 15. 2007More template benchmarks
As you know, there are lies, damned lies and
So here are some more template benchmarks: http://template-bench.iliakantor.ru/, prepared by Ilia Kantor. Curiously enough, Smarty is on the 5th place, right after the stuff written in C, so think twice before saying it's slow. Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
Yay to Smarty!
 I still believe smarty would win from some cleaning-up and Smarty3 can be faster than Smarty2
Yeah, but first it was on place 6 when i entered the site.
On place 2 there is PHP itself so what kind of test is it ?
Hey Antony,
One of the new developers at my job is advocating ETS (http://ets.sourceforge.net/). I don't have any experience with it, but I was wondering how it stacks up against the competition, in terms of speed. Is there any way you can add this system to your benchmarks?
Sure, try mailing Ilia (see his email on the mentioned page), I guess he has no problems adding another template engine to the comparison.
Haha ... Whoops. I missed the part about Ilia being the person who actually arranged these benchmarks. What I get for not reading thoroughly.
The benchmark is pluggable, Howto is available on main page (besides many existing modules).
Feel free to implement your own engine benchmark and publish it by mailing iliakan (at server) gmail.com.
It's still 433% slower than pure PHP templates. So yes, while Smarty isn't slow compared to other crappy templating systems, it's still slow.
the truth is somewhere between
 most of really highloaded projects doesn't use smarty. guess why
>most of really highloaded projects doesn't use smarty
Most of the people don't speak Italian, does this mean there is something wrong with Italian? >guess why Because of rumors like this? Use an opcode cache and Smarty becomes no different from PHP+HTML mess, but it's much more user friendly when it comes to maintenance.
>>Use an opcode cache and Smarty becomes no different from PHP+HTML mess
In theory - yes But we are engineers. I think that you "believe" in what you're saying but never tried to measure. Measurement is the only one correct way of talking to the nature, so let's use scientific method instead of talking about Italian. No different? Prove your words! You may take any Smarty code you like which makes something closer to real life than Ilya's benchmarks (if you don't believe to my lebowski-bench results, http://alexeyrybak.com/blitz/lebowski-bench-big.gif). It must produce view-stage code only, no database calls, no sleeps, external requests or whatever. And I'll do a PHP-analog which is at least 30% faster with ab/http_load. Deal?
Antony, thanks for the link, now I'm gonna bench {{macro}}(http://efiquest.org/?p=14)
|
QuicksearchArchivesCategoriesSyndicate This Blog |
