[loggerhead/MERGE] Switch Loggerhead from KID to ZPT template engine

John Arbash Meinel john at arbash-meinel.com
Tue Jun 10 16:16:21 BST 2008

Hash: SHA1

Martin Albisetti wrote:


|> I can see *supporting* zpt as a tradeoff.  At this point I'm not certain
|> we should be ditching kid at the same time.  At least, not in Loggerhead
|> trunk.
| Code is a mess as-is, not sure that allowing to choose templating
| engines will help in that area.
| I'll look into a few more templating engines, and see if I can come up
| with something better.

Can you keep documentation on which ones you try and how they perform? I think
it would be nice to know, and might provide some insight.

Like I just did "python template fast" and found Mako, which claims this benchmark:

Insanely Fast. An included bench suite, adapted from a suite included with
Genshi, has these results for a simple three-sectioned layout:
Mako:	1.10 ms	    	Myghty:	4.52 ms
Cheetah:1.10 ms		Genshi:	11.46 ms
Django:	2.74 ms		Kid:	14.54 ms
Since a speed test is always a flashpoint for controversy, and you can modify
the bench to show different variances, the point here is not that "Mako is
faster"; its not meant as a competitive point. The point is, "Mako is as fast as
any of the other currently popular approaches".

I don't see ZPT there, but it does show that Genshi is in the "Kid" tier, not
the Cheetah/Django/Myghty tier.

Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org


More information about the bazaar mailing list