View source for L20n/Weekly Update/2009-07-06

__NOTOC__
<small>[[L20n/Weekly_Update/2009-06-29|&laquo; previous meeting]] | [[L20n:Weekly_Meetings|Weekly Meetings]] | [[L20n/Weekly_Update/2009-07-13|next week &raquo;]]</small>

Meeting Details

* 9:00 AM Mountain View Time (UTC -7)
* 650-903-0800 x92 Conf# 206 (US/INTL)
* irc.mozilla.org #l10n-drivers for backchannel
* ''(notetaker: Seth)''

== Attendees ==
Seth, Gandalf, Axel, Pike

== Notes ==
*Q3 L20n goal summary
**Evaluating JSON support
**Code optimization
**Trigger discussion with tool authors on how to create a UI paradigm to expose richness of l20n

*Code optimization
**Initial code optimizations did not bring us as much Pike thought
**Inlining to remove function call overhead wasn't as fast as we thought
**Tracing turned on was slower than when it was turned off
**Gandalf's ideas about performance:
***Tweaking Firebug
***Give to Jquery team to take a look at it
***Use Venkman
***Set up meeting with JS Team
***[http://bitbucket.org/jhiatt/l20n-test/ l20n-test]...
***[https://files.getdropbox.com/u/355217/profile.html A tweak of the test by Gandalf with Venkman profiler turned on].

*Evaluating JSON support
**Need to get test case more realistic to resolve a few variables before we start a thorough evaluation of JSON
**Using JS expressions
***Evaluate the code as JS and not trigger the l20n parser
***Could be tricky with subtleties that are useful to have in practice

== To-dos ==
*Get better ideas on current profiling results by evaluating Firebug and also comparing to Venkmann
**  [https://files.getdropbox.com/u/355217/profile.html A tweak of the test by Gandalf with Venkman profiler turned on to start the comparison].
*** Very hard to compare these results against the Firebug profile.  
***  Gandalf will try l20n.js 
*Get further feedback on code from JS and Jquery team
*Investigate Hg issue for jhiatt, consider WONTFIX-ing bug