[Chandler-dev] Re: wx tarball, SWIG and functional tests update

[Philippe Bossut]

Hi,

Heikki Toivonen wrote:
> Robin Dunn wrote:
>   
>> So it looks like SWIG isn't the culprit, unless I just got lucky and the
>> totally random failures didn't fail for me three times in a row.
>>     
>
> Thanks Robin, the ability to use the newer swig would actually buy us
> some benefits. I understand there have been some performance
> improvements, and we already have the other swig-dependent (well, only
> m2crypto besides wx) tarballs generated using swig 1.3.29.
>   
Glad we established that. Now we have to create the wx branch (from rev 
9544 which corresponds to wx tarball 36 as mentioned by Robin) and start 
the process described earlier in that thread. 2 questions:
- Heikki/Bear: are you ready to create that branch and flip it (so it 
becomes the de facto trunk)?
- John: how far are you from starting the increment/test play?

Also, David, once we get the branch created, it is imperative that you 
stop incremental updates.
> Alec did report some issues earlier that he attributed to swig upgrade,
> but he may have worked around those in Python.
>   
He did for what I know though Robin should be interested by his observation.

Cheers,
- Philippe