Do not skip site import on python startup.

Do not skip site import on python startup.

Remove the -S option when invoking the bindings generator to allow implicit 'import site' by the python runtime.

BUG=410787
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=181375

[elecro, 2014-09-04 12:16:39]

pgal.u-szeged@partner.samsung.com changed reviewers:
+ haraken@chromium.org

[haraken, 2014-09-04 12:19:25]

haraken@chromium.org changed reviewers:
+ jl@opera.com

[haraken, 2014-09-04 12:19:25]

I'm not much familiar with python enough to judge this change.

jl@ ?

[Jens Widell, 2014-09-04 12:25:00]

On 2014/09/04 12:19:25, haraken wrote:
> I'm not much familiar with python enough to judge this change.
> 
> jl@ ?

Sounds strange that -S and virtualenv interacts this way. But I'm not familiar
with either, really.

Given the recent request on blink-dev@ to drop this (quite questionable) use of
-S for other reasons, I think we should do that instead. The complexity added
here seems unjustified for an unproven optimization.

[elecro, 2014-09-04 12:39:08]

On 2014/09/04 12:25:00, Jens Widell wrote:
> On 2014/09/04 12:19:25, haraken wrote:
> > I'm not much familiar with python enough to judge this change.
> > 
> > jl@ ?
> 
> Sounds strange that -S and virtualenv interacts this way. But I'm not familiar
> with either, really.
> 
> Given the recent request on blink-dev@ to drop this (quite questionable) use
of
> -S for other reasons, I think we should do that instead. The complexity added
> here seems unjustified for an unproven optimization.

Hmm.. it seems I've missed that discussion on blink-dev@. But I'm all for
removing the -S option :)
I'll update the CL to remove the -S options.

[elecro, 2014-09-04 12:50:04]

Updated to remove the -S options.

[haraken, 2014-09-04 12:52:40]

LGTM

[Jens Widell, 2014-09-04 13:18:07]

On 2014/09/04 12:39:08, elecro wrote:
> On 2014/09/04 12:25:00, Jens Widell wrote:
> > On 2014/09/04 12:19:25, haraken wrote:
> > > I'm not much familiar with python enough to judge this change.
> > > 
> > > jl@ ?
> > 
> > Sounds strange that -S and virtualenv interacts this way. But I'm not
familiar
> > with either, really.
> > 
> > Given the recent request on blink-dev@ to drop this (quite questionable) use
> of
> > -S for other reasons, I think we should do that instead. The complexity
added
> > here seems unjustified for an unproven optimization.
> 
> Hmm.. it seems I've missed that discussion on blink-dev@. But I'm all for
> removing the -S option :)

This is the thread:
https://groups.google.com/a/chromium.org/d/topic/blink-dev/m-9S04U12Xs/discus...

The suggestion met some resistance, but I think it was against the declared
purpose of using system libraries rather than against the actual change to drop
-S.

LGTM too. You might send a mail to blink-dev when this lands, saying the change
has happened.

[elecro, 2014-09-04 14:37:47]

The CQ bit was checked by pgal.u-szeged@partner.samsung.com

[commit-bot: I haz the power, 2014-09-04 14:38:12]

CQ is trying da patch. Follow status at

https://chromium-status.appspot.com/cq/pgal.u-szeged@partner.samsung.com/5389...

[commit-bot: I haz the power, 2014-09-04 16:06:05]

Committed patchset #2 (id:20001) as 181375