Update the list of bots in rebaseline.py.

Update the list of bots in rebaseline.py.

This script was not updated for the new Ubuntu bots. Also, it was
attempting to rebaseline some bots for which we don't keep results
(NVPR, ASAN).

R=borenet@google.com, borenet

Committed: https://code.google.com/p/skia/source/detail?r=14472

[Stephen White, 2014-04-30 15:21:10]

borenet@: PTAL. Thanks!

[borenet, 2014-04-30 15:40:42]

LGTM, but:

1. Can we (finally) get rid of this hard-coded list?
2. Is rebaseline server not working, or not in a good-enough state to replace
this script?

[Stephen White, 2014-04-30 16:11:20]

On 2014/04/30 15:40:42, borenet wrote:
> LGTM, but:
> 
> 1. Can we (finally) get rid of this hard-coded list?

Sure. I think this dates from when the bots were named differently on autogen
(now GS)
and in the expectations. Logged as
https://code.google.com/p/skia/issues/detail?id=2500.
Looks we could just pull it from a dirlist of expectations/gm/Test-*;
would that work?

> 2. Is rebaseline server not working, or not in a good-enough state to replace
> this script?

Partly I'm just a Luddite, but also note that this script is able to work when
the
rebaseline server is down (this happened recently).

[Stephen White, 2014-04-30 16:11:54]

Committed patchset #1 manually as r14472 (presubmit successful).

[epoger, 2014-04-30 16:55:10]

On 2014/04/30 16:11:20, Stephen White wrote:
> On 2014/04/30 15:40:42, borenet wrote:
> > LGTM, but:
> > 
> > 1. Can we (finally) get rid of this hard-coded list?
> 
> Sure. I think this dates from when the bots were named differently on autogen
> (now GS)
> and in the expectations. Logged as
> https://code.google.com/p/skia/issues/detail?id=2500.
> Looks we could just pull it from a dirlist of expectations/gm/Test-*;
> would that work?
> 
> > 2. Is rebaseline server not working, or not in a good-enough state to
replace
> > this script?
> 
> Partly I'm just a Luddite, but also note that this script is able to work when
> the
> rebaseline server is down (this happened recently).

Yesterday's http://skbug.com/2474 ('rebaseline server is down') was caused by
the part of rebaseline_server that automatically gets the list of platforms to
rebaseline... so if rebaseline.py didn't have the hard-coded list, it would have
failed too.

rebaseline.py is unsupported.