managing chromes use section of docs

managing chromes use section of docs

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/193b1346552a49460488274cb91799c8b0c78181

[hcm, 2015-02-18 02:58:00]

hcm@google.com changed reviewers:
+ bungeman@google.com, reed@google.com, robertphillips@google.com

[bungeman-skia, 2015-02-18 15:58:36]

https://codereview.chromium.org/934893003/diff/100001/site/dev/chrome/blink.md
File site/dev/chrome/blink.md (right):

https://codereview.chromium.org/934893003/diff/100001/site/dev/chrome/blink.m...
site/dev/chrome/blink.md:31: You usually only need to fire off the layout bots.
This is good as far as it goes, I suppose, but it's too bad that it doesn't go
on to state how to actually remove the suppression. Just suppressing means you
aren't done yet.

The benefit to this setup is that Skia doesn't need code suppressions. The
downside now is that you cannot use the auto rebaseline bot (as [
NeedsRebaseline ] in this file will not be rebaselined). As a result, the
baselines need to be manually updated in Blink, and then these suppressions
removed from Chromium. This was useful in the days before the auto rebaseline
bot when you'd have to update the baselines manually anyway, but these days this
may actually be more (and easier to get wrong) work.

https://codereview.chromium.org/934893003/diff/100001/site/dev/chrome/blink.m...
site/dev/chrome/blink.md:84: ### Rebaseline
This is more or less right (I think this section was originally written by
fmalita), but it is a bit hand-wavy (Rebaseline like any Blink rebaseline?) and
a bit out of date now.

I think the more modern version of this is to:

Create a CL removing the code suppression from Blink's
public/blink_skia_config.gyp while simultaneously adding [ NeedsRebaseline ]
lines to blink's LayoutTests/TestExpectations file. Then the auto rebaseline bot
will take care of the work of actually checking in the new images. This is
generally acceptable for up to 600 or so rebaselined images. Above that you
might still use [ NeedsRebaseline ], but it's best to coordinate with the
gardener/sheriff. This should go through the CQ cleanly.

Basically, this paragraph should probably replace most of 2 and the first
sentence of 3. I'd keep the rest of the warning in 3 (starting with 'Be
careful...').

[hcm, 2015-02-18 20:38:26]

On 2015/02/18 15:58:36, bungeman1 wrote:
> https://codereview.chromium.org/934893003/diff/100001/site/dev/chrome/blink.md
> File site/dev/chrome/blink.md (right):
> 
>
https://codereview.chromium.org/934893003/diff/100001/site/dev/chrome/blink.m...
> site/dev/chrome/blink.md:31: You usually only need to fire off the layout
bots.
> This is good as far as it goes, I suppose, but it's too bad that it doesn't go
> on to state how to actually remove the suppression. Just suppressing means you
> aren't done yet.
> 
> The benefit to this setup is that Skia doesn't need code suppressions. The
> downside now is that you cannot use the auto rebaseline bot (as [
> NeedsRebaseline ] in this file will not be rebaselined). As a result, the
> baselines need to be manually updated in Blink, and then these suppressions
> removed from Chromium. This was useful in the days before the auto rebaseline
> bot when you'd have to update the baselines manually anyway, but these days
this
> may actually be more (and easier to get wrong) work.
> 
>
https://codereview.chromium.org/934893003/diff/100001/site/dev/chrome/blink.m...
> site/dev/chrome/blink.md:84: ### Rebaseline
> This is more or less right (I think this section was originally written by
> fmalita), but it is a bit hand-wavy (Rebaseline like any Blink rebaseline?)
and
> a bit out of date now.
> 
> I think the more modern version of this is to:
> 
> Create a CL removing the code suppression from Blink's
> public/blink_skia_config.gyp while simultaneously adding [ NeedsRebaseline ]
> lines to blink's LayoutTests/TestExpectations file. Then the auto rebaseline
bot
> will take care of the work of actually checking in the new images. This is
> generally acceptable for up to 600 or so rebaselined images. Above that you
> might still use [ NeedsRebaseline ], but it's best to coordinate with the
> gardener/sheriff. This should go through the CQ cleanly.
> 
> Basically, this paragraph should probably replace most of 2 and the first
> sentence of 3. I'd keep the rest of the warning in 3 (starting with 'Be
> careful...').

Alright PTAL at proposed new steps for first section and integrated updates to
second section.  

Per Mike, if other specific fixes I can quickly address with this check-in,
please comment.  If there are general differing thoughts on what this process
should be, per Mike's suggestion I'll start a shared doc for you guys to pitch
in your thoughts...and go ahead and get these files over to skia.org in an OK
state.

[bungeman-skia, 2015-02-18 20:42:25]

lgtm with nit.

https://codereview.chromium.org/934893003/diff/140001/site/dev/chrome/blink.md
File site/dev/chrome/blink.md (right):

https://codereview.chromium.org/934893003/diff/140001/site/dev/chrome/blink.m...
site/dev/chrome/blink.md:80: lines to blink's LayoutTests/TestExpectations
[file](http://src.chromium.org/viewvc/blink/trunk/LayoutTests/TestExpectations).
nit: s/blink/Blink/

[hcm, 2015-02-18 20:49:52]

New patchsets have been uploaded after l-g-t-m from bungeman@google.com

[hcm, 2015-02-18 20:50:48]

On 2015/02/18 20:42:25, bungeman1 wrote:
> lgtm with nit.
> 
> https://codereview.chromium.org/934893003/diff/140001/site/dev/chrome/blink.md
> File site/dev/chrome/blink.md (right):
> 
>
https://codereview.chromium.org/934893003/diff/140001/site/dev/chrome/blink.m...
> site/dev/chrome/blink.md:80: lines to blink's LayoutTests/TestExpectations
>
[file](http://src.chromium.org/viewvc/blink/trunk/LayoutTests/TestExpectations).
> nit: s/blink/Blink/

thanks, fixed.

[hcm, 2015-02-18 20:50:51]

The CQ bit was checked by hcm@google.com

[commit-bot: I haz the power, 2015-02-18 20:51:04]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/934893003/160001

[commit-bot: I haz the power, 2015-02-18 20:56:44]

Committed patchset #9 (id:160001) as
https://skia.googlesource.com/skia/+/193b1346552a49460488274cb91799c8b0c78181