Revert 215785 "[telemetry] Add a webdriver backend with support ..."

Revert 215785 "[telemetry] Add a webdriver backend with support ..."

> [telemetry] Add a webdriver backend with support for IE.
> BUG=none
> TEST=tools/perf/run_measurement --browser=internet-explorer octane & kraken
> 
> Review URL: https://chromiumcodereview.appspot.com/20672002

TBR=chrisgao@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=215796

[kochi, 2013-08-06 05:05:29]

Committed patchset #1 manually as r215796.

[chrisgao (Use stgao instead), 2013-08-06 21:12:37]

Hi Takayoshi,

Did r215785 cause some break?
It would be great if you could ping me the links or log of the break.

I can't find a related break in the main or perf waterfall.

Thanks,
chrisgao


On Mon, Aug 5, 2013 at 10:05 PM, <kochi@chromium.org> wrote:

> Committed patchset #1 manually as r215796.
>
>
https://codereview.chromium.**org/22336006/<https://codereview.chromium.org/2...
>

[kochi, 2013-08-06 23:58:48]

Hi,

Sorry not pinging you in advance.
I sent a mail to chromium-dev/chromium-os-dev/telemetry about the breakage,
sorry if it didn't get to you.

I missed pasting the link to the actual log yesterday, but it looked
something like this:
http://build.chromium.org/p/chromium.chromiumos/builders/ChromiumOS%20%28x86%...

Changes to some python files in Chromium tree telemetry files broke the
Chromium OS
package builders (as it depends on those files).

===================
As a sheriff yesterday and today, I had to revert 2 CLs about Telemetry
changes
which broke chromiumos build_packages bot (crrev.com/215573 &
crrev.com/215796).

The Chromium OS Telemetry package and its dependent autotest-chrome and
autotest-tests-ownershipapi packages started to fail when these original
CLs were
committed to the Chromium tree, and buildpackages bots reported failures,
because these packages are partially using Chromium's code directly (IIUC).

Looks like buildpackages builders are not tree closers, so people can work
on
Chrome OS package change after telemetry-related packages are committed to
the tree,
but can we have a trybot for catching this case?  Or is the practice
(e.g. change chromium, update chromium os package then commit them in order
etc.)
described anywhere?


On Wed, Aug 7, 2013 at 6:12 AM, Chris Gao <chrisgao@chromium.org> wrote:

> Hi Takayoshi,
>
> Did r215785 cause some break?
> It would be great if you could ping me the links or log of the break.
>
> I can't find a related break in the main or perf waterfall.
>
> Thanks,
> chrisgao
>
>
> On Mon, Aug 5, 2013 at 10:05 PM, <kochi@chromium.org> wrote:
>
>> Committed patchset #1 manually as r215796.
>>
>>
https://codereview.chromium.**org/22336006/<https://codereview.chromium.org/2...
>>
>
>


-- 
Takayoshi Kochi

[chrisgao_google.com, 2013-08-07 00:19:42]

Hi,

Thank you for the revert and the link.
Will look into how to resolve the failure before recommitting the code.

Thanks again,
chrisgao


On Tue, Aug 6, 2013 at 4:58 PM, Takayoshi Kochi <kochi@chromium.org> wrote:

> Hi,
>
> Sorry not pinging you in advance.
> I sent a mail to chromium-dev/chromium-os-dev/telemetry about the
> breakage, sorry if it didn't get to you.
>
> I missed pasting the link to the actual log yesterday, but it looked
> something like this:
>
>
http://build.chromium.org/p/chromium.chromiumos/builders/ChromiumOS%20%28x86%...
>
> Changes to some python files in Chromium tree telemetry files broke the
> Chromium OS
> package builders (as it depends on those files).
>
> ===================
> As a sheriff yesterday and today, I had to revert 2 CLs about Telemetry
> changes
> which broke chromiumos build_packages bot (crrev.com/215573 &
> crrev.com/215796).
>
> The Chromium OS Telemetry package and its dependent autotest-chrome and
> autotest-tests-ownershipapi packages started to fail when these original
> CLs were
> committed to the Chromium tree, and buildpackages bots reported failures,
> because these packages are partially using Chromium's code directly (IIUC).
>
> Looks like buildpackages builders are not tree closers, so people can work
> on
> Chrome OS package change after telemetry-related packages are committed to
> the tree,
> but can we have a trybot for catching this case?  Or is the practice
> (e.g. change chromium, update chromium os package then commit them in
> order etc.)
> described anywhere?
>
>
> On Wed, Aug 7, 2013 at 6:12 AM, Chris Gao <chrisgao@chromium.org> wrote:
>
>> Hi Takayoshi,
>>
>> Did r215785 cause some break?
>> It would be great if you could ping me the links or log of the break.
>>
>> I can't find a related break in the main or perf waterfall.
>>
>> Thanks,
>> chrisgao
>>
>>
>> On Mon, Aug 5, 2013 at 10:05 PM, <kochi@chromium.org> wrote:
>>
>>> Committed patchset #1 manually as r215796.
>>>
>>>
https://codereview.chromium.**org/22336006/<https://codereview.chromium.org/2...
>>>
>>
>>
>
>
> --
> Takayoshi Kochi
>