Revert of roll_dep: Avoid large commit messages by providing the log as a link

Revert of roll_dep: Avoid large commit messages by providing the log as a link (patchset #1 id:1 of https://codereview.chromium.org/1257233006/ )

Reason for revert:
Blink merge is done, blink people won't have to use it anymore.

Original issue's description:
> roll_dep: Avoid large commit messages by providing the log as a link
> 
> The majority of chromium projects are on googlesource.com, which can
> display a range of commits on the web. Use this feature instead of
> dumping logs into the commit message.
> 
> Particularly, this avoids creating large commit messages as these logs
> sometimes contain upwards of 50 commits.
> 
> Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=296184

TBR=jochen@chromium.org,spang@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=296840

[M-A Ruel, 2015-09-23 13:45:29]

Created Revert of roll_dep: Avoid large commit messages by providing the log as
a link

[commit-bot: I haz the power, 2015-09-23 13:45:45]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1364763002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1364763002/1

[commit-bot: I haz the power, 2015-09-23 13:45:53]

Committed patchset #1 (id:1) as
http://src.chromium.org/viewvc/chrome?view=rev&revision=296840

[spang, 2015-09-23 15:09:39]

On 2015/09/23 13:45:53, commit-bot: I haz the power wrote:
> Committed patchset #1 (id:1) as
> http://src.chromium.org/viewvc/chrome?view=rev&revision=296840

My patch had nothing to do with blink, but with trace-viewer (now catapult) and
the fact that in the old (and now reverted) state, the signal/noise ratio for
roll commit messages is basically zero. In fact reverting removes the most
useful part (the googlesource link) since merely clicking on it verifies the git
hash is correct.

So I don't think this was properly handled. Is there an actual reason to revert?

[M-A Ruel, 2015-09-23 16:33:21]

On 2015/09/23 15:09:39, spang wrote:
> On 2015/09/23 13:45:53, commit-bot: I haz the power wrote:
> > Committed patchset #1 (id:1) as
> > http://src.chromium.org/viewvc/chrome?view=rev&revision=296840
> 
> My patch had nothing to do with blink, but with trace-viewer (now catapult)
and
> the fact that in the old (and now reverted) state, the signal/noise ratio for
> roll commit messages is basically zero. In fact reverting removes the most
> useful part (the googlesource link) since merely clicking on it verifies the
git
> hash is correct.
> 
> So I don't think this was properly handled. Is there an actual reason to
revert?

Yes, next time please ensure you look at git log/blame. Also the rationale for
the CL wasn't obvious, Jochen is not working on catapult so I couldn't see any
obvious connection there. Not as "I own the code" but more as "let's get input
first".

In that case I'm happy to bring the "skip descriptions for project X and Y" but
not as a blanket switch over. The link is nice, I just hadn't time to add it
first and reverting was easier at that point.

[spang, 2015-09-23 17:13:41]

On 2015/09/23 16:33:21, M-A Ruel wrote:
> On 2015/09/23 15:09:39, spang wrote:
> > On 2015/09/23 13:45:53, commit-bot: I haz the power wrote:
> > > Committed patchset #1 (id:1) as
> > > http://src.chromium.org/viewvc/chrome?view=rev&revision=296840
> > 
> > My patch had nothing to do with blink, but with trace-viewer (now catapult)
> and
> > the fact that in the old (and now reverted) state, the signal/noise ratio
for
> > roll commit messages is basically zero. In fact reverting removes the most
> > useful part (the googlesource link) since merely clicking on it verifies the
> git
> > hash is correct.
> > 
> > So I don't think this was properly handled. Is there an actual reason to
> revert?
> 
> Yes, next time please ensure you look at git log/blame. Also the rationale for
> the CL wasn't obvious, Jochen is not working on catapult so I couldn't see any
> obvious connection there. Not as "I own the code" but more as "let's get input
> first".

I definitely looked at the blame, however you were on vacation.

If you look our most active deps, none of them include a log. So I would say
this tool is out of whack with the rest of the project's gardening practices. 

> 
> In that case I'm happy to bring the "skip descriptions for project X and Y"
but
> not as a blanket switch over. The link is nice, I just hadn't time to add it
> first and reverting was easier at that point.

[spang, 2015-09-23 17:27:12]

On 2015/09/23 17:13:41, spang wrote:
> On 2015/09/23 16:33:21, M-A Ruel wrote:
> > On 2015/09/23 15:09:39, spang wrote:
> > > On 2015/09/23 13:45:53, commit-bot: I haz the power wrote:
> > > > Committed patchset #1 (id:1) as
> > > > http://src.chromium.org/viewvc/chrome?view=rev&revision=296840
> > > 
> > > My patch had nothing to do with blink, but with trace-viewer (now
catapult)
> > and
> > > the fact that in the old (and now reverted) state, the signal/noise ratio
> for
> > > roll commit messages is basically zero. In fact reverting removes the most
> > > useful part (the googlesource link) since merely clicking on it verifies
the
> > git
> > > hash is correct.
> > > 
> > > So I don't think this was properly handled. Is there an actual reason to
> > revert?
> > 
> > Yes, next time please ensure you look at git log/blame. Also the rationale
for
> > the CL wasn't obvious, Jochen is not working on catapult so I couldn't see
any
> > obvious connection there. Not as "I own the code" but more as "let's get
input
> > first".
> 
> I definitely looked at the blame, however you were on vacation.
> 
> If you look our most active deps, none of them include a log. So I would say
> this tool is out of whack with the rest of the project's gardening practices. 
> 

To be specific, those that don't include: blink, v8, webrtc, angle

skia is a notable exception.


> > 
> > In that case I'm happy to bring the "skip descriptions for project X and Y"
> but
> > not as a blanket switch over. The link is nice, I just hadn't time to add it
> > first and reverting was easier at that point.