Sent out a proposed CL to remove the check to
https://codereview.chromium.org/1956993002/

On Mon, Apr 11, 2016 at 9:04 AM, Toon Verwaest <[email protected]>
wrote:

> Ping. Was just bitten by this again :)
>
> On Thu, Jan 21, 2016 at 11:04 PM 'Daniel Ehrenberg' via v8-dev <
> [email protected]> wrote:
>
>> To clarify, I wasn't arguing that all of those workflows should have BUG=
>> lines.
>>
>> Dan
>>
>> On Thu, Jan 21, 2016 at 10:41 AM, Adam Klein <[email protected]> wrote:
>> > I'd like to separate the questions of requiring LOG= lines and requiring
>> > BUG= lines. I think the former is clearly not doing much for us (given
>> my
>> > exploration of the data). For the latter, I think there's a variety of
>> > workflows used by folks on the team, and forcing tracking bugs for every
>> > patch seems like a major shift in process.
>> >
>> > On Thu, Jan 21, 2016 at 4:03 AM, Michael Hablich <[email protected]>
>> wrote:
>> >>
>> >> I agree with Dan. I think it make more sense to simply use a tracking
>> bug.
>> >>
>> >>   +Michael Achenbach any objections?
>> >>
>> >>
>> >> On Wed, Jan 20, 2016 at 10:37 PM Adam Klein <[email protected]>
>> wrote:
>> >>>
>> >>> Some stats on the ChangeLog in the 4.9 branch:
>> >>>
>> >>> - Out of 385 minor revisions, 60 contained non-trivial ChangeLogs
>> >>> - A total of 75 commits were mentioned in those 60 revisions
>> >>>
>> >>> Does anyone here object to getting rid of the LOG= requirement? With
>> >>> fewer than 1 in 5 minor revisions containing a commit log, I highly
>> doubt
>> >>> that it's a useful signal. I'll send a message out to v8-users unless
>> I hear
>> >>> objections.
>> >>>
>> >>> On Mon, Jan 11, 2016 at 1:06 PM, Jakob Kummerow <
>> [email protected]>
>> >>> wrote:
>> >>>>
>> >>>> On Mon, Jan 11, 2016 at 8:29 PM, Adam Klein <[email protected]>
>> wrote:
>> >>>>>
>> >>>>> Looking through that ChangeLog it doesn't look particularly useful
>> >>>>
>> >>>>
>> >>>> Agreed.
>> >>>>
>> >>>>>
>> >>>>> (most releases have no log lines, though I suspect many of them do
>> have
>> >>>>> significant changes).
>> >>>>
>> >>>>
>> >>>> I'm not so sure -- our tagged versions are really fine-grained these
>> >>>> days, sometimes every single commit is tagged, so it could well be
>> that many
>> >>>> of those aren't ChangeLog-worthy (which of course doesn't help with
>> the
>> >>>> ChangeLog's lack of usefulness.)
>> >>>>
>> >>>>> What if we relaxed the presubmit to allow BUG= lines without LOG=
>> >>>>> lines? That would solve my main problem, while still letting people
>> add
>> >>>>> LOG=Y lines if they think it's appropriate. The fact that the
>> presubmit only
>> >>>>> complains for patches that have a BUG line seems like a perverse
>> incentive
>> >>>>> to avoid using tracking bugs.
>> >>>>
>> >>>>
>> >>>> The rationale was that visible features should be tracked by bugs
>> >>>> anyway, and requiring the LOG line forces people to make a decision
>> whether
>> >>>> a given change should be mentioned in the log or not, as we felt
>> that we
>> >>>> couldn't come up with a reasonable way to automate it.
>> >>>>
>> >>>> I suspect that making the LOG line optional is pretty much
>> equivalent to
>> >>>> getting rid of it, as people will start to forget about it.
>> >>>>
>> >>>> Which is fine with me :-)
>> >>>>
>> >>>>>
>> >>>>> On Mon, Jan 11, 2016 at 12:43 AM, Jakob Kummerow
>> >>>>> <[email protected]> wrote:
>> >>>>>>
>> >>>>>> Yeah, the ChangeLog is still generated from LOG=y commits:
>> >>>>>>
>> https://chromium.googlesource.com/v8/v8/+/branch-heads/4.8/ChangeLog
>> >>>>>> (Note that the master branch's ChangeLog is not updated
>> automatically;
>> >>>>>> syncing it to the latest branch-head is a manual process that
>> apparently
>> >>>>>> hasn't been done in half a year.)
>> >>>>>>
>> >>>>>> We've discussed getting rid of the ChangeLog repeatedly over the
>> >>>>>> years. Nobody on our team needs/uses it, however there do seem to
>> be some
>> >>>>>> external people reading it; or at least that was the conclusion
>> last time
>> >>>>>> this came up. We could ask on v8-users if anyone would be unhappy
>> if we
>> >>>>>> dropped it, but that question doesn't necessarily reach everyone
>> who might
>> >>>>>> care, so it's hard to be sure.
>> >>>>>>
>> >>>>>>
>> >>>>>> On Mon, Jan 11, 2016 at 9:23 AM, 'Michael Hablich' via v8-dev
>> >>>>>> <[email protected]> wrote:
>> >>>>>>>
>> >>>>>>> My impression is that LOG=Y should be used for CLs which are
>> notable/
>> >>>>>>> should be on a changelog. I think some infra scripts are using
>> this
>> >>>>>>> information but machenbach@ surely knows more about this.
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> On Fri, Jan 8, 2016 at 8:45 PM Adam Klein <[email protected]>
>> wrote:
>> >>>>>>>>
>> >>>>>>>> Back in the day, when we were still on SVN and all V8 releases
>> were
>> >>>>>>>> blobs of merged changes, the LOG=Y served a good purpose, giving
>> those SVN
>> >>>>>>>> commits on the release branch human-readable change descriptions.
>> >>>>>>>>
>> >>>>>>>> Now that our releases are merge commits on branches with full CL
>> >>>>>>>> descriptions, and moreover the release commits contain a link to
>> a full list
>> >>>>>>>> of changes, I'm not sure the LOG stuff is buying us anything
>> other than nags
>> >>>>>>>> from the presubmit script.
>> >>>>>>>>
>> >>>>>>>> How would folks feel about dropping it?
>> >>>>>>>>
>> >>>>>>>> - Adam
>> >>>>>>>
>> >>>>>>> --
>> >>>>>>> --
>> >>>>>>> v8-dev mailing list
>> >>>>>>> [email protected]
>> >>>>>>> http://groups.google.com/group/v8-dev
>> >>>>>>> ---
>> >>>>>>> You received this message because you are subscribed to the Google
>> >>>>>>> Groups "v8-dev" group.
>> >>>>>>> To unsubscribe from this group and stop receiving emails from it,
>> >>>>>>> send an email to [email protected].
>> >>>>>>> For more options, visit https://groups.google.com/d/optout.
>> >>>>>>
>> >>>>>>
>> >>>>>
>> >>>>
>> >
>> > --
>> > --
>> > v8-dev mailing list
>> > [email protected]
>> > http://groups.google.com/group/v8-dev
>> > ---
>> > You received this message because you are subscribed to the Google
>> Groups
>> > "v8-dev" group.
>> > To unsubscribe from this group and stop receiving emails from it, send
>> an
>> > email to [email protected].
>> > For more options, visit https://groups.google.com/d/optout.
>>
>> --
>> --
>> v8-dev mailing list
>> [email protected]
>> http://groups.google.com/group/v8-dev
>> ---
>> You received this message because you are subscribed to the Google Groups
>> "v8-dev" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>> For more options, visit https://groups.google.com/d/optout.
>>
> --
> --
> v8-dev mailing list
> [email protected]
> http://groups.google.com/group/v8-dev
> ---
> You received this message because you are subscribed to the Google Groups
> "v8-dev" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
>

-- 
-- 
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
--- 
You received this message because you are subscribed to the Google Groups 
"v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to