Regression notifications (was Re: [squeak-dev] The Trunk: Tools-fbs.450.mcz)

Frank Shearar frank.shearar at gmail.com
Wed Mar 6 19:18:07 UTC 2013


On 6 March 2013 14:51, Bert Freudenberg <bert at freudenbergs.de> wrote:
>
> On 2013-03-06, at 14:24, Frank Shearar <frank.shearar at gmail.com> wrote:
>
>>> getting the regressions sent to squeak-dev would be awesome!
>>
>> The rules for email notification are:
>>
>> If configured, Jenkins will send out an e-mail to the specified
>> recipients when a certain important event occurs.
>> Every failed build triggers a new e-mail.
>> A successful build after a failed (or unstable) build triggers a new
>> e-mail, indicating that a crisis is over.
>> An unstable build after a successful build triggers a new e-mail,
>> indicating that there's a regression.
>> Unless configured, every unstable build triggers a new e-mail,
>> indicating that regression is still there.
>
> Well, if we can filter these to only send a mail if there are failures that were not in the previous build, that would be what we want.

Not as far as I know.

>> For lazy projects where unstable builds are the norm, Uncheck "Send
>> e-mail for every unstable build".
>>
>> SqueakTrunk is always unstable (because we always have failing tests
>> since some are markers for desired behaviour)
>
> Would it be hard not to log expected failures?

Yes, but only by using expectedFailures, which is kind've lame. There
are two tests - the Locale one and the one ReleaseBuilder one - that
I'd really like to see fixed one way or the other, and we could always
just expectedFailure the others. A separate branch could unexpect one
when someone's ready to tackle the issue.

>> so we'd get a mail on
>> every build. But I don't know yet what mail address Jenkins will use
>> here.
>>
>> frank
>
> When I was experimenting with email notifications I'd always use a couple private emails for a couple of days first ...

For sure; for testing I'd just mail myself.

frank

> - Bert -
>
>
>


More information about the Squeak-dev mailing list