[Bioc-devel] Automatic build failure notifications

[Dan Tenenbaum]

----- Original Message -----
> From: "Hervé Pagès" <hpages at fredhutch.org>
> To: "Emma Schymanski" <Emma.Schymanski at eawag.ch>, "bioc-devel" <bioc-devel at r-project.org>
> Cc: "Dan Tenenbaum" <dtenenba at fredhutch.org>
> Sent: Tuesday, October 11, 2016 10:55:20 AM
> Subject: Re: Automatic build failure notifications

> OK, will do. Thanks Emma for your feedback.
> 

I like the idea of opting in for this. Sometimes there are false positives (for example, a package might fail because its dependency failed for some reason, or there is an issue with the build system) that the package author can do nothing about. I imagine that there might be a lot of traffic on this list asking what should be done about such emails, or complaining about receiving such emails. The way the notifications are currently set up is that they are sent once a week and that could get frustrating for people if the emails are false positives.

These sort of false positives were less common in release which was probably why these notifications were only enabled in release before.

On the other hand, failing to opt in would mean some maintainers would not see true positives. 


My $0.02.
Dan


> H.
> 
> On 10/10/2016 11:22 PM, Schymanski, Emma wrote:
>> Hi Herve, Markus, all,
>>
>>
>>
>> It would really be great if these automatic notifications would be on
>> again –
>>
>> and if this would happen for devel and release. After my last email Dan T.
>>
>> suggested I sign up for the RSS feed – I followed the instructions but even
>>
>> there I seem only to get an old feed of the release build report – which
>> for us is
>>
>> a different error [one that is relates to Java … and has been going on
>> for ages]
>>
>>
>>
>> Manually checking the builds every days is a pain when we know we are
>>
>> waiting for a fix – if there would be a way to automate devel build reports
>>
>> (maybe an opt in so those that don’t want the traffic don’t go crazy), that
>>
>> would be great. I’m sure there’d be quite a few of us that would really
>>
>> appreciate that! We’d have found our error much quicker that way...
>>
>>
>>
>> I would vote for daily in the run up to the release, if doable. A number of
>>
>> us seem to still have some need for this ;)
>>
>>
>>
>> Thanks,
>>
>> Emma
>>
>>
>>
>>>> We would also
>>
>>>> appreciate it to have the e-mail system for failed builds / checks back
>>
>>>> online :-)
>>
>>
>>
>>>Just to clarify, in the past we've only activated the automatic build
>>
>>>failure notifications for the release version of Bioconductor (except
>>
>>>for the current release i.e. BioC 3.3, for which we forgot).
>>
>>>
>>
>>>Our plan was to activate them for BioC 3.4 after it's released.
>>
>>>But we could do it now if nobody objects. We usually run the script
>>
>>>once a week as a cron job but we could change this to run it 2 or
>>
>>>3 times/week until the release (which is in 8 days from now), and
>>
>>>then go back to once a week after the release.
>>
>>>
>>
>>>Cheers,
>>
>>>H.
>>
>>
>>
>>
>>
> 
> --
> Hervé Pagès
> 
> Program in Computational Biology
> Division of Public Health Sciences
> Fred Hutchinson Cancer Research Center
> 1100 Fairview Ave. N, M1-B514
> P.O. Box 19024
> Seattle, WA 98109-1024
> 
> E-mail: hpages at fredhutch.org
> Phone:  (206) 667-5791
> Fax:    (206) 667-1319