[Bioc-devel] no more commits allowed before release ?

[Kevin Rue-Albrecht]

Thanks Robert, that is a great idea indeed. I like doing checks before
submitting a new release to the tracker. Avoids bad surprises.

Just to be clear, in this case, I wasn't expecting to see any change in the
build report (I had OKs everywhere, even when the bug existed). The bug
only occurred when using my package with microarray data, which I forgot to
test since I did relevant changes in my code. The only data I provide at
the moment with the package for testing is RNA-Seq, which I use in the
examples and the vignette. I occasionally test the package with an in-house
microarray dataset.

>From this experience, I believe it is best if I include a subset of my
microarray dataset as well. But at this stage I think it's better to wait
until the release is out. The package is now is a good working state, and
I'd rather not take risks just before the big day :)

Cheers
Kevin

On 10 October 2014 16:38, Robert M. Flight <rflight79 at gmail.com> wrote:

> If you want to be really sure that the changes fixed the problem without
> pushing to Bioconductor master (and seeing a build report), you can set up
> a Travis CI on a bugfix branch and see what happens there. The r-travis
> wiki has some good documentation
> https://github.com/craigcitro/r-travis/wiki, and my own package uses it,
> as well as others.
> https://github.com/rmflight/categoryCompare/blob/master/.travis.yml
>
> -Robert
>
> Robert M Flight, PhD
> Bioinformatics PostDoctoral Scholar
>  Resource Center for Stable Isotope Resolved Metabolomics
> Markey Cancer Center
> University of Kentucky
>  Lexington, KY
>
> Twitter: @rmflight <https://twitter.com/rmflight>
> Web: rmflight.github.io
> EM rflight79 at gmail.com
> PH 502-509-1827
>
> The most exciting phrase to hear in science, the one that heralds new
> discoveries, is not "Eureka!" (I found it!) but "That's funny ..." - Isaac
> Asimov
>
>
>
>
> On Fri, Oct 10, 2014 at 11:15 AM, Dan Tenenbaum <dtenenba at fhcrc.org>
> wrote:
>
>>
>>
>> ----- Original Message -----
>> > From: "Kevin Rue-Albrecht" <kevin.rue at ucdconnect.ie>
>> > To: "Dan Tenenbaum" <dtenenba at fhcrc.org>
>> > Cc: bioc-devel at r-project.org
>> > Sent: Friday, October 10, 2014 8:03:25 AM
>> > Subject: Re: [Bioc-devel] no more commits allowed before release ?
>> >
>> >
>> > Hi Dan,
>> >
>> >
>> > Ok good, you have just given me the most recent commit that I did
>> > today, so the ones I was most concerned about (from yesterday)
>> > should be there too, that's good for me. The changes should show up
>> > in the next build report.
>>
>> I do see a commit with the commit message "Bug fix...".
>>
>> > Maybe the build system just started the checks just before I pushed
>> > those changes yesterday morning (10am, Irish time.. so around 2am
>> > Seattle time, I believe), which could have left them out of the
>> > build report.
>> >
>>
>> Changes need to be committed by 4:20 PM Seattle time in order to show up
>> in the following day's build report.
>>
>>
>> >
>> >
>> > I always run CMD check locally before submitting. That wasn't the
>> > problem, it was just a scenario that I hadn't tested in a while. I
>> > am considering throwing some unit tests in the package soon, but I
>> > won't have time before the release.. I'll have to spend some time to
>> > do it correctly as it'd be my first time writing unit tests for an R
>> > package (did it in other languages before).
>> >
>> >
>> >     * Are the unit tests run by the BiocCheck ? In which case, I
>> >     would need to add another testing dataset corresponding to the
>> >     scenario that I fixed yesterday.
>>
>>
>> Yes, unit tests are run by BiocCheck, our howto page shows you how to
>> make this happen:
>> http://www.bioconductor.org/developers/how-to/unitTesting-guidelines/
>>
>> Dan
>>
>> >
>> >
>> > Cheers
>> > Kevin
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> > On 10 October 2014 15:40, Dan Tenenbaum < dtenenba at fhcrc.org > wrote:
>> >
>> >
>> >
>> >
>> > ----- Original Message -----
>> > > From: "Kevin Rue-Albrecht" < kevin.rue at ucdconnect.ie >
>> > > To: bioc-devel at r-project.org
>> > > Sent: Friday, October 10, 2014 7:31:36 AM
>> > > Subject: [Bioc-devel] no more commits allowed before release ?
>> > >
>> > > Dear all,
>> > >
>> > > I fixed a bug affecting one particular situation in my package
>> > > (GOexpress)
>> > > yesterday, and pushed the changes to GitHub. However, the webhook
>> > > did
>> > > not
>> > > trigger to update the page, due to the feature freeze if I
>> > > understand
>> > > the
>> > > schedule correctly
>> > > ( http://www.bioconductor.org/developers/release-schedule/
>> > > ).
>> > >
>> >
>> > I see a commit at 2014-10-10 13:13:54 +0100 (your time) / 2014-10-10
>> > 05:11:29 -0700 (Seattle time).
>> >
>> > Is that the one? The commit message is " Little addition of details
>> > to the description of DAM and NCN participation in the project."
>> >
>> > We do not disable commits to trunk. We have disabled commits to the
>> > current release branch (Bioconductor 2.14) because 2.14 builds have
>> > stopped.
>> >
>> > > I suppose I need to wait the release is out to push these changes
>> > > again to
>> > > update the SVN ? Or is it still possible to include those changes
>> > > in
>> > > the
>> > > release? From the schedule, it still sounds possible "Package
>> > > maintainers
>> > > should limit changes to "show-stopper" bugs and documentation
>> > > improvements."
>> > >
>> >
>> > It is possible, but there are very few build cycles left till the
>> > release, so please build and check your package locally before
>> > committing changes and then keep an eye on the build report to make
>> > sure that the package built ok.
>> >
>> > Dan
>> >
>> >
>> > > Sincerely,
>> > > Kevin
>> > >
>> > > --
>> > > Kévin RUE-ALBRECHT
>> > > Wellcome Trust Computational Infection Biology PhD Programme
>> > > University College Dublin
>> > > Ireland
>> > > http://fr.linkedin.com/pub/k%C3%A9vin-rue/28/a45/149/en
>> > >
>> > > [[alternative HTML version deleted]]
>> > >
>> > > _______________________________________________
>> > > Bioc-devel at r-project.org mailing list
>> > > https://stat.ethz.ch/mailman/listinfo/bioc-devel
>> > >
>> >
>> >
>> >
>> >
>> > --
>> >
>> > Kévin RUE-ALBRECHT
>> > Wellcome Trust Computational Infection Biology PhD Programme
>> > University College Dublin
>> > Ireland
>> > http://fr.linkedin.com/pub/k%C3%A9vin-rue/28/a45/149/en
>> >
>>
>> _______________________________________________
>> Bioc-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>>
>
>


-- 
Kévin RUE-ALBRECHT
Wellcome Trust Computational Infection Biology PhD Programme
University College Dublin
Ireland
http://fr.linkedin.com/pub/k%C3%A9vin-rue/28/a45/149/en

	[[alternative HTML version deleted]]