[mapserver-dev] RFC-7.2: MapServer Git Push Management

[Stephan Meißl]

On 10/03/2012 09:43 AM, thomas bonfort wrote:
> Thanks Tom for taking this.
> 
> On Tue, Oct 2, 2012 at 9:04 PM, Lime, Steve D (DNR)
> <Steve.Lime at state.mn.us> wrote:
>> Thanks for doing this Tom. Couple of things to discuss:
>>
>>
>>
>> 1)      We’ve talked about getting rid of HISTORY.TXT and somehow relying on
>> commit comments. Any thoughts on this?
> 
> How I envision this is to keep HISTORY.TXT as a means to highlight the
> noteworthy additions to mapserver in major releases. The reason for
> *not* updating HISTORY.TXT in stable branches is twofold:
>  - The changes we publicize between stable versions need to be
> complete. The commit log is a good source for this, as it does not
> rely on manually updating the history file.
>  - The goal is to ease fixes being committed to the stable branch
> rather than only the development one, and updating HISTORY in the
> stable branch will always cause merge conflicts when merging the
> stable branch back into the master one.
> 
+1 for easier merges
> 
>>
>> 2)      In the Git Commit Practices section there’s mention of msautotest.
>> That’s changed now with the test on commit testing and “make test” support.
> 
> The tests are still added to msautotest. make test and the continuous
> integration are just wrappers around running the original msautotest
> suite. I would like to include a requirement (or at least a very
> strong recommendation) that all new features get a batch of autotests,
> if possible also exercising corner cases.
> Regarding tests, and given that our test suite is now fully passing, I
> would also like to include a requirement that a commit does not break
> any of the existing tests (or that the test suite be updated with
> expected results if required). Using pull requests rather than
> directly committing to the branches is a practical way of testing
> proposed changes before inclusion.
> How do you guys feel about adding these to RFC7.2 ? /me feels a little
> bit awkward at forcing down these kind of rules, but think that they
> will be beneficial in the long run.
> 
+1 IMHO, using the proposed process and also documenting it is a great
sign of quality to users
>>
>> 3)      Should the coding style (astyle) be mentioned?
> Good point. I'll update the RFC accordingly.
>>
>>
>>
>> Steve
>>
>>
>>
>> From: mapserver-dev-bounces at lists.osgeo.org
>> [mailto:mapserver-dev-bounces at lists.osgeo.org] On Behalf Of Tom Kralidis
>> Sent: Tuesday, October 02, 2012 12:57 PM
>> To: mapserver-dev at lists.osgeo.org
>> Subject: [mapserver-dev] RFC-7.2: MapServer Git Push Management
>>
>>
>>
>> FYI given our move to Git/GitHub, RFC 7.1 needs updating.
>>
>> http://www.mapserver.org/development/rfc/ms-rfc-7.2.html now obsoletes 7.1,
>> and thus we need review/comment and subsequent vote.
>>
>> ..Tom
>>
>>
>> _______________________________________________
>> mapserver-dev mailing list
>> mapserver-dev at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/mapserver-dev
>>
> _______________________________________________
> mapserver-dev mailing list
> mapserver-dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/mapserver-dev
>