[Qgis-developer] Please stop spamming commit logs!

Anita Graser anitagraser at gmx.at
Fri Oct 10 02:26:57 PDT 2014


As Jürgen mentioned, we discussed this:
http://hub.qgis.org/wiki/quantum-gis/PSC_Meeting_7_Feb_2014
My interpretation is that the PSC agreed on allowing giving credit to
funders in commit messages. Obviously I promised to write something
along these lines in the governance docs, but that might have gotten
lost in my todo pile:


16:20 <@anitagraser> 1. would be "giving credit to funders"
16:21 <@anitagraser> ideas were: a) in code/comment, b) in commit
messages, c) in release log, .... any more?
----------------------------- >8 -----------------------------
16:21 < pcav> I think commits are most appropriate
16:22 < pcav> but we should agree on a standard formula
----------------------------- >8 -----------------------------
16:22 < pcav> like "funded by ..." and a max N of words
----------------------------- >8 -----------------------------
16:22 < pcav> and decide links yes/no (I'd vote no)
----------------------------- >8 -----------------------------
16:23 < duiv> if we only do it in the release log, it does not
accumulate in the code
----------------------------- >8 -----------------------------
16:24 < duiv> and we do not have trouble with 'first version funded by
A', second by B and finally fixed by C
----------------------------- >8 -----------------------------
16:25 <@anitagraser> vote for b)?
----------------------------- >8 -----------------------------
16:25 < pcav> +1
16:25 <@anitagraser> +1
16:26 < jef> we vote for what? allow commiters to add a comment about
how funded their work and forbid to put it into the source?
16:27 < duiv> if it is ok for the funders: +1 but if I would for
example have raised a couple of thousand euro's to implement
something... not sure if I would be ok with a commit msg
16:27 < jef> s/add a comment/add a comment into the commit message/
16:27 < pcav> they can always refer to the link
16:27 < pcav> and put it in their own announcements
----------------------------- >8 -----------------------------
16:32 < jef> duiv: a comment in the commit is acceptable to me -
although it's unnessary clutter, too.
----------------------------- >8 -----------------------------
16:33 < duiv> ok, so the verdict? We prefer to do credits in commit
msgs, not in code anymore?
16:33 < pcav> me and anitagraser voted
16:33 < duiv> +1
16:34 < pcav> jef: ?
16:34 < jef> 0
16:34 < jef> guidelines... ;)
----------------------------- >8 -----------------------------
16:35 <@anitagraser> somewhere here
http://www.qgis.org/en/site/getinvolved/governance/index.html?
----------------------------- >8 -----------------------------
16:38 <@anitagraser> duiv: i can write two or three lines to start with
16:39 < duiv> plz do, and add it in the governance docs in one go

Best wishes,
Anita



On Fri, Oct 10, 2014 at 10:18 AM, Matthias Kuhn <matthias.kuhn at gmx.ch> wrote:
> I agree with Nathan and Nyall. The commit history isn't a very "official"
> thing, so there is some room for attribution and other additional
> information.
>
> Personally I prefer a meaningful commit message with some "spam" in it over
> a commit message that contains (almost) no useful information like "Fix
> #1234", "Fix #4567 [Meaningless title of an issue report]" "Followup 65443"
> (That one is not so bad, but could be improved with some prose). I often
> find myself looking at the commit history to find information about why
> something was done.
> IF something needs to be fixed in the commit log, then we should rather
> focus on this than on a bit of pride, fun and attribution.
>
> I also think that a list of funders/sponsors for a particular version would
> be nice. This could be directly below the changelog or linked there (I am
> rather thinking of a list than each individual change).
>
> Regards,
> Matthias
>
>
>
> On 10/10/2014 10:01 AM, Nathan Woodrow wrote:
>
> I'm with Nyall.  I see no direct reason this is a bad idea.  If I do
> something in my free time I don't care about getting recognition for it
> because my direct work on the project is enough and if someone wants to see
> what I do they can check my commit history or blog, however if I commit
> something, say a feature that took be a while to make, on work time under my
> employers name for them I think it's worth noting that they sponsored that
> work.  Their contribution can be traced to that single commit and pulled
> from the log.
>
> The place it gets tricky is if you are running your own business committing
> all the time for work reasons.
>
> On the same note I do think it's worth having a page for each release with a
> list of users who were active and sponsors that did the work.  This would
> mean everyone gets highlighted for their work.  It doesn't need to be name +
> feature type of list, just a list of names is enough..
>
> - Nathan
>
> On Fri, Oct 10, 2014 at 5:50 PM, Nyall Dawson <nyall.dawson at gmail.com>
> wrote:
>>
>> On 10 October 2014 18:11, Alessandro Pasotti <apasotti at gmail.com> wrote:
>> > Hi,
>> >
>> > I noticed that somebody started to add commercials to commit logs:
>> >
>> > Sponsored by ....
>> >
>> > Funded by ...
>> >
>> > etc. etc.
>> >
>> > We should take this seriously, mostly ever developer works for a
>> > company or run its own business.
>> >
>> > Imagine if everybody starts adding those (not really useful) sentences
>> > to every commit.
>> >
>>
>> Is this really an issue? It seems rather trivial. I personally am
>> strongly in favour of these attributions in the commit log. Reasons
>> are:
>>
>> - It gives credit to sponsors. That's important! Look at how many cool
>> features were added in 2.6 thanks to sponsorship...
>> - It gives credit to developers who donate their free/company time.
>> That's also important. QGIS wouldn't exist if it wasn't for these
>> developers donating their time
>> - The commit log is basically for developers or power
>> users/contributors only. It's a fairly harmless place to advertise
>> these sponsorship messages. For a while there was a few "sponsored by"
>> messages in code comments - that's a much worse/more intrusive place
>> for these messages.
>> - It lets us blow off steam when release pressures ramp up :P see
>> 68c49fe09, 34f00d106 and 2427546d8
>>
>> So, +1 for allowing these messages.
>>
>> Nyall
>> _______________________________________________
>> Qgis-developer mailing list
>> Qgis-developer at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/qgis-developer
>
>
>
>
> _______________________________________________
> Qgis-developer mailing list
> Qgis-developer at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/qgis-developer
>
>
>
> _______________________________________________
> Qgis-developer mailing list
> Qgis-developer at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/qgis-developer


More information about the Qgis-developer mailing list