Re: [gentoo-dev] The state and future of the OpenRC project

[Christopher Schwan <cschwan@students.uni-mainz.de>]

[-- Attachment #1: Type: text/plain, Size: 2818 bytes --]

On Monday 09 June 2014 21:45:26 hasufell wrote:
> Probably because no one mentored them on how to fix these QA issues.
> Otherwise... if that's attitude, then that's just sad and has to be
> fixed by those who run that overlay (review, contribution guidelines).
> 
> And I still think that the top 1 reason people run an overlay is because
> it's easier than contributing directly.
> A lot of overlay maintainers I tried to convince on getting more
> involved even said that.
> 
> Even sunrise workflow has proven too slow and cumbersome... look at the
> commit history, it's constantly decreasing.
> 
> Sure, reasons may vary, but there is not much positive to say about
> current gentoo workflow.

Since I was mentioned here - I am one of sage overlay[1] developers - it is 
maybe worth sharing my point of view as someone that is not a Gentoo 
developer. The recent discussion[2] with hasufell on our overlay might also be 
interesting.

I have to agree that indeed it is easier to contribute to an overlay, in my 
case because of

   a) git and
   b) github.

As far as I am up-to-date the main Gentoo repository is still managed by cvs, 
right? Thats something I really would not like to work with. Not because cvs 
is inferior to git (I hope no one feels offended) but because now I am so used 
to and pleased with the workflow that comes with git that I can not even think 
of changing that.

What I really love about github (and I don't think that similar platform 
differ much here) is the fact that it allows me to comment on /everything/. I 
can comment on commits that maybe lack something, which could then result in a 
new Issue; a user could be faster than I and create a pull request that fixes 
this problem. Maybe there is another mistake which I tell him there so he 
adjusts his commit and I finally accept it. Sometimes, after some months I 
wonder why we decided about some things the way we did and I can look it up in 
the Issue. There is the _complete_ discussion with direct references to the 
code and all people involved.

If there is a lession that I learned here, than that centralizing 
communications and development in one place was beneficial for us. Whether 
this also applies to the Gentoo project - well, I don't know. Right now there 
are several mailing lists, IRC channels and the Bugzilla so in comparison to 
our little overlay it is quite decentralized. Adding another way, e.g. via 
github, could complicate the situation but it will defintely get the project 
more user contributions (if that is needed).

In any case, I believe that the migration to git would be a huge step forward 
and it would attract at least one pontential new developer ;). Any update on 
this?

[1] https://github.com/cschwan/sage-on-gentoo
[2] https://github.com/cschwan/sage-on-gentoo/issues/294

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 490 bytes --]