[Sugar-devel] Engineering Team

[Simon Schampijer]

On 11/20/2010 02:06 PM, Aleksey Lim wrote:
> On Sat, Nov 20, 2010 at 04:27:49AM +0000, Martin Dengler wrote:
>> On Sat, Nov 20, 2010 at 02:53:06AM +0000, Aleksey Lim wrote:
>>> Hi all,
>>>
>>> (Just trying to summarize previous discussions, including recent one
>>> on #sugar with some kind of motion)
>>>
>>>
>>> = The problem =
>> [...]
>>> = Engineering Team [the solution] =
>> [...]
>>
>> I'm not opposed to the discussion.  While we (continue to) have it, am
>> I wrong in thinking the de factor situation is that:
>>
>> - anyone who is allowed push to sugar* repos is allowed to approve and
>>    / or push a patch or other code (and is trusted to not push
>>    contentious things)
>
> The practice was (if got it right), only glucose maintainers (discussing
> with their peers) might say "please push"
> http://wiki.sugarlabs.org/go/Development_Team/Release/Modules#Glucose

Yes, commit access means that you can do the commit yourself. Not that 
you commit without review.

> When Tomeu stepped aside, we have "unmaintained" for sugar. Another
> important module is maintained by Simon, but he is busy with OLPC work.

And this is a typical case. That is why separating the workload makes sense.

> The idea is having several people who can say "please push" for all
> glucose modules. Maybe better to have such people per module, but we
> have only 4 modules (I guess people will manage to work together),
> moreover there might situations when patches/features affect several
> core modules eg ds and shell.

Actually the previous model was mainly like that. For example as the 
shell co-maintainer I was as well eligible to give my ok for a push to 
the shell. Back then when we did the original separation of the work 
load we just decided on one main maintainer per module and added peers 
so we were not blocking on one person (someone is on vacation and a new 
release is needed, someone steps down etc).

>> - people who are not allowed to push to sugar should do this to get
>>    code in: http://wiki.sugarlabs.org/go/Development_Team/Code_Review
>
> All patch/feature providers need to follow this rule (including glucose
> maints who need to patch). That was in previous scheme and I think would
> be useful to keep this way for the new one.

Yes, we always reviewed all the patches, including the ones from 
maintainers. I think this is important and should stay like this.

Regards,
    Simon