Re: [Fedora-spins] Proposal: Spins process amendment for Fedora 20 cycle
On 07/29/2013 01:20 AM, Ian Malone wrote:
On 27 July 2013 17:24, Kevin Fenzi <kevin(a)scrye.com> wrote:
> On Sat, 27 Jul 2013 08:22:25 +0200
> Brendan Jones <brendan.jones.it(a)gmail.com> wrote:
>
>> On 07/26/2013 05:36 PM, Kevin Fenzi wrote:
>>> So, there was some feedback (mostly from QA), but pretty much
>>> nothing from spin owners.
>>>
>>> I guess next step is for me to write up the proposal again with
>>> feedback pulled in and post to devel?
>>>
>>> Any other feedback before I do so?
>>> Or anyone else have desire to write up things? :)
>>>
>>> kevin
>>>
>> The Fedora Jam team (music-creation SIG) can manage at least 2 test
>> matrices for F20. I think we have some good support there.
>
> Cool.
>
>> But our case is somewhat easier - we are simply adding packages to
>> the KDE spin. It is not as if we are providing an alternate desktop
>> or anything. Any errors we catch will probably be picked up by QA as
>> well.
>
> Sure, but there could be errors in the packages you add, their
> interaction with the rest of the desktop or compose issues, etc.
True, though what we mainly find if anything (unless a distro-wide or
KDE spin bug) seems to be compose issues. Actually it's a pity the
formulas idea hasn't really caught anyone's imagination yet.
I'm really not sure how this is supposed to work. I am hesitant to do
anything outside of RPM. Matthew M's flock proposal also makes me
somewhat uneasy. Maybe I need to keep up with the times...
If rings or formulas are just sets of packages, we have that already. If
they are alternate config files based on use cases in /etc/sysconfig I
can see the merit in that, but I would rather install an application
which changes those things and let me know when its doing it.