| 1 |
On Mon, Dec 2, 2019 at 4:10 AM Michał Górny <mgorny@g.o> wrote: |
| 2 |
> |
| 3 |
> === |
| 4 |
> Should Gentoo developers be expected to be able to find a way to work |
| 5 |
> together to build a somewhat consistent distribution, or should it be |
| 6 |
> allowed for individual developers to 'run their own shops' and ignore |
| 7 |
> everybody else? |
| 8 |
> === |
| 9 |
> |
| 10 |
> To explain what I mean with an extreme example: in the past we had |
| 11 |
> a developer who blocked adding systemd units to his own packages because |
| 12 |
> of his own personal agenda against systemd. This was obviously harmful |
| 13 |
> to systemd users who couldn't install use those packages normally, |
| 14 |
> and was harmful to consistent user experience across Gentoo. |
| 15 |
|
| 16 |
Without wanting to start a flame war, is there a more contemporary |
| 17 |
example, or even a hypothetical? |
| 18 |
|
| 19 |
That particular systemd issue was resolved years ago by escalation to |
| 20 |
the Council, which set a policy that was a reasonable compromise, and |
| 21 |
it appears to have worked as there has been little argument over that |
| 22 |
particular matter since. |
| 23 |
|
| 24 |
I'd assume that if other specific issues come up, they would be also |
| 25 |
escalated, and Council would again set a reasonable policy that |
| 26 |
everybody could live with. |
| 27 |
|
| 28 |
I guess we could ask the Council to have a purely ideological |
| 29 |
discussion, but IMO it isn't necessary. If we didn't want to have a |
| 30 |
mechanism to ensure a reasonable level of consistency we wouldn't have |
| 31 |
the Council in the first place. Having a mechanism to resolve these |
| 32 |
conflicts won't prevent them from happening in the first place. I |
| 33 |
think that Gentoo's key attributes of being flexible/FOSS/etc probably |
| 34 |
attract people with strong opinions on matters, and there are bound to |
| 35 |
be cases where they can't be resolved by mutual agreement. |
| 36 |
|
| 37 |
I'll go ahead and toss out a more recent example without naming |
| 38 |
names/etc (not that anybody couldn't find it if they went looking). |
| 39 |
There was some heated discussion over disagreement on appropriate use |
| 40 |
or naming of a USE flag. Ultimately the package maintainer made the |
| 41 |
call (at least implicitly via their commit). There was some |
| 42 |
disagreement over that call. Nothing was formally escalated to |
| 43 |
QA/Council/etc or any other technical governing body. Is that an |
| 44 |
example of the system working as intended (if the issue were important |
| 45 |
enough it would have been escalated, and work got done without undue |
| 46 |
delay), or is it an example of it being broken (nobody wants to have |
| 47 |
to deal with escalating things, so maybe the maintainer should have |
| 48 |
sought approval/consensus/etc first, plus if it were escalated now |
| 49 |
we're changing things already in the repo)? I could cite the |
| 50 |
specifics if needed - my goal is to raise a contemporary example that |
| 51 |
might or might not be what you're getting at, without dragging actual |
| 52 |
debate over that specific issue into this thread... |
| 53 |
|
| 54 |
-- |
| 55 |
Rich |
Archives