[Nix-dev] [***SPAM***] Re: Announcing free-nix: the free Linux distributionbased on the Nix package manager

Michael Raskin 7c6f434c at mail.ru
Tue Jun 26 18:41:01 CEST 2012


>> What exactly *is* NixOS's policy?
>Also not clearly defined, but so far, it seems to have been
>functioning reasonably well without people feeling the need to fork.

Its lack also has created a few weird and unfiortunate outcomes, and
currently we seem to have a chance to get more.

>> Eelco revoked all access of all regular contributors without prior
>> warning. Until it suddenly happened, there was not the slightest
>> indication that we would be losing access to the infrastructure that we
>> have helped build over the last couple of years. Is that in your opinion
>> the proper way to handle a community project?
>
>You should not be so impatient here. Eelco has after years and years
>of complaints from people decided to switch stuff over to git. That
>should make most of us happy, shouldn't it? The transition has mostly
>been extremely painless. Now, Eelco has made a questionable initial

What exact definition of "painless" you use? Once history is there, SVN
is better in showing it to you than Git - so if only we have access to 
old revisions via SVN, we have no real problem with history; but there 
appeared a brand-new bottleneck even for minor package 
additions/updates. 

And I guess most contributors had some private workflow related to SVN
upstream, so it is not that nixos-checkout is the only thing to update.

>decision about commit access. You can complain about that, but I think
>it's much more worthwhile to make him understand that he's wrong on
>this particular decision than to use it as an incentive to split the
>community.

We already had related problem on community-culture level when for some
changes it was useless to discuss them (because no one replied) and it 
was discouraged to actually do something (because no reply is no ok).

Now it just reached some focal point.

>>  > I could somehow understand it if your decision was perceived by me as
>>  > the outcome of a failed attempt of convincing the project to go in
>>  > this direction, but afaics, you haven't even tried.
>>
>> You shouldn't jump to conclusions. I have expressed my dissatisfaction
>> with Eelco's nontransparent decisions to him in private e-mail. However,
>> he brushed me off, and told me that I don't need commit rights to NixOS
>> because we have a distributed VCS now, so I can commit wherever I want
>> instead of being restricted to his repository. Go figure.
>
>Alright. Several of us have had this discussion, I guess. And there's
>been a thread on the mailing list, too. My impression was always that
>Eelco had this opinion that we could all commit and send pull
>requests, but that he never indicated that he couldn't be convinced.
>On the contrary, I remember several occasions on which Eelco has
>indicated already that the current restrictions are preliminary, and
>that he's likely to move to a hybrid model.
>
>I still feel that just forking already is overreacting.

Well, isn't the reason for all the mess with the switch that cherry 
picking back from the fork is always easy?





More information about the nix-dev mailing list