Re: [gentoo-pms] kdebuild-1 conditionales

[Christian Faulhammer]

Hi,

after a week of real-life work, I was able to catch up with the whole
mail mess, thus doing a dump of thoughts here.

Ciaran McCreesh <ciaran.mccreesh@××××××××××.com>:
> On Thu, 10 Dec 2009 21:48:48 +0100
> Christian Faulhammer <fauli@g.o> wrote:
> > in commit 6a281c0bc6b951c0885c8787fa5c353a4f4e3d0d I disabled
> > kdebuild-1 by default.  My proposal now is to drop the KDEBUILD
> > conditionals as a whole as the overlay has gone anyway, where it was
> > used.  We can add a sentence in the introduction or wherever which
> > says something along the lines like "kdebuild-1 was the first EAPI
> > like format that supported extended features added to official EAPIs
> > later on and was heavily tested in the official Gentoo KDE overlay".
> > This would ease maintenance a bit.
>
> As we've already discussed:
>
> * Stop committing things that aren't typo fixes without posting them
> to this list for review.

 They are still administrative things reflecting a council decision and
setting the repo to official document generation by default.
 The whole "get rid of detailled kdebuild-1 description" has nothing to
do with denying that kdebuild-1 was one of the first steps towards
EAPI in Gentoo, but it eases the maintenance burden. LaTeX code is not
easier to read when a lot of conditionals are applied.  Put a warning
in Paludis when an kdebuild-1 is detected and I also support your news
item here.

> * Don't commit the EAPI 3 / 4 changes until the Council are done
>   changing things, and until we have a patch for the new definition of
>   EAPI 3. We don't have a definition for the new EAPI 3 yet. We also
>   don't have approved summaries of any of the meetings where these
>   things happened. Any changes done now are wasted effort.

 As I spoke with council members and people attending it before doing
the commits, I think I know about the intentions.  What Zac or anyone
else is doing is not to be intermixed with my actions as we acted
independently from each other.  So let's stick to one topic and I will
justify my commits now:

Disable kdebuild-1 by default: We had the discussion several times and
your only argument now is that there might be consumers of an
never-approved EAPI out there.
Update bash version: This reflects a council decision and two people
had input from you and others about the patch.  I discussed it with
Thomas Anderson on #-dev and we agreed on a wording which I committed.
3 to 4 move: Purely administrative and has been worked on by two people
(ulm and myself).
Cheat note: The commit comment is wrong and is not what I intended to
say in the blob itself.  So I will revert that piece of code as it was
a shoot from the hip and not thought through.

Anyway, yes, reviewing is necessary, but if essential changes from my
point of view are blocked or stonewalled through that means, I may
choose to take action.eas

> * Don't mess with kdebuild until you're sure that no-one has any
>   kdebuild packages installed.

 Don't be too academic.  To be sure is not possible.  And please don't
speak about bridge construction and failure possibilites when you don't
know about how an engineering process works.

> * When the heck did "use the highest EAPI" become policy?

 Maybe I mixed up some discussion in -dev with some policy agreement.

V-Li

-- 
Christian Faulhammer, Gentoo Lisp project
<URL:http://www.gentoo.org/proj/en/lisp/>, #gentoo-lisp on FreeNode

<URL:http://gentoo.faulhammer.org/>