| 1 |
On Sun, May 21, 2006 at 12:10:40PM +0900, Georgi Georgiev wrote: |
| 2 |
> Just two points: |
| 3 |
> |
| 4 |
> - standards should not be set by the primary package manager |
| 5 |
> - the primary package manager does not have to be developed by Gentoo. |
| 6 |
> |
| 7 |
> More about it below: |
| 8 |
> |
| 9 |
> maillog: 20/05/2006-14:54:18(+0200): Paul de Vrieze types |
| 10 |
> > The primary package manager is the package manager that sets the |
| 11 |
> > standards for the tree. All ebuilds in the tree must function |
| 12 |
> > with the primary package manager. As the primary package manager |
| 13 |
> > sets the standard it does not have to maintain compatibility with |
| 14 |
> > other package managers. |
| 15 |
> |
| 16 |
> I personally hate the way this sounds. It implies that the package |
| 17 |
> manager comes before the standards while it should be the other way |
| 18 |
> around. Plus, it would not solve the main problem -- that there are no |
| 19 |
> set standards for the tree. You accept the GLEP like this and there will |
| 20 |
> continue to be no standards. |
| 21 |
|
| 22 |
<snipping lots of good reasons about why implementation should not |
| 23 |
define standards> |
| 24 |
|
| 25 |
So... where's the standard? :) |
| 26 |
Right, no doc yet that's official, thus at this juncture, what's |
| 27 |
there now (portage) is the effective standard. |
| 28 |
|
| 29 |
Said in the last thread, chunking out a formal EAPI=0 definition from |
| 30 |
the tree/portage implementation, tiz a good thing. But until that's |
| 31 |
done (and folks agree it's the standard), portage (primary manager) |
| 32 |
rules, thus doc it out (as I've suggested to ciaran for the |
| 33 |
slot value and use/slot dep restrictions he's added) if you're after |
| 34 |
changing the existing definition. |
| 35 |
|
| 36 |
Not saying I like it, but it's the reality of current situation- the |
| 37 |
intention of the glep is to prevent lock in, and to keep the tree |
| 38 |
unified in terms of support (avoid fracturing of the env the tree has |
| 39 |
been written against), either a doc standard is created for EAPI=0, or |
| 40 |
portage defines the standard (since it's primary). |
| 41 |
|
| 42 |
|
| 43 |
> The process should go like this: |
| 44 |
> |
| 45 |
> 1. Standars are set (by the council or whatever). |
| 46 |
> 2. They are implemented in the official package manager. |
| 47 |
> 3. Other package managers follow suit. |
| 48 |
|
| 49 |
Council really shouldn't be involved sans big changes imo, and big |
| 50 |
changes imo should require gleps (both from an archive standpoint, and |
| 51 |
from fitting the council in via existing process of gleps). |
| 52 |
|
| 53 |
One concern out of all of this is ensuring that their isn't |
| 54 |
ping/ponging back and forth as to which manager is 'official'; arms |
| 55 |
race in terms of features supported by each manager is a good thing |
| 56 |
imo, but need to keep that from causing chaos for devs in terms of |
| 57 |
changing standards. |
| 58 |
|
| 59 |
~harring |
Archives