Re: custom vs. derivative
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Thu, Apr 10, 2008 at 02:00:48PM +0200, Andreas Tille wrote:
> On Thu, 10 Apr 2008, Jonas Smedegaard wrote:
>> you widened [the CDDNamingProposals wiki page] from naming CDD to
>> naming other things too, lengthening the discussion and blurring the
>> topic IMO.
>
> Well, I learned that the whole thing can finally only explained to
> users in comparison to other things.
True. And I want CDD defines in comparison to Debian. New ideas
emerge, and some of them die. The CDD concept is tied to Debian. CDD
might very well get _influenced_ by other ideas, and we can evolve. But
we do not need to define other things than our offspring IMO.
What we experienced in discussions last week in Extremadura was not
(after clarifying, and other viewpoints was still kept) inability to
understand, but different *interests* of what CDD *ought* to be.
This spawned several discussions (intermixed with each other at times):
1) rename CDD to stand apart from non-Debian customizations
2) "correct" CDD definition to include more kinds of distros
3) "correct" CDD definition to include only the originally intended
4) name CDD-related stuff
5) replace CDD definition with definition(s) based on 4)
My interest is 1) and 3), and the intend of my latest posts has been 3)
to support 1). You seem to want it as part of 4) which is fine by me,
but you seem alone in that quest.
> So I agree with you that our main focus should be on "thingy 1" but we
> will loose many friends if we call them "all the (indifferent) rest".
> IMHO this builds the bridge between mathematical preciseness and real
> life.
We also needs a kernel in our CDDs, so should make friends with the
kernel team... And what about programming languages... And packaging
styles...?
I honestly believe that we do not loose or upset anyone by clearly
defining who we are ourselves. Except perhaps those that wrongly
thought that they were part of us. But really you do not avoid that by
blurring the discussion with talking about all species of Debian-related
distros: In the end they will still be hurt when they realize they do
not fit our definition of ourselves.
You _do_ avoid unfitting distros getting hurt by redefining yourself to
match all those your want to make friends. Is that your current intend?
> Well, in the discussion at OSWC in Malaga 2004 with Bdale Garbee,
> Colin Watson, Martin Michlmayer and others we exactly thought about
> the idea to drop the concept of doing full releases and instead
> releasing single CDD-like pieces of Debian.
I like that idea. But a (related but) *different* one than CDD.
It is a bolder one, because it involves changes to fundamental Debian
structures. CDD does not require interaction with ftpmaster ;-)
The CDD can help push that bolder idea: CDDs help ensure that multiple
subdistributions maintained in parallel from same core distributions
works *without* changing infrastructure. Next step is then to convince
the d-i team to roll out testing releases tied to CDDs. Then it is
*much* easier to discuss if we want to tag some of those
subdistributions as "stable" with different release cycles.
>> Here is my view on the above, illustrated like the referenced text:
>>
>>
>> unstable -> testing -> stable -> official stable release
>> | | |
>> | | +-> CDD_A stable -> CDD_A stable release
>> | | |
>> | | +-> CDD_B stable -> CDD_B stable release
>> | | |
>> | | +-> ...
>> | |
>> | +-> testing snapshot release
>> | |
>> | +-> CDD_A testing -> testing CDD_A snapshot release
>> | |
>> | +-> CDD_B testing -> testing CDD_B snapshot release
>> | |
>> | +-> ...
>> |
>> + -> CDD_A unstable -> unstable CDD_A snapshot release
>> |
>> + -> CDD_B unstable -> unstable CDD_B snapshot release
>> |
>> +-> ...
>
> Well, this is our old fashioned view as it currently is.
No - this is *not* a common view. Maybe common to you and me, but we
also tend to agree with the CDD definition as is.
Debian Edu see themselves as a CDD. They promote their current release
as stable.
> So please do not build an artificial wall, when we might need a
> doorway (perhaps with a door that never opens - but it should be there
> just in case).
If you deliberately want a vague definition then why discussing at all?
Debian Edu is happy that they are considered a CDD now.
Debian gets bad reputation aming users if a CDD promoted as stable is of
lower quality than Debian stable.
The CDD concept gets bad reputation among Debian developers, if
non-technical users file bugreports about packages only officially
available in non-stable distributions, but promoted as stable in a CDD.
>> The drawing intends to reflect current workflow of Debian itself:
>
> Yep - this is the problem of the drawing. It contains nothing
> which is _new_ and thus belongs to a different place inside the
> docs.
I am only happy if my drawing - and with it most of my latest posting,
has been a complete waste of time ;-)
>> separates "stable" as distribution and as release, and adds testing
>> release. CDDs pays full respect to Debian work by now distorting it:
>
> And while they do so we perhaps waste a big chance to enrole the
> real power of the CDD concept.
IMO power is bad if it distorts Debian work.
Please elaborate.
- Jonas
- --
* Jonas Smedegaard - idealist og Internet-arkitekt
* Tlf.: +45 40843136 Website: http://dr.jones.dk/
- Enden er nær: http://www.shibumi.org/eoti.htm
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
iD8DBQFH/isvn7DbMsAkQLgRAm11AJwKpeUJZ0I+qHPfrgPSzbAjibiUHQCfc2hf
hTA8aaz658GgB5dQpCTm2wo=
=P/As
-----END PGP SIGNATURE-----
Reply to: