[Mageia-dev] Release cycle - what is actually POSSIBLE?

[Romain d'Alverny]

Le 9 oct. 2010 à 09:57, Margot <margot at otfordduckscomputers.co.uk> a écrit :
> There has been a lengthy debate about users' wishes for the Mageia
> release cycle, but one important voice has been missing from this
> debate: the collective voice of the devs who will be responsible
> for producing the releases.
>
> Before we start having polls and surveys, it would be useful to
> hear from our devs.

Before devs may comment on the technical in's and out's of a proposal,
there must exist such a specific proposal, understood and agreed on by
several people. To this day, we only have discussions, so a lot of
different views and tracks coming from a common statement.

The wiki page I drafted a few days ago
(http://mageia.org/wiki/doku.php?id=rollingdebate - maybe misnamed in
this regard) serves as a draft for this documentation/proposal
purpose:
 - first gather views on perceived issues with the release/update
lifecycle of the system and apps;
 - then dive in depth into these to get common issues and understand
the real bottom line (that may not ne obvious as of today)
 - then have proposals + analysis of feasability.

This involves makers and users at all steps. That requires not
discussing specificities of solutions before having agreed on the
problem to solve.

Without that (making sure all think about the same thing in a
documented way), the discussion will stay cloudy. However no proposal
has been formally posted over there, so it may have been wrongly
understood as a method or it may just not be a fit here? I don't know.
However, if you _do_ have such a proposal to post, please take the
time to post it as an article somewhere. Or propose another structured
way.

As Oliver reminded it, the first milestone we have is releasing our
first ISO to test-drive the whole infrastructure and project
organisation.

That includes test-driving methods to state, discuss, refine
issues/solutions or new ideas and then decide to put them into
production or not (what we could do right here, not relying only on ml
conversations but on stable docs too).

Cheers,

Romain