Ubuntu Docteam problems post string freeze

[Claude Paroz]

Le samedi 08 octobre 2005 à 09:25 +0100, Matthew East a écrit :
> On Sat, 2005-10-08 at 02:49 -0300, Matthew Thomas wrote:
> > On 8 Oct, 2005, at 12:01 AM, Robert Stoffers wrote:
> > > ...
> > > There have been several bugs filed lately, some duplicates, for several
> > > known issues with our documentation. Mostly it has been a problem that
> > > is caused by the fact we are writing about software that is still 
> > > having changes made to it.
> > >
> > > One other noticeable problem was with the name of the next release of 
> > > Ubuntu, Dapper Duck, in that we didn't know what it was going to be 
> > > called until after string freeze.
> > 
> > Nor should you have to! The idea that the help for an operating system 
> > should need to mention the *codename* for the *next release* is truly 
> > bizarre. Please mention this in the Rationale section of the BoF notes 
> > <https://wiki.ubuntu.com/AutomaticUpgrade>.
> > 
> > > ...
> > > My solution to this problem is simple: ubuntu-docs shouldn't be 
> > > included in the initial release of each new version of Ubuntu, but 
> > > instead pushed out as an update after release.
> > 
> > I can think of four other options:
> > 1.  Make the feature freeze earlier and/or harder.
> > 2.  Make the documentation freeze later and/or softer.
> > 3.  Increase the number of writers working in the time between those
> >      two freezes.
> > 4.  Reduce the scope of the documentation.
> 
> All of these options are better than releasing ubuntu-docs after release
> of the operating system in my opinion.
> 
> I have another one:
> 
> 5. enforce the freezes.
> 
> In terms of (2), it might be a good idea to have the doc string freeze
> later, and to allow _translations_ to come into the operating system
> after release. This would be a good balance to strike in my opinion. We
> just need to make sure that the lessons of this freeze are learned and
> put foward at UBZ... 

I'm definitely in favour of (2), with a translator point of view.
While I understand the importance of freeze concept for code, I don't
see the point in freezing documentation at all, except maybe to keep
updates as small as possible.
IMHO correction and translation of documentation should not be subject
to freeze time (as it's the case for po translation for app strings in
language packs).

Claude