[Aptitude-devel] Simple and direct question: Should we go ahead with a new release?

[Manuel A. Fernandez Montecelo]

2012/1/31 Daniel Hartwig <mandyke at gmail.com>:
>> I don't think that a 0.7 version number can be justified just with a
>> maintainer change and a bunch of bug fixes. (But I wont' cry if you do
>> it nevertheless. :-)
>
> I agree.  A bump to 0.7 indicates changes more substantial than those
> currently made.

I don't see 0.7/0.6.5 as matter of bugfixes, but as a matter of taking
over development and a new direction of the project, and I would leave
"0.6" and below as a kind of "the legacy of Daniel Burrows".  I
wouldn't think that this is necessary if we were previously active in
the project and we merely continued it without Daniel Burrows being
much involved anymore.  But for me it's also important to remark to
users that the change in the version is also because of more than mere
bugfixes, but a change in the development project.

Also, Linux 2.6.x->3.0 and similar with Emacs and other projects is
more a matter of the leading numbers (the "major version") not meaning
much anymore.  For me, aptitude is well past the 1.0 landmark anyway.

I won't argue about this, so if others are more comfortable with
0.6.5, so be it.


>>> Also, I guess that we should target experimental first for a few weeks?
>
> Disagree with this.  Keep in mind these are primarily bug fixes and
> translation updates, hardly justifying the use of experimental.  I
> have kept an eye on the changes being committed and don't foresee any
> major issues.  An upload to unstable is fine IMO.

OK, if you prefer so.


> A couple of issues to resolve before releasing:
> [...]
> I will look at both of these in the next 24 h and we should have
> aptitude suitable for release after that.

Have you worked on this?  Can we release now?

The freeze for the upcoming release is coming (June is the initial
date proposed).  Also, and apart that I was busy with SDL and other
packages, I am hesitant to continue looking at bugs in aptitude
because I think than closing ~50 bugs at once and the size of the
changelog somehow tell me that this already too much for a release.

IMO, if we continue incorporating small fixes, we should release once
every 15 to 30 days or so, until we start to hit bugs or development
requirements needing bigger changes and more thought.

Cheers.