Re: [Pytables-users] CVS branching incorrect?

[Francesc A. <fa...@ca...>]

Hi *Norbert* ;)

A Diumenge 05 Desembre 2004 13:59, Norbert Nemec va escriure:
> I just realized that the branching in the CVS seems to be a bit mixed up:=
 When=20
> I pull pytables from CVS without any sophisticated branching options, I g=
et=20
> an older version than the 0.9.1 which I can download as .tar.gz
>=20
> For example, the last patch of AttributeSet.py (changing the behavior for=
=20
> non-existant attributes) is in the .tar.gz but not in the MAIN branch of =
the=20
> CVS.
>=20
> In general, I do not really see, by the pytables CVS needs any branches a=
t=20
> all. Branches are a nice thing if you want to fork of a experimental vers=
ion=20
> that should be handled separated from the regular maintenance.

I've decided to create a branch for 0.9 updates called Release-0_9_patches,
and continue the development in MAIN for 1.0. The normal thing is to do some
merges from branch Release-0_9_patches to MAIN with some frequence.

> For the rather regular and centralized development of pytables one workin=
g=20
> trunk containing all the patches should IMO be enough and help avoid some=
=20
> confusion.

Well, I've just followed the recommendations in:

https://www.cvshome.org/docs/manual/cvs-1.11.16/cvs_5.html

which clearly advocates for the solution that I've recently adopted and I
must say that I find out this approach to be very useful.

Cheers,

=2D-=20
=46rancesc Altet
Who's your data daddy? =A0PyTables