Le 18 févr. 2006 à 20:25, David R. Morrison a écrit :

On Jan 20, 2006, at 9:44 AM, David R. Morrison wrote:

I am imposing a "chill" on CVS, for the 10.4-transitional tree (both stable and unstable). During the "chill", I ask that people only make urgent commits, and that they try to avoid changing packages which have a GCC tag. (Of course, bug fixes are always welcome, particularly if they involve correcting the GCC tag.)

This "chill" will be followed by a proper CVS freeze in a few days, to enable construction of the 10.4 tree.
Two questions:

1 - Should the specific version number on dependencies be updated systematically

Example to make clear what I mean:

libgnome2 mentions orbit2 (>= 2.10.0-1) as dependency, but orbit2 in 10.4 stable tree is at 2.10.3-1008

Is this ok, or should the number in libgnome2 be modified so that it ensures that the old deb from 10-4 transitional be not used? Sorry if my question is silly, but I need to have a clear idea to what should be done exactly.

2- There is no gnome-vfs2-ssl package on 10.4 stable tree, as in 10.4- transitional stable tree; Nevertheless some packages mention it as an or'ed dependency. Is it ok, or would it be safer to drop the gnome- vfs2-ssl part?

Some conflicts today:

Trying to update-all today - that is February 20th 2006 - 6:35: 44 UTC time (I already installed glade2 and bluefish yesterday), there is a conflict between libgnome2-dev and gnome-libs-dev:

It seems that gdk-pixbuf uses gnome-libs-dev, gnome-desktop uses libgnome2-dev and gnome-libs-dev, but gnome-libs-dev is marked as conflicting with libgnome2-dev. Something probably is not perfect here.

Packages directly affected (and probably a good mesure of other ones in gnome):
gdk-pixbuf  glade2 gnome-desktop gnome-panel

(Note that I've begun from an empty 10.4 tree and bootstrap from head, so nothing was installed previously and I just installed glade2 thereafter).

Cheers,
Michèle
<http://micmacfr.homeunix.org>

Attachment: 36C471DED4B09EEB30A0281F2608DB2FE6F9E147.gpgkey
Description: Binary data



Reply via email to