Re: A new address of an APT source registered
* "WA" == Wichert Akkerman <wichert@cs.leidenuniv.nl> wrote:
WA> True. But you have to admit that since the original `release' of
WA> the slink GNOME packages a *lot* of bugfixes have been made in
WA> most GNOME packages.
I can't really say, as in fact I don't even maintain a single GNOME
aware package, so maybe someone else can comment on this? I just
compiled a part of the slink update. Mostly there are only "New
upstream release" changelog entries. I don't have a script to compare
versions at hand, but from a quick look on some core packages, the
slink versions are at most one or two subsubversions behind the potato
ones (gnome-libs: 1.0.9 -> 1.0.10; gnome-panel, gnome-core: 1.0.5 ->
1.0.7; gnome-control-center: same versions).
WA> And I am getting reminded too often that our GNOME packages are
WA> much more buggy then the RPM's that are being distributed, which
WA> is not something I'm happy about.
This is something I do not understand. There are no such reports on
debian-user or in this list (the instructions specifically request
problems to be reported here). And the list has been quiet for a
while. No reports from a maintainer about some buggy package from the
slink upgrade (thous requesting a fix) as well.
In fact, the responses on debian-user are more of "I installed it, no
problems, works great". If a newer version is needed to fix bugs, then
it can be compiled (how is the project to make the day 25 hours long
doing? ;-), but I am not aware of such a case.
If you have mails with specific problems, please forward them to the
list. If we don't hear about bugs, we can't fix them.
Actually, the update as I understand it was intended as a one shot
thing. Create a recent (1.0) working set of packages for slink and
distribute them. Just like a interim stable release.
If the maintainers from the respective packages give their specific OK
for a newer unstable version to be recompiled when they don't have a
slink box (by posting to debian-gtk-gnome), then someone will
recompile the package. This is just like it was handled before.
Ciao,
Martin
Reply to: