Bug#654174: Multiple binnmus for iceweasel 9

On Mon, Jan 2, 2012 at 09:13:29 +0100, Mike Hommey wrote:

nmu gjs_1.30.0-2 . ALL . -m "Rebuild against libmozjs9d"
nmu gnome_shell_3.2.1-8 . ALL . -m "Rebuild against gjs rebuilt against libmozjs9d"

Won't work, gjs fails to build against the new mozjs.

Cheers,
Julien



To UNSUBSCRIBE, email to
with a subject of "unsubscribe". Trouble? Contact

Hi,

Mike Hommey (02/01/2012):

nmu gjs_1.30.0-2 . ALL . -m "Rebuild against libmozjs9d"
nmu gnome_shell_3.2.1-8 . ALL . -m "Rebuild against gjs rebuilt against libmozjs9d"

already done independently when src:mozjs was introduced.

nmu openvrml_0.18.8-5 . ALL . -m "Rebuild against libmozjs9d"
nmu libproxy_0.3.1-4+b1 . ALL . -m "Rebuild against libmozjs9d"
nmu gxine_0.5.906-1+b3 . ALL . -m "Rebuild against libmozjs9d"
nmu edbrowse_3.4.8-1+b1 . ALL . -m "Rebuild against libmozjs9d"
nmu dehydra_0.9.hg20110609-2+b3 . ALL . -m "Rebuild against libmozjs9d"
nmu couchdb_1.1.1-1 . ALL . -m "Rebuild against libmozjs9d"
nmu python-xpcom_1:8.0~hg20111006-1 . ALL . -m "Rebuild against libmozjs9d"

Looking at the transition page[1], concentrating on packages in testing,
for dependency level 1, a local binNMU build led to the following
results:
| couchdb build OK, deps OK.
| dehydra #655779
| edbrowse build OK, deps OK.
| firetray #653290
| gecko-mediaplayer build OK, no deps.
| gnash build OK, no deps.
| gnome-chemistry-utils #657273
| gtk-vnc build OK, no deps.
| icedtea-web build OK, no deps.
| libjdic-java #631039 or similar.
| libproxy build OK, recommends OK.
| libreoffice not tried (ETOOBIG), no deps.
| openvrml #652790 (second message)
| packagekit build OK, no deps.
| pyxpcom #654613

1. http://release.debian.org/transitio...asel9.html

Do we want to trigger binNMUs for packages in the unknown state anyway?
(Where there are no runtime build dependencies)

Mraw,
KiBi.






To UNSUBSCRIBE, email to
with a subject of "unsubscribe". Trouble? Contact

(Still learning the ropes.)

Cyril Brulebois (25/01/2012):

Looking at the transition page[1], concentrating on packages in testing,
for dependency level 1, a local binNMU build led to the following
results:
| couchdb build OK, deps OK.
| dehydra #655779
| edbrowse build OK, deps OK.
| openvrml #652790 (second message)
| pyxpcom #654613

I guess we may need to remove the 3 FTBFSing packages from testing. dak
rm -Rn -s testing says:
- happy with removing dehydra,
- happy with removing openvrml,
- not happy with removing pyxpcom:
| # Broken Depends:
| sugar-browse-activity-0.84: sugar-browse-activity-0.84
| sugar-browse-activity-0.86: sugar-browse-activity-0.86
| sugar-hulahop: python-hulahop [amd64 armel i386 ia64 kfreebsd-amd64 kfreebsd-i386 mips mipsel powerpc s390 sparc]
|
| # Broken Build-Depends:
| sugar-hulahop: python-xpcom (>= 1.9~rc2)

One would need to investigate what to do with sugar-*.

In the meanwhile, I scheduled binNMUs for couchdb and edbrowse (and
dehydra, oops).

Mraw,
KiBi.






To UNSUBSCRIBE, email to
with a subject of "unsubscribe". Trouble? Contact

Hello Mike.

Cyril Brulebois (25/01/2012):

I guess we may need to remove the 3 FTBFSing packages from testing.
dak rm -Rn -s testing says:
- happy with removing dehydra,
- happy with removing openvrml,
- not happy with removing pyxpcom:
| # Broken Depends:
| sugar-browse-activity-0.84: sugar-browse-activity-0.84
| sugar-browse-activity-0.86: sugar-browse-activity-0.86
| sugar-hulahop: python-hulahop [amd64 armel i386 ia64 kfreebsd-amd64 kfreebsd-i386 mips mipsel powerpc s390 sparc]
|
| # Broken Build-Depends:
| sugar-hulahop: python-xpcom (>= 1.9~rc2)

One would need to investigate what to do with sugar-*.

In the meanwhile, I scheduled binNMUs for couchdb and edbrowse (and
dehydra, oops).

Those are ready, like pyxpcom which you uploaded yesterday. My hints
file currently says:
| # 2012/01/26 - iceweasel 9 transition (#654174)
| remove autofill-forms/0.9.8.3-3
| age-days 5 uppity/1.5.8-3
| age-days 0 pyxpcom/1:9.0~hg20111212-1
| easy iceweasel/9.0.1-1 gxine/0.5.906-2 oolite/1.76-1 pyxpcom/1:9.0~hg20111212-1 uppity/1.5.8-3 -autofill-forms/0.9.8.3-3

dehydra and openvrml don't seem to require any special handling for the
moment, so I left them alone for now.

According to my dry-run tests, that should let iceweasel migrate during
the next run, provided nothing from others' hint files interfere with
the above one (I'm starting little, so I only ran the dry-run tests with
my tiny hints file).

Once that's done, I'll see what to do with the old binaries in testing.

Mraw,
KiBi.






To UNSUBSCRIBE, email to
with a subject of "unsubscribe". Trouble? Contact