[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: gnuab.org unreleased analysis

Thank for review, it's been on my TODO for a while..

On Mon, Sep 10, 2007 at 06:32:37PM -0400, Joey Hess wrote:
> The following packages in unreleased have older versions than unstable
> and the new versions don't have the armel patch applied. I filed bugs
> on several of these since the patch wasn't in the BTS. We probably should
> update the versions in unreleased in the meantime:

> gdb 6.6.dfsg-1+armel


needs gobjc + gcj-4.2 build dependencies. I'm currently bootstrapping
gcj so that leaves just removing gobjc build dependency on armel.
Then we need to poke drow.

> gettext 0.16.1-1+armel
> libtool 1.5.22-4+armel

This just need gcj + reschedule of build.

> nspr 2.6.18-7+armel

#436213. Mike was happy anough to add my patch to webkit while
the nspr patch seems to have falled to void.. Let's see if poking helps.

> python-numeric 24.2-7+armel
> 	Changelog for the +armel version says "Disable lapack and blas
> 	support.", but it was a binary-only upload so I don't know what
> 	changes were made to the source and couldn't produce a patch.
> 	This needs to be dealt with.

Long term solution is the g77->gfortran transition[0]. As part of that I
need to find out why refblas3 doesn't build on selected architectures[1]
and then upload lapack3 to experimental. Then it would be more natural
to upload experimental versions of refblas3, lapack3 and python-numeric
to unreleased.. Now I just need to find time to execute the plan.

> gcc-defaults 1.56+armel
> 	Newer version in unstable has the armel patches.
> 	But new version hasn't built on armel yet for some reason.

It needs gcj-4.2 as well[2]:

For building states, you can use buildd.net[3] to find
where the package is. The Failed/Dep-Wait/Building/Not-For-Us
lists need walking through as well.

> linux-kernel-headers 2.6.18-7+armel
> 	Removed from unstable.

Needs manual gnuab admin action to remove afaik.

> freedts #441736
> 	The patch in this one could use a better explanation than I
> 	managed.

I'll comment.


[0] http://lists.debian.org/debian-toolchain/2007/07/msg00000.html
[1] http://lists.debian.org/debian-toolchain/2007/08/msg00012.html (amd64 and armel build failures are identical)
[2] http://experimental.debian.net/fetch.php?&pkg=gcc-defaults&ver=1.61&arch=armel&stamp=1188871681&file=log&as=raw
[3] http://unstable.buildd.net/index-armel.html

"rm -rf" only sounds scary if you don't have backups

Attachment: signature.asc
Description: Digital signature

Reply to: