On Wed, Jan 02, 2002 at 12:04:01AM +0100, Martin Schulze wrote:

> samba-common  stable    2.0.7-3.4   alpha, arm, i386, m68k, powerpc, sparc
> samba-common  updates   2.0.7-4     alpha, arm, i386, powerpc
> samba         stable    2.0.7-3.4   alpha, arm, i386, m68k, powerpc, sparc
> samba         updates   2.0.7-4     alpha, arm, i386, powerpc
> smbclient     stable    2.0.7-3.4   alpha, arm, i386, m68k, powerpc, sparc
> smbclient     updates   2.0.7-4     alpha, arm, i386, powerpc
> smbfs         stable    2.0.7-3.4   alpha, arm, i386, m68k, powerpc, sparc
> smbfs         updates   2.0.7-4     alpha, arm, i386, powerpc
> swat          stable    2.0.7-3.4   alpha, arm, i386, m68k, powerpc, sparc
> swat          updates   2.0.7-4     alpha, arm, i386, powerpc

>       ChangeLog says:

>       * Permanently fix problem with NMU's being built against incorrect
>         kernel
>         interfaces (closes: #94380, #95015, #102226)

>       * add uploaders: header to control file

>       This upload most probably fixes the problem with the old alpha
>       version not being able to run properly due to a bad build
>       environment.  This problem may be solved by a general
>       change...  may be...  Steve Langasek should speak up...

Samba upstream takes advantage of the best system facilities 
(libc/kernel) available at compile time.  Because Debian releases 
usually include a baseline kernel and an 'experimental' kernel, Eloy and 
I have introduced packaging code in unstable that prevents Samba from 
detecting facilities that it should not be compiled against.  The
2.0.7-4 upload backports these packaging mods to potato, both correcting
the problems with past alpha security NMUs and safeguarding against
the possibility of future problems with security NMUs in potato.

>       MISSING m68k
>       MISSING sparc

I have not heard from either of these porting teams yet, I imagine 
because of the holidays.  I can build a sparc binary on kubrick if 
necessary, but I'm not aware of any m68k machines I can get access to 
without again going through the porting team.  If anyone is in a 
position to do the binary NMU for m68k in time for 2.2r5, I'd certainly
appreciate it.

Steve Langasek
postmodern programmer

Attachment: pgpmXzySiIZ8T.pgp
Description: PGP signature

Reply via email to