[Pkg-gnutls-maint] libgcrypt11

[Andreas Metzler]

On 2006-06-19 Andreas Metzler <ametzler at downhill.at.eu.org> wrote:
> On 2006-06-18 James Westby <jw+debian at jameswestby.net> wrote:
[...]
> > As for the rest of the packaging I cannot understand most off the
> > diff.gz from the last version, it seems to make lots of changes to the
> > source, many of which I cannot see the point of. For instance it
> > includes whole files that were removed upstream a couple of years ago.

> Afaik Matthias synced upstream's CVS (or SVN) repositories into
> private git repositories http://netz.smurf.noris.de/cgi/gitweb?o=age
> somehow this syncing seems to have missed most of the file deletions.
> As the packages where built from the repository also all autogenerated
> files (autof*,  *.ps, *pdf, ...) will also differ. (See also #367358).

> So for extracting the interesting stuff we will be able to
> a) probably ignore any file additions, except caused by c
> b) ignore any auto* files

> however 
> c) we will need to find out first whether the 1.2.2-1 package contains
> upstream's 1.2.2 or some later changes. There cannot be many as the
> package appeared shortly after (Tue, 18 Oct 2005) 1.2.2's release (Oct
> 5).
[...]

Afaict the whole package is basically unpatched:
:/tmp/GCR$ diff -ur libgcrypt-1.2.2-upstream/ libgcrypt11-1.2.2-1/ | \
  filterdiff -x '*/aclocal.m4' -x '*/config.guess' -x '*/config.sub' \ 
  -x '*/configure' -x '*/Makefile.in' -x '*/stamp-vti'  -x '*/version.texi' \
  -x '*/install-sh' -x '*/ltmain.sh' -x '*/missing'  -x '*/depcomp' \
  -x '*/mdate-sh' -x '*/INSTALL' -x '*/debian/*' -x '*/doc/*' \
  -x '*/m4/*' | diffstat
 Makefile.in                           |    0
 configure.ac                          |    6 +++---
 libgcrypt11-1.2.2-1/cipher/ac.c       |    1 -
 libgcrypt11-1.2.2-1/cipher/blowfish.c |    1 -
 libgcrypt11-1.2.2-1/cipher/cipher.c   |    1 -
 libgcrypt11-1.2.2-1/cipher/crc.c      |    2 +-
 libgcrypt11-1.2.2-1/cipher/md4.c      |    2 +-
 libgcrypt11-1.2.2-1/cipher/md5.c      |    2 +-
 libgcrypt11-1.2.2-1/cipher/rmd160.c   |    2 +-
 libgcrypt11-1.2.2-1/cipher/rndegd.c   |    1 -
 libgcrypt11-1.2.2-1/cipher/rndlinux.c |    1 -
 libgcrypt11-1.2.2-1/cipher/rndunix.c  |    1 -
 libgcrypt11-1.2.2-1/cipher/sha1.c     |    2 +-
 libgcrypt11-1.2.2-1/cipher/sha256.c   |    2 +-
 libgcrypt11-1.2.2-1/cipher/tiger.c    |    2 +-
 libgcrypt11-1.2.2-1/mpi/config.links  |   10 +++++++---
 libgcrypt11-1.2.2-1/src/global.c      |    1 -
 libgcrypt11-1.2.2-1/src/misc.c        |    1 -
 libgcrypt11-1.2.2-1/src/module.c      |    1 -
 libgcrypt11-1.2.2-1/src/sexp.c        |    1 -
 libgcrypt11-1.2.2-1/tests/Makefile.am |    2 +-
 21 files changed, 18 insertions(+), 24 deletions(-)

All changes to *.c look like this one:

--- libgcrypt-1.2.2-upstream/src/misc.c 2005-07-29 15:45:42.000000000 +0200
+++ libgcrypt11-1.2.2-1/src/misc.c      2006-06-25 09:57:58.000000000 +0200
@@ -23,7 +23,6 @@
 #include <stdlib.h>
 #include <string.h>
 #include <stdarg.h>
-#include <assert.h>
 #include <unistd.h>

Which is this change (and identical ones in other subdirectories) in
upstream SVN:
| Revision 1003  - (view) (download) (as text) - [select for diffs]
| Modified Mon Aug 23 13:34:51 2004 UTC (22 months ago) by mo
| File length: 16099 byte(s)
| Diff to previous 974 , to selected 1095
| 
| 2004-08-23  Moritz Schulte  <moritz at g10code.com>
| 
|         * global.c: Do not include <assert.h>.
|         * sexp.c: Likewise.
|         * module.c: Likewise.
|         * misc.c: Likewise.

As debian/changelog does not say anything about intentionally *not*
merging this upstream change I think this is a mismerge. Hmm, you have
added patches/*assert*patch which include this patch. 

libgcrypt11-1.2.2-1/mpi/config.links is #284609, I have added this one
in svn.

You have already extracted the patch to gcrypt.texi, but I think it is
wrong, because e^(-1) is rendered as

 (
e -1)

whereas the vanilla version renders correctly. (tested using texi2dvi
and xdvi)

cu andreas
-- 
The 'Galactic Cleaning' policy undertaken by Emperor Zhark is a personal
vision of the emperor's, and its inclusion in this work does not constitute
tacit approval by the author or the publisher for any such projects,
howsoever undertaken.                                (c) Jasper Ffforde