Bug#595094: linux-base: /etc/initramfs-tools/conf.d/resume^old causes trouble

Package: linux-base
Version: 2.6.33-1~experimental.4
Severity: normal

Hello

Since upgrading to 2.6.32-5-amd64 my system does not boot every time. It then
hangs after identifying all SATA discs in script called "local-premount" with a
message saying that it cannot find the resume drive /dev/sda2 and I should
press enter.
As my USB keyboard is not working at this point I usually do a hard reset and
then most of the time the boot continues.

I was wondering where the /dev/sda2 was coming from as my GRUB and fstab
contain UUID resp. /dev/disk/by-name labels. The only occurence which grep
found was a /etc/initramfs-tools/conf.d/resume^old with a timestamp from
2008-01-14 next to a file "resume" which contains the correct UUID.

As I would never use a ^ character in a filename I looked in /var/lib/dpkg/info
and found linux-base.postinst which does so an has an interesting comment next
to it:

                    {packages => 'initramfs-tools',
                     path => '/etc/initramfs-tools/conf.d/resume',
                     list => \&initramfs_resume_list,
                     update => \&initramfs_resume_update,
                     # udev will source all files in this directory,
                     # with few exceptions.  Such as including a '^'.
                     suffix => '^old'},

Bug#595095: Preseed options for hands-off install

Package: grub2
Version: 1.98+20100804-4

I am attempting to migrate an automated preseed installation from Lenny
to Squeeze. In Lenny, when installing with only one internal hard drive,
grub picked it without requiring the device to be explicitly specified.
If it happened that there were more than one physical device present,
the installer would prompt for the correct drive. When upgrading to
grub2 during the late-command script, grub2 just used that previously
selected drive when running the postinstall / legacy upgrade script.

Now in Squeeze, it seems impossible to perform a preseeded installation
without putting a specific device (/dev/sda) in the preseed file. I
preferred the way the Lenny grub installer handled it. Sometimes if
another device is present there is a race at boot and an sd card or usb
key will sometimes be selected by the kernel to be the /dev/sda device
node. This means that the device might be inadvertently overwritten.

Is it possible to make a completely hands-off preseed installer without
putting a specific device node in the preseed file using grub2?

Does this make sense? Let me know if I need to clarify what I'm asking.

Tony

Bug#594944: auctex: preview no longer works with 'geometry' package

tag 594944 + unreproducible
thanks

Hi,

thanks for your report.  Unfortunately I cannot reproduce this bug: your
sample document works fine here, with or without geometry.

I will try to delve into preview-latex log; meanwhile, it would be
helpful if you could retry your sample document on some other machine
and let me know if you can reproduce the bug.

--

-- 
Servus, Davide

Bug#595097: debhelper: dh_fixperms doesn't on image files, .css nor .js files

Package: debhelper
Version: 8.0.0
Severity: wishlist

dh_fixperms doesn't set sane permissions for image files (.png, .gif,
..jpeg, to name a few), cascading stylesheet files (.css) nor
JavaScript fragments (.js).

This could be useful while packaging web applications that provide this
kind of collateral material. I found myself having to write explicit
find statements in debian/rules to overcome this shortcoming of
dh_fixperms while packaging webgui.

A simple solution could be

    # .. fix JavaScript fragments
	complex_doit("find $tmp -type f",
		"-perm -5 -name '*.js'  $find_options -print0",
		"2>/dev/null | xargs -0r chmod 644");

    # .. fix CSS files
	complex_doit("find $tmp -type f",
		"-perm -5 -name '*.css'  $find_options -print0",
		"2>/dev/null | xargs -0r chmod 644");

    # .. fix common webapp image file formats
	complex_doit("find $tmp -type f",
		"-perm -5 \( -name '*.png' -o -name '*.jpg' -o -name '*.gif' \) $find_options -print0",
		"2>/dev/null | xargs -0r chmod 644");

Bug#588552: [Evolution] Bug#588552: Bug#588552: evolution-alarm-notify does not start automaticly anymore

OK, so evolution-alarm-notify can no longer be launched from evolution
itself...  Sad, 

Any ways, good to know.  Do you think it's best to start
evolution-alarm-notify before evolution, or after?

Thanks,

--

-- 
Javier.

Bug#594872: espeak: hungarian language phoneme is full broken and unusable

Hello,

Could you check that the attached patch fixes it?

Samuel

diff --git a/platforms/big_endian/espeak-phoneme-data.c b/platforms/big_endian/espeak-phoneme-data.c
index 2a8dd22..815fdd4 100644
--- a/platforms/big_endian/espeak-phoneme-data.c
+++ b/platforms/big_endian/espeak-phoneme-data.c
 <at>  <at>  -286,6 +286,24  <at>  <at>  void swap_phondata  (const char *infile, const char *outfile,
             xread = fread (env_buf, 1, 128, in);
             fwrite (env_buf, 1, 128, out);
         }
+        else if (line[0] == 'Q') {
+            unsigned char pb[4];
+            unsigned length;
+            char *buf;
+
+            xread = fread (pb, 1, 4, in);
+            fwrite (pb, 1, 4, out);
+
+            length = (pb[2] << 8) + pb[3]; // size in words
+            length *= 4;
+
+            buf = (char *) malloc (length);
+
+            xread = fread (buf, length, 1, in);
+            fwrite (buf, length, 1, out);

Bug#595082: slow at processing entries with large number of authors

tags 595082 - moreinfo + confirmed
retitle 595082 pybtex: slow at processing entries with large number of authors
thanks

* Jameson Rollins <jrollins <at> finestructure.net>, 2010-08-31, 18:53:
>> FWIW, it doesn't hang here:
>>
>> $ grep bogo /proc/cpuinfo
>> bogomips        : 4620.37
>> bogomips        : 4620.37
>>
>> $ time pybtex foo
>>
>> real    1m16.624s
>> user    1m16.205s
>> sys     0m0.212s
>>
>> Maybe you need to be a bit more patient? :)
>
>Hey, Jakub.  I guess you're right.  I am *not* particularly patient, but
>I think this would probably stretch the patients of anyone:
>
>0 $ time pybtex foo
>
>real	4m36.079s
>user	2m54.335s
>sys	0m5.476s
>0 $
>
>I guess my name for this issue isn't exactly accurate, but I would still

Bug#594607: libdbi upload to SID reverted (was: Freeze exception for libdbi and libdbi-drivers)

Clint Byrum writes:
 > I believe all that is needed is to bump 'LIB_CURRENT' in configure.in from 0 -> 1, and autoreconf.

Hi,

I hope it's not too late, but I've prepared a 0.8.4 test tarball:

http://libdbi.sourceforge.net/downloads/libdbi-0.8.4.tar.gz

This isn't officially released yet, but I could do so in no
time. Please have a look at the tarball and let me know if this
serves your immediate needs.

I've used the 0.8.3 sources and changed the following things:

- bumped the package version number to 0.8.4

- updated some autotools-related things in configure.in, autogen.sh,
  and Makefile.am (recent versions need AC_CONFIG_MACRO_DIR set
  properly, otherwise I wouldn't be able to bootstrap the build on my
  FreeBSD development box)

- LIB_CURRENT=1, LIB_REVISION=0, LIB_AGE=0

- fixed the --disable-docs behaviour of configure (this had been fixed
  long ago in HEAD)

regards,
Markus

Bug#594274: debian-policy: Don't track generated README documents in VCS

Bill Allombert <Bill.Allombert <at> math.u-bordeaux1.fr> writes:

> On Fri, Aug 27, 2010 at 03:50:02PM -0700, Russ Allbery wrote:
> > Bill, do you mind if we make Emacs a build requirement for Policy
> > for the *.org files that are in there right now? I think you were
> > the one who expressed concerns previously about the org files.
>
> emacs and org-mode I think.
> I do not care as long as all needed packages are listed in the
> Build-Depends field, […]

The versions of Emacs in Debian include Org mode. AFAIK, the ‘org-mode’
package is merely a more current version of Org mode than that included
with Emacs.

If that's true, then a dependency on the ‘emacs’ package alone should be
sufficient for rendering Org source files.

--

-- 
 \        “I don't accept the currently fashionable assertion that any |
  `\       view is automatically as worthy of respect as any equal and |
_o__)                                   opposite view.” —Douglas Adams |
Ben Finney <ben <at> benfinney.id.au>

Bug#595032: "icon" value ignored in "startmenu" entries

  Hello Bill,

Determining parent section name by its title (not by package name, as in 
my original menu file) with the rest of full section path - thank you 
for this hint, it was my mistake, not bug. Now update-menus works for 
IceWM as I expect.

My purpose is to reconstruct manually written IceWM toolbar into menu 
files and then generate from them customized menus for another window 
managers like XFWM.

By the way: in --menufilesdir option directory name required trailing slash.

Regards
Tomasz

On 31.08.2010 18:52, Bill Allombert wrote:
> On Tue, Aug 31, 2010 at 05:10:19PM +0200, Tomasz [arctgx] Nowiński wrote:
>>
>> ?package(local.image): \
>> 	needs="X11"\
>> 	title="Image"\
>> 	section="Applications"\
>> 	command="" \
>> 	icon=geeqie.xpm
>> ?package(local.geeqie):\
>> 	needs="X11"\
>> 	title="Geeqie"\
>> 	section="image"\
>> 	icon="geeqie.xpm"\