Bug#346072: slocate writes to zero-length array -> segfault
Package: slocate
Version: 3.0.beta.r1-1
An array is allocated on the stack with zero length and then written to
which corrupts the stack. Shows up as a segfault on i386 and an Ubuntu
setup:
http://bugzilla.ubuntu.com/show_bug.cgi?id=21525
Maybe related to #345171 seen on PowerPC.
Thanks to Chris Moore <[EMAIL PROTECTED]> for tracking it down. Patch
attached that corrects the length.
-Paul
--
This country is covered in white fluffy snow. Helsinki, FI
--- slocate-3.0.beta.r1.orig/debian/changelog
+++ slocate-3.0.beta.r1/debian/changelog
@@ -1,3 +1,9 @@
+slocate (3.0.beta.r1-1ubuntu1) dapper; urgency=low
+
+ * Closes: Ubuntu#21525 . Population of zero-length array.
+
+ -- Paul Sladen <[EMAIL PROTECTED]> Thu, 5 Jan 2006 02:35:43 +
+
slocate (3.0.beta.r1-1) unstable; urgency=low
* New release. 3.0 beta r1. This is a complete redesign/rewrite.
--- slocate-3.0.beta.r1.orig/src/slocate.c
+++ slocate-3.0.beta.r1/src/slocate.c
@@ -483,7 +483,7 @@
int search_db(struct g_data_s *g_data, char *database, char *search_str)
{
int fd = -1;
- char ch[0];
+ char ch[1];
int buf_len;
char buffer[BLOCK_SIZE];
int ret = 0;
Bug#389511: Please send patch upstream
It would be useful if this fix could be filed and sent upstream: https://launchpad.net/distros/ubuntu/+source/acpi-support/+filebug -Paul -- High on a tall bridge, surrounded by noisy lorries. London, GB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#341822: (no subject)
Confirmed, this is: https://launchpad.net/bugs/40956 and going to occur whenever the package is built on a buildd that is primarily controlled by a serial console (eg. most of them! :-). My proposal is to build with '--with-serial' with as standard to solve the problem. This bug *severely* screws up people's machines, as keystrokes get lost or corrupted and disappear from /dev/tty0. John Lines: are you happy if I NMU up a patch via DELAYED? This bug has been open for 9 months so far. -Paul -- High on a tall bridge, surrounded by noisy lorries. Southampton, GB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#341822: Patch using --with-serial
Tag: patch Adjusted version of patch uploaded to Ubuntu, please sign and sponsor this upload if you're happy with it. -Paul -- High on a tall bridge, surrounded by noisy lorries. Southampton, GB diff -u plptools-0.14/debian/rules plptools-0.14/debian/rules --- plptools-0.14/debian/rules +++ plptools-0.14/debian/rules @@ -32,6 +32,7 @@ ./configure $(confflags) --prefix=/usr \ --with-mountdir=/var/lib/plptools/mnt \ --mandir=/usr/share/man --with-initdir=none --enable-kde \ + --with-serial=/dev/ttyS0 \ --disable-rpath touch configure-stamp diff -u plptools-0.14/debian/changelog plptools-0.14/debian/changelog --- plptools-0.14/debian/changelog +++ plptools-0.14/debian/changelog @@ -1,0 +1,8 @@ +plptools (0.14-5) unstable; urgency=low + + * Build with '--with-serial=/dev/ttyS0' so that 'ncpd' doesn't default +to opening '/dev/tty0' (the primary console) and corrupting keyboard + input. (Closes: #341822) + + -- Paul Sladen <[EMAIL PROTECTED]> Tue, 15 Aug 2006 11:35:49 +0100 + -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQBE4d/Hc444tukM+iQRAmE6AKCKG+YGP2NxIyxHTg3/0BHubVRLlwCfRB3v V9utjotyeHdrXemeJhFq6iE= =r+jS -END PGP SIGNATURE-
Bug#605102: devscripts pulls in Postfix by default: Should Suggest instead of Recommend bsd-mailx | mailutils
Source: devscripts Version: 2.10.69 Forwarded from: https://bugs.launchpad.net/ubuntu/+source/devscripts/+bug/485264 The inclusion of 'bsd-mailx | mailutils' in Recommends combined with Recommends-by-default causes a full-blown MTA (Postfix/Exim) install on laptops/desktop that do not already have or need one. Historically, an MTA was previously only suggested (see following commit and bug). The included utilities all fail gracefully in the case of there not being an MTA. Previous commit (wholesale Suggests->Recommends migration): - +devscripts (2.10.23) unstable; urgency=low + + * Move the current Suggests: to Recommends: so that they are pulled in by +default but may be removed if desired (Closes: #474559) + * bug/presubj: Update in line with the change to Recommends: + + -- Adam D. Barratt Sun, 06 Apr 2008 16:53:46 +0100 - Bug (noting that some items have already been moved back to Suggests): - http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=474559 - Original LP bug report: - Because Recommended packages get installed by default, installing devscripts on a fresh install pulls in a complete MTA as an indirect result of Recommending these mailers. Setting up an MTA is not easy to automate, so users will either be pestered with Debconf questions and may need to edit configuration files and do testing to get the MTA running succesfully and securely, or they will have to manually uninstall it after installing devscripts, or they will need to install it with apt-get --no-install-recommends and cherry-pick all the other packages they need. This is IMHO overkill for most users and defeats the purpose of Recommending certain packages. All of the utilities which might need access to a mailer are described as such in the package description, and all will fail gracefully (as far as I can tell) with a descriptive message if one is not available. Compare with the reportbug package, whose sole purpose is reporting bugs by email, which merely Suggests a mailer. Kind regards, Alain - Patch against git head to follow as soon as debbugs sends me this bug number for the changelog. -Paul -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#605102: Patch
[ Paul Sladen ] * debian/control: move "bsd-mailx | mailx" from Recommends: -> Suggests: to avoid installing full-blown MTA (Postfix/Exim) on desktop machines that do not already have one. (Closes: #605102) (LP: #485264) * debian/control: git-core -> git post package-name migration I would support/encourage a larger review of Suggests/Recommends for devscripts. Many items really don't need to be there and some choices (preferring 'subversion' first) are possibly better adjusted to reflect the times (if they /have/ to be there at all...). If anyone would prefer to pull these changes via git, please let me know; ditto for signing+sponsoring the changes directly (left as UNRELEASED). -Paul From 192a533e740cccdb3c122e3c52f9656a5f53b513 Mon Sep 17 00:00:00 2001 From: Paul Sladen Date: Sat, 27 Nov 2010 13:02:41 + Subject: [PATCH 1/2] debian/control: move "bsd-mailx | mailx" from Recommends: -> Suggests: to avoid installing full-blown MTA (Postfix/Exim) on desktop machines that do not already have one. (Closes: #605102) (LP: #485264) --- debian/changelog |9 + debian/control |4 ++-- 2 files changed, 11 insertions(+), 2 deletions(-) diff --git a/debian/changelog b/debian/changelog index 4348154..215b38b 100644 --- a/debian/changelog +++ b/debian/changelog @@ -1,3 +1,12 @@ +devscripts (2.10.70) UNRELEASED; urgency=low + + [ Paul Sladen ] + * debian/control: move "bsd-mailx | mailx" from Recommends: -> Suggests: +to avoid installing full-blown MTA (Postfix/Exim) on desktop machines +that do not already have one. (Closes: #605102) (LP: #485264) + + -- Paul Sladen Sat, 27 Nov 2010 12:02:00 +0100 + devscripts (2.10.69) unstable; urgency=low [ James Vega ] diff --git a/debian/control b/debian/control index e2466e8..7f55ff4 100644 --- a/debian/control +++ b/debian/control @@ -17,13 +17,13 @@ Recommends: at, curl | wget, dctrl-tools, debian-keyring, debian-maintainers, dput | dupload, equivs, fakeroot, gnupg, libauthen-sasl-perl, libcrypt-ssleay-perl, libparse-debcontrol-perl, libsoap-lite-perl, libterm-size-perl, libtimedate-perl, liburi-perl, libwww-perl, - libyaml-syck-perl, lintian, lsb-release, bsd-mailx | mailx, + libyaml-syck-perl, lintian, lsb-release, man-db, patch, patchutils, ssh-client, strace, unzip, wdiff, www-browser, subversion | cvs | darcs | tla | bzr | git-core | mercurial, lzma, xz-utils, sensible-utils, libjson-perl Suggests: build-essential, cvs-buildpackage, devscripts-el, gnuplot, libfile-desktopentry-perl, libnet-smtp-ssl-perl (>= 1.01-2), mutt, - svn-buildpackage, w3m + svn-buildpackage, w3m, bsd-mailx | mailx Description: scripts to make the life of a Debian Package maintainer easier Contains the following scripts, dependencies/recommendations shown in brackets afterwards: -- 1.7.2.3 From c632d1dd724e664b394a870a048fa15069bbe8fd Mon Sep 17 00:00:00 2001 From: Paul Sladen Date: Sat, 27 Nov 2010 13:05:00 + Subject: [PATCH 2/2] debian/control: git-core -> git post package-name migration --- debian/changelog |1 + debian/control |2 +- 2 files changed, 2 insertions(+), 1 deletions(-) diff --git a/debian/changelog b/debian/changelog index 215b38b..07332a9 100644 --- a/debian/changelog +++ b/debian/changelog @@ -4,6 +4,7 @@ devscripts (2.10.70) UNRELEASED; urgency=low * debian/control: move "bsd-mailx | mailx" from Recommends: -> Suggests: to avoid installing full-blown MTA (Postfix/Exim) on desktop machines that do not already have one. (Closes: #605102) (LP: #485264) + * debian/control: git-core -> git post package-name migration -- Paul Sladen Sat, 27 Nov 2010 12:02:00 +0100 diff --git a/debian/control b/debian/control index 7f55ff4..f139764 100644 --- a/debian/control +++ b/debian/control @@ -19,7 +19,7 @@ Recommends: at, curl | wget, dctrl-tools, debian-keyring, debian-maintainers, libterm-size-perl, libtimedate-perl, liburi-perl, libwww-perl, libyaml-syck-perl, lintian, lsb-release, man-db, patch, patchutils, ssh-client, strace, unzip, wdiff, www-browser, - subversion | cvs | darcs | tla | bzr | git-core | mercurial, lzma, + subversion | cvs | darcs | tla | bzr | git | mercurial, lzma, xz-utils, sensible-utils, libjson-perl Suggests: build-essential, cvs-buildpackage, devscripts-el, gnuplot, libfile-desktopentry-perl, libnet-smtp-ssl-perl (>= 1.01-2), mutt, -- 1.7.2.3
Bug#603157: ITP: ttf-ubuntu-font-family
On Wed, 23 Nov 2011, [ISO-8859-1] G�rkan Seng�n wrote: > would be interested to see it in main, not just in "contrib". Everyone would like to see the Ubuntu Font Family in 'main'. At this point in time, 'main' is sadly not an option because: (a) UFF comes with complete source [not just .ttfs], and (b) UFF requires a build-system that is not wholly in 'main'. > then it's contrib only... Uploading to 'contrib' would be the logical destination. I don't think there's anything blocking that? I'm happy to /help/ via pkg-fonts—but because I've been working closely on the Ubuntu Font Family and I am presented employed by Canonical I don't want to be seen to be actively forcing anything. From a Debian-perspective I absolutely want to see it in Debian and moving up the repos whenever possible. > to make the font really free? The full source code for the typeface is there, and everyone is free to modify, share, remix and improve it without any requirement to ask permission first or the possibility to be denied. (It is perfectly okay for the FTP masters to have found this to be non-DFSG free, that is the FTP masters' viewpoint). > what do you think are the chances for Ubuntu/Canonical At the point that a better libre font licence comes along, the Ubuntu Font Family can be relicensed. All of the necessary legal permissions to quickly perform such a relicensing are (intentionally) in-place: http://font.ubuntu.com/ufl/FAQ.html I think it's been consistently stated, that the Ubuntu Font Licence-1.0 is being used as an *interim licence* solution. There are likely to be further revisions of the UFL, and also to other libre font licences, as the community and understanding of open font production expands. > They (Canonical, Ubuntu) after all mention the differences between > free software and other software. Font Software licensing has some unique requirements not found in other general-purpose software-, or content-licences (where GPL and CC-BY-SA are already clearly the preferred). There is a necessarity to allow embeding of the font in documents, without causing those documents to come under the same licensing conditions as the font. A font that cannot be /used/ by target users is not much use. Various of the available font licences try to cover the font-specific criteria in several ways: http://font.ubuntu.com/ufl/FAQ.html#embedding For instance, the GPL+Font Exception approaches this via: "As a special exception, if you create a document … this font does not by itself cause the resulting document to be covered by the GNU General Public License" The second aspect is that a font-name field is a "well-known identifier/key" linking to a particular set of metrics (this is partially why we have the Liberation Fonts that Red Hat kindly sponsored). For instance, the SIL OFL approaches this via a "Reserved Name" method: "No Modified Version of the Font Software may use the Reserved Font Name(s) unless explicit written permission is granted" There was a keen desire to a find a way to incorporate both of the above aspects, and others, into a leen "single-page" licence, in a way that did have the burden on seeking prior permission. Nor was likely to be overly scary to existing font foundries. As of 18 months ago, no out-of-the-box, font-specific, widely-used licence was available. Canonical set about sponsoring various work in the area. The 'Ubuntu' in the UFL name is there to discourage its use until it's been debugged. The list of open bugs/comments for the UFL-1.0 is visible at: https://launchpad.net/ubuntu-font-licence/+bugs The UFL-1.0 can't be changed, but work and advice is greatly sought for feeding into an UFL-1.1 and onwards; or functional equivalent. In the mean-time, the UFL-1.0 came about by looking for the closest starting point (SIL OFL 1.1) and making the minimum number of changes. For convenience, a colourful diff is available at: http://font.ubuntu.com/ufl/ofl-1.1-ufl-1.0.diff.html Yes, future relicensing is not only an option, but an intention. Yes, the meantime, I think 'contrib' would be a wondeful home. -Paul PS. I had another chat with Zack in Orlando about other methods of meeting a permissive, well-known name/Reserved Name criteria; Zack has been working issues around the DFSG and trademark compatibility, which may open up some options: http://www.debian-news.net/2011/11/12/bits-from-the-dpl-for-october-2011/ http://lists.debian.org/debian-project/2011/10/msg00028.html -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#688252: $EDITOR + /usr/bin/editor default
Hello Ben, I'm not the maintiner, but I saw a related question on 'ubuntu-users' so I started hunting. The related previous bugs on this topic are: 1. * enable ENV_EDITOR Fixed in sudo (1.4.1-1) on 1996-03-05. 2. "visudo does not default to vi"=20 http://bugs.debian.org/388659 Fixed in sudo (1.6.9p9-1) on 2007-12-03 3. "sudo: ignores configured editor preferences" http://bugs.debian.org/474197 Fixed in sudo (1.6.9p15-2), on 2008-04-16 with a revert + rationale: * revert the fix for 388659 such that visudo once again defaults to using /usr/bin/editor. I was always ambivalent about this change, it has caused more confusion and frustration than it cured, …" So my understanding is that since (April 2008) $EDITOR should be respected, and the fall-back should be to /usr/bin/editor. Could you clarify more precisely what you're seeing? -Paul -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#576474: Add Apple iPad USB IDs
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Package: usbmuxd Version: 1.0.3-1 Severity: minor Add new USB ID 05ac:129a (iPad Wifi). Upstream: http://git.marcansoft.com/?p=usbmuxd.git;a=commitdiff;h=803f6ccf Downstream: http://launchpad.net/bugs/555358 -Paul -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFLuRkTc444tukM+iQRArfLAJ9YInsddxu0KpsZPXPG098JJ+GTCACeIz3H fbwK66IqocRb9Xp6b11WGjc= =CEui -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#563011: iPod Touch 3G USB IDs
Package: usbmuxd Range extended 0x1290-0x1299 (iPod Touch 3G) and again to 0x1290-0x129a (iPad Wifi) -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#645354: [Pkg-fonts-devel] Bug#645354: Bug#645354: ttf-nanum: contained fontconfig setting files force to make it default font
On Sat, 15 Oct 2011, Changwoo Ryu wrote: > 2011-10-14 (금), 15:19 -0400, Scott Moser: > > Package: ttf-nanum (http://pad.lv/835304). > > Jinkyu Yi submitted the following fix to Ubuntu. > That "fix" will break default Debian desktop in Korean language. Not everything can be the one true default without regard to others. Fonts that install themselves as such can cause issues: "Microsoft Sans Serif maps to Thai font even on european setups" https://bugs.launchpad.net/ubuntu/+source/thaifonts-scalable/+bug/539008 > This is an ugly way. … But I strongly disagree with such a way; Can you offer Scott et al guidance on a more idea solution? -Paul -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#635068: Lintian Wishlist: validiate .ttf/.otf font files to check for DRM bits (fsType != 0)
Package: devscripts
Severity: Wishlist
X-Debbugs-CC: Debian Fonts Task Force
Based on the suggestion by Paul Wise on pkg-fonts, it would be useful
if Lintian were able to validate .ttf/.otf Truetype-based files and
issue warnings for those fonts that have DRM bits set.
The presence of such bits is designed to prevent the prevent full Free
use of the fonts; thread at:
http://lists.alioth.debian.org/pipermail/pkg-fonts-devel/2011-July/thread.html#6916
The DRM bits ('OS/2'.fsType != 0) can restrict the user's freedom
by preventing activities such installing/embedding/printing/editing.
Catching these bits will allow informing upstream font developers that
their fonts have such restrictions and therefore encouraging upstream
authors to free their fonts further. Many are simply aware, or have
been the victim of default export settings.
Proof that this is valuable and can generate results is at:
"Technical: Metadata: set OS/2.fsType=0 (Installable Embedding)"
https://bugs.launchpad.net/ubuntu-font-family/+bug/648406
Indeed, such a check proved so useful have it was added to the build
process for the Ubuntu Font Family, causing an error condition and
bailing out if required. For Lintian the check can simply be a
warning, or perhaps a error condition for fonts entering 'main'.
Example GPLv3 Python/fonttools code to check can be found at:
http://bazaar.launchpad.net/~sladen/ubuntu-font-family/midstream/view/head:/midstream/validate-libre-ttf.py
and for resetting the DRM bits:
http://bazaar.launchpad.net/~sladen/ubuntu-font-family/midstream/view/head:/midstream/patch-truetype.py
--remove-drm
Paul Wise notes that something can probably be done with existing
Perl-based font libaries too. Full documentation on the
Truetype/Opentype file format and the meaning of the 'fsType' DRM
flags can—and the GUI interface in Fontforge to change them—can be
found at:
http://partners.adobe.com/public/developer/opentype/index_os2.html#fst
http://fontforge.sourceforge.net/fontinfo.html#TTF-Values
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#139861: sort -u incorrect behaviour
More examples: $ echo -e "日\n本\nで\nは" | sort -u | wc -l 4 $ echo -e "日\n本\nで\nは" | sort | wc -l 3 Something is quite wrong (eg. this is definitely *incorrect behaviour*, rather than merely a difference of opinion over implementation). -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#676665: 'packages.dat' URL diff (cherry-picked from Odakyufan openbve=1.4.1.2a)
On Fri, 8 Jun 2012, Ronny Standtke wrote:
> seems to be http://www.openbve.org/.
There are few different people offering continuity forks; I'm tempted
to hold off doing any more newer Debian releases until things have
regrouped and stabilised and instead do a cherry-picked backport for
the 'packages.dat' location that Openbve tries to fetch.
> The site openbve tries to connect to (http://trainsimframework.org/)
Attached is the cherry-picked diff for updating the URLs; that should
fix this bug #676665, should be minimal enough to be Ubuntu SRU-able:
Update in-game 'packages.dat' fetch locations
This changes from a single 'trainsimframework.org' location to two
sites hosted by Anthony Bowden and Odakyufan respectively.
This was cherry-picked from the Odakyufan Openbve 1.4.1.2a tarball at:
http://odakyufan.zxq.net/openbve/
http://www.mediafire.com/download.php?xa6n5vp9euna0eu
md5sum *
1771e60bb900e9b9711949803748ce85 openbve1412a.7z
Ronny: are you in a position to test in this fix works for you; if so
we can package and upload it.
-Paul
Update in-game 'packages.dat' fetch locations
This changes from a single 'trainsimframework.org' location to two
sites hosted by Anthony Bowden and Odakyufan respectively.
This was cherry-picked from the Odakyufan Openbve 1.4.1.2a tarball at:
http://odakyufan.zxq.net/openbve/
http://www.mediafire.com/download.php?xa6n5vp9euna0eu
md5sum *
1771e60bb900e9b9711949803748ce85 openbve1412a.7z
diff --git a/openBVE/OpenBve/OldCode/formMain.GetAddOns.cs
b/Source/openBVE/OpenBve/OldCode/formMain.GetAddOns.cs
index 56bf6fa..3d5958b 100644
--- a/openBVE/OpenBve/OldCode/formMain.GetAddOns.cs
+++ b/Source/openBVE/OpenBve/OldCode/formMain.GetAddOns.cs
@@ -76,7 +76,9 @@
private void ConnectToServer() {
string[] urls = new string[] {
-
"http://trainsimframework.org/common/packages.dat";
+ "http://www.railsimroutes.net/packages.dat";,
+ "http://odakyufan.zxq.net/packages.dat";
};
string[] names = new string[] {
- "trainsimframework.org"
+ "railsimroutes.net",
+ "odakyufan.zxq.net"
};
Bug#827354: ITP: splat-vo -- Spectral Analysis Tool for Virtual Observatory
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Package: wnpp
Owner: Paul Sladen
Severity: Wishlist
* Package Name: splat-vo
* Version: 3.11-3
* Upstream Author: Margarida Castro Neves; Petr Skoda
* URL: http://www.g-vo.org/pmwiki/About/SPLAT
* Programming Language: Java
* License: GPL-2+
* Description: Spectral Analysis Tool for Virtual Observatory
Aladin is an desktop client for viewing and analysis of Virtual
Observatory (VO) spectral data. This graphical tool can display,
analyse, compare and modify already extracted spectra stored in FITS,
TEXT, and NDF/NDX formats. It can also interact with the Virtual
Observatory registery and data-centres to dynamically query and
download data for a particular section of the sky on-demand.
Splat-VO can work in multiple coordinate systems and can plot and
overlay multiple spectra at the same time. Plots are displayed in a
visually rich format. Analysis can be performed by selecting and
zooming multiple, and fitting to emission and absorption lines.
An example tutorial can be found at:
http://www.euro-vo.org/sites/default/files/documents/GAVO_splat-tutorial.pdf
As of 2016 upstream work is in the 'Splat-VO/Starjava' branch at:
https://github.com/SPLAT-VO/starjava/tree/splatmaster/splat
In the top-level of this repository also contains the rest of
dependencies, collectly known as 'Starjava', after the collaboration
project which created them. The Starjava master repo is at:
https://github.com/Starlink/starjava
There are a large number of dependencies on libraries contained in
the rest of Starjava, and these will need packaging first. Splat also
contains some JNI native code optimisation functions, with
pure-Java fallback implementations.
Licensing information is found in 'SPLAT_CONDITIONS' and 'COPYING':
https://github.com/SPLAT-VO/starjava/blob/splatmaster/splat/src/docs/SPLAT_CONDITIONS
And states that it is GPL-2+:
1. The Software is owned by the Council for the Central Laboratory
of the Research Councils ("CCLRC").
.
2. This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.
.
The GNU General Public License text is included in the file COPYING
which should be found in the same directory as this file.
There are several copyrights given, in the documentation at:
http://star-www.dur.ac.uk/~pdraper/splat/sun243.htx/node50.html
This is given as at least:
Copyright (C) 2001-2005 Council for the Central Laboratory of the Research
Councils
Copyright (C) 2006-2007 Particle Physics and Astronomy Research Council
Copyright (C) 2007-2009 Science and Technology Facilities Council
Copyright (C) 2010-2013 Peter W. Draper
Copyright (C) 2013 German Astrophysical Virtual Observatory (GAVO)
Upstream-Name: SPLAT-VO
Upstream-Author: Margarida Castro Neves
Source: https://github.com/SPLAT-VO/starjava/tree/splatmaster/splat
Section: science
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
iD8DBQFXYRrcc444tukM+iQRAvPAAJ9JD9JkIalfueKCV01vtyjkqdQzRQCgpATQ
XgiFQKO3H2KKDWG8E7o1+68=
=ubil
-END PGP SIGNATURE-
Bug#818373: Assembling complete set of cds.astro.* library source
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Package: libcds-astro-java
The Aladin 9.x upstream source distribution contains a 'cds.astro.jar'
with 31 class files, compiled on 10 May 2012. Source code matching
this is present in historic Aladin source snapshots: it is present in
Aladin 5.018, Aladin 7.014, and Aladin 7.533 under the 'cds/astro/*'
sub-directory.
Aladin 8.040 and later source releases do not have this 'cds/astro/'
sub-directory. The source code appears to have been split out to 'CDS
Units' distributed separately:
http://cds.u-strasbg.fr/cdsdevcorner/units1.2/src.zip
This is only a subset (9 .java files + 1 Makefile). A testcase for
'tUnit.java' was added, and the 'Makefile' refers other testcases.
This CDS Units package also includes one bugfix to 'Parsing.java',
ahead 'cds/astro/Parsing.java' in the Aladin 7.533 release (however
this bugfix says it is older, and is dated internally as May 2007).
The copies of source code in the Aladin 7.014 and Aladin 7.533 have
"GNU General Public License version 3" boilerplate added to the top of
each source file.
There It should be possible to retrieve and use this copy of the code
extracted form the Aladin 7.xxx releases:
https://anonscm.debian.org/cgit/debian-astro/packages/aladin.git/tree/cds/astro?h=upstream/7.533%2bdfsg
And to highlight the extra Makefile, unittest and 'parseSexa2()' bug
fix available in CDS-Units
List of differing files:
cds.astro.* package code ('-': not in v1.2 'CDS Units' 'Src.jar')
- -Astrocoo.java
Astroformat.java
- -Astroframe.java
AstroMath.java
- -Astropos.java
Astrotime.java
- -CheckPosition.java
Converter.java
- -Coocube.java
- -Coo.java
- -Ecliptic.java
Editing.java
- -Equatorial.java
- -FK4.java
- -FK5.java
- -Galactic.java
- -ICRS.java
+Makefile (added; +unittests)
- -Mtype.java
- -Otype.java
*Parsing.java (function parseSexa2() added in 2007-05)
- -Proj3.java
- -Qbox.java
- -QboxNumber.java
- -Supergal.java
+tUnit.java
Udef.java
Unit.java
Tests 't*.java' ('?': not found in v1.2 Src.jar):
?tAstroMath.java
?tAstrocoo.java
?tAstropos.java
?tConverter.java
?tCoo.java
?tOtype.java
?tParsing.java
?tPierre.java
?tQbox.java
?tQbox2.java
?tQboxNumber.java
tUnit.java (in v1.2; but source says 2000-03-04)
nb. 'tUnit.java' (only) is also found separately at:
http://cds.u-strasbg.fr/cdsdevcorner/units1.0/tUnit.java
Historic Aladin 5.x source discovered at:
http://www.asiaa.sinica.edu.tw/~chyan/Aladin5/Aladin5_src/cds/astro/
http://www.asiaa.sinica.edu.tw/~chyan/Aladin5/Aladin-src5.018.jar
Historic Aladin 7.x source was obtained from backups of a colleague.
(Any other historic source dumps would be appreciated!)
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
iD8DBQFXYmdgc444tukM+iQRAkKmAJ0Z50UamZG/ktrUuOtw9qsylVXcVACglR11
n+FWGWUt9zq+MQRFnPUbzRc=
=R9yg
-END PGP SIGNATURE-
Bug#680475: 680475 - jsamp: jsamp + libjsamp-java + libjsamp-java-doc
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 retitle 680475 ITP: jsamp -- Java Simple Application Messaging Protocol for Virtual Observatory thanks Upstream has noted that the main library can also be called on the commandline as an executable. Simplify naming + version based on feedback with Upstream authors. -Paul -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFXY+eJc444tukM+iQRAoq+AKCnn7RyxweXZ7R9p2QuacPLWwJDoACgg2V2 PJ+DMrSnFuLHtgUFMMonP3g= =2KFi -END PGP SIGNATURE-
Bug#818376: ITP: libcds-astro-java: CDS Astronomical Units conversions
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: libcds-astro-java * Version: 1.2 * Upstream Author: François Ochsenbein, Strasbourg astronomical Data Center (CDS) * URL: http://cds.u-strasbg.fr/resources/doku.php?id=units * Programming Language: Java * License: Educational-use only (non-free) * Description: Java-based library for units conversion in the astronomical Virtual Observatory CDS Astro is a units conversion library that provides library functionality under the 'cds.astro' namespace. It is one of the dependencies of the Aladin Virtual Observatory (VO) client; see: The library was written by the Strasbourg astronomical Data Center (CDS) and source code is made available by CDS as a zip file from: http://cds.u-strasbg.fr/cdsdevcorner/units1.2/src.zip (This states it is from November 2006, but contains files modified in February 2007). This package is one of the dependencies for the CDS Aladin package, see: 'vo-aladin', ITP: #818373. At present the read me information on the homepage at: http://cds.u-strasbg.fr/resources/doku.php?id=units contains a link to: http://cds.u-strasbg.fr/resources/doku.php?id=using_conditions which includes restrictions on use, only allowing for: "the reproduction and representation as a private copy or for educational and research purposes outside any lucrative use" / "la reproduction et la representation a titre de copie privee ou des fins d'enseignement et de recherche et en dehors de toute utilisation lucrative" This is incompatible with the GPLv3 licence under which the rest of Aladin is already published and distributed. Contact has been made with the authors in the of gaining a clear confirmation of GPLv3 licencing. This would need to be done prior to acceptance of Aladin and its dependencies into Debian. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFW6X2Cc444tukM+iQRAoVLAKCGyofE4nZKFv34gltfjGT9+AzEYQCfQFpj wznT/Edi56v5MZsGd3inYh0= =MDFw -END PGP SIGNATURE-
Bug#818375: ITP: libsavot-java: Simple Access to VOTable (SAVOT)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: libsavot-java * Version: 4.0 * Upstream Author: Andre Schaaff (CDS), Laurent Bourges (JMMC) * URL: https://github.com/aschaaff/savot/ * Programming Language: Java * License: GPL-3 * Description: SAVOT: Java-based VO-Table (Virtual Observatory) library Simple Access to VOTable (SAVOT) is a Java library for VO-Tables which lives under the 'cds.savot' namespace. SAVOTv4 is a library written in Java, and maintined by the Strasbourg astronomical Data Center (CDS). It has been released by one of its co-authors, Andre Schaaff under the GPLv3 on Github: . https://github.com/aschaaff/savot/ This package is one of the dependencies for the CDS Aladin package, see: 'vo-aladin', ITP: #818373. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFW6Xoec444tukM+iQRAgFvAJ9hIMnqVUMZAry3g444wLuXcgIr4gCfRkpP ShwLE/DlSaRbM+o5RiCsa2Q= =TT79 -END PGP SIGNATURE-
Bug#818378: ITP: libcds-moc-java: Multi-Order Coverage Virtual Observatory library
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: libcds-moc-java * Version: 4.5 * Upstream Author: Pierre Fernique, Strasbourg astronomical Data Center (CDS) * URL: http://wiki.ivoa.net/twiki/bin/view/IVOA/MocInfo * Programming Language: Java * License: GPL-3 * Description: Java-based library for Multi-Order Coverage access in the astronomical Virtual Observatory The Multi-Order Coverage (MOC) library provides access to functionality under the 'cds.moc' namespace: . http://wiki.ivoa.net/twiki/bin/view/IVOA/MocInfo The library is maintained by the Strasbourg astronomical Data Center (CDS) and published by the International Virtual Observatory Alliance (IVOA) under the GPLv3. A GPLv3 licence statement is included at top of each .java source file when downloaded from: http://wiki.ivoa.net/internal/IVOA/MocInfo/MocSrc.jar It is one of the dependencies of the Aladin Virtual Observatory (VO) client; see: 'vo-aladin', ITP: #818373. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFW6YDEc444tukM+iQRAh/1AKDCugMNbvde2HSyT4p0RHrc3NmnuwCfWegp gK9Fmp/PqI2Q2/sq6DT2L6w= =wzT2 -END PGP SIGNATURE-
Bug#818373: ITP: vo-aladin
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: vo-aladin * Version: 9.013 * Upstream Author: Pierre Fernique, Strasbourg astronomical Data Center (CDS) * URL: http://aladin.u-strasbg.fr/ * Programming Language: Java * License: GPL-3 * Description: Interactive sky atlas for astronomical images and datasets Aladin is an desktop client for viewing and analysis of Virtual Observatory (VO) astronomical datasets in real-time. Data, tables and images are dynamically fetched and displayed to the user, via the network of registries and and data centres offering access using VO protocols. . Aladin is written in Java, released under the GPLv3 and maintined by the Strasbourg astronomical Data Center (CDS). . http://aladin.u-strasbg.fr/ The main code base is written in Java and published under the GPLv3 as Src.jar files, at: http://aladin.u-strasbg.fr/java/nph-aladin.pl?frame=downloading These 'jarballs' are unversioned; a tool to download and renumber the upstream sources jarballs is available at: https://github.com/sladen/vo/blob/master/aladin-meta/fetch-aladin-source.py Several internal dependencies are used, (included by default as binary .class object in the source distribution: 'microhub' + dependencies, 'kxml2', 'jsamp', 'moc', 'cds-astro', 'savot'. These dependences need separately packaging and uploading first, and are listed in more detail at: https://github.com/sladen/vo/blob/master/aladin-meta/dependencies.md The current application-to-application data sharing protocol is SAMP, meaning that it should be possible to safely disable the the 'microhub') dependency. The Plastic Manager (supplied by Microhub) and its further dependencies appear to be frequently disabled in the main code anyway, and can likely be fully commented out the via the 'USE_PLASTIC_REQUESTED=false' code-paths. 'libkxml2-java' is already packaged in Debian as: https://packages.debian.org/search?keywords=libkxml2-java 'jsamp' is available in 'debian-astro' revision control, based on previous work by Florian Rothmaier: http://anonscm.debian.org/cgit/debian-astro/packages/jsamp.git/ 'MOC' presently has an unknown licence: http://wiki.ivoa.net/twiki/bin/view/IVOA/MocInfo Both the 'CDS-Astro' and 'SAVOT3' libraries have a non-free educational-use-only licence, requiring explicit permission for anything else. The Present restrictive CDS licence text is in: http://cds.u-strasbg.fr/resources/doku.php?id=using_conditions "reproduction and representation as a private copy or for educational and research purposes" / "la reproduction et la representation a titre de copie privee ou des fins d'enseignement et de recherche" Which is linked from the download pages at: http://cds.u-strasbg.fr/resources/doku.php?id=3Dunits "Read before any use of available Java classes" http://cds.u-strasbg.fr/resources/doku.php?id=3Dsavot_v3.0=20 "Read before any use of available Java classes" Although SAVOT3 appears under the educational-only (non-free) licence, there is a SAVOT4 GPL3 implementation from Andre Schaaff (CDS) to be found on Github. If this is API-compatible, it should be possible to package and switch to this: https://github.com/aschaaff/savot/ Communication is underway in order to try to get moc, cds-astro and, if needed SAVOT3, clearly under the GPL-3 licence aswell as this is required prior to inclusion in Debian. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFW6XOGc444tukM+iQRAkxQAKCTKJ95bFDfQxzelodiBFKhQ5bGXACdFBIN nFTvBYEMRZ2JqrBnOAVgXRM= =btVR -END PGP SIGNATURE-
Bug#860811: python-pymoc should Suggest: python-healpy
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Source: pymoc Severity: Wishlist Currently in upstream pymoc, there is a soft dependency on healpy. The only place where is healpy is imported and called is into the plotter (util/plot.py), in one function plot_moc(). Ideally this relationship should be declaring in the packaging metdata. However, Healpy is only available for some architectures (64-bit ones), and so declaring a hard dependency on Healpy would make Pymoc uninstallable for several architectures (including i386). Package: python-healpy Architecture: any-ia64 any-amd64 any-arm64 any-ppc64 any-ppc64el any-sparc64 Healpy itself in-turn relies on a large number of other packages. Therefore it would be preferable to only declare a soft Suggests: or Recommends: could be added so that specialist functionaltiy can be easily used by users on 64-bit architectures without inconviencing other users. Christian Dersch notes that Fedora currently uses a Suggests: against Healpy. Ole Streicher notes that the situation could probably be further improved afterwards by wrapping the 'import healpy' in a try: except: clause and skipping that part of the plotting when unavailable. This would allow the code to complete, even if some aspects, eg. grid lines, would not be visible in the resulting output.(*) -Paul (*) Hopefully upstream may be able to offer input as to whether it is better to: (a) run through without drawing some of the output; vs. (b) generate a stacktrace in these corner cases. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFY+LI0c444tukM+iQRAo8vAKCLRj+RZCmGOtNhV7xKa2fjMtHEOQCghDrU QkYfYnvxEW2/CcF9lg+eLqw= =iQjJ -END PGP SIGNATURE-
Bug#860816: unblock: pymoc/0.4.2-1
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: unblock Please unblock pymoc/0.4.2-1 for stretch. This fixes a FTBFS on stretch/i386 reported by Lucas Nussbaum during the full-rebuild. This was triggered by a failing testcase on 32-bit systems: https://bugs.debian.org/860629 https://github.com/grahambell/pymoc/issues/3 The fix has come from upstream and arrived within 24 hours; the patch contains 2-lines of real code changes (in lib/pymoc/io/fits.py). The remainder of the diff are two new test case and accompanying changelog/documentation tweaks. This would appear to meet criteria 1; (a FTBFS marked as 'serious', with a targetted fix with optional documentation changes) uploaded via unstable: https://release.debian.org/stretch/freeze_policy.html Full debdiff attached. -Paul -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFY+Motc444tukM+iQRAmRwAKCD2SkybjBrURrCI0pce0t+UskA1ACePnbp /UU7IwJdFfiUuPPeJKwYlTw= =J+Lg -END PGP SIGNATURE- diff -Nru pymoc-0.4.1/CHANGES.txt pymoc-0.4.2/CHANGES.txt --- pymoc-0.4.1/CHANGES.txt 2016-09-26 13:05:27.0 +0200 +++ pymoc-0.4.2/CHANGES.txt 2017-04-20 05:15:32.0 +0200 @@ -1,3 +1,8 @@ +0.4.2 2017-04-19 + +- Corrected bug reading MOCs in FITS format on 32-bit systems with + 64-bit NUNIQ columns. + 0.4.1 2016-08-08 - Corrected bug in intersection routine. diff -Nru pymoc-0.4.1/debian/changelog pymoc-0.4.2/debian/changelog --- pymoc-0.4.1/debian/changelog2016-09-26 13:12:24.0 +0200 +++ pymoc-0.4.2/debian/changelog2017-04-20 15:21:00.0 +0200 @@ -1,3 +1,10 @@ +pymoc (0.4.2-1) unstable; urgency=low + + * Import upstream bugfix release (0.4.2) + + Fix large FITS import on 32-bit architectures (Closes: #860629) + + -- Paul Sladen Thu, 20 Apr 2017 15:21:00 +0200 + pymoc (0.4.1-1) unstable; urgency=low * Initial release (Closes: #839744) diff -Nru pymoc-0.4.1/debian/copyright pymoc-0.4.2/debian/copyright --- pymoc-0.4.1/debian/copyright2016-09-26 13:12:24.0 +0200 +++ pymoc-0.4.2/debian/copyright2017-04-20 15:21:00.0 +0200 @@ -5,7 +5,7 @@ Files: * Copyright: 2014-2016, Graham Bell - 2015-2016, East Asian Observatory + 2015-2017, East Asian Observatory 2013-2014, Science and Technology Facilities Council License: GPL-3+ diff -Nru pymoc-0.4.1/doc/.gitignore pymoc-0.4.2/doc/.gitignore --- pymoc-0.4.1/doc/.gitignore 1970-01-01 01:00:00.0 +0100 +++ pymoc-0.4.2/doc/.gitignore 2017-04-20 05:15:32.0 +0200 @@ -0,0 +1 @@ +/_build diff -Nru pymoc-0.4.1/.gitignore pymoc-0.4.2/.gitignore --- pymoc-0.4.1/.gitignore 1970-01-01 01:00:00.0 +0100 +++ pymoc-0.4.2/.gitignore 2017-04-20 05:15:32.0 +0200 @@ -0,0 +1,5 @@ +*~ +*.pyc +/MANIFEST +/build +/dist diff -Nru pymoc-0.4.1/lib/pymoc/io/fits.py pymoc-0.4.2/lib/pymoc/io/fits.py --- pymoc-0.4.1/lib/pymoc/io/fits.py2016-09-26 13:05:27.0 +0200 +++ pymoc-0.4.2/lib/pymoc/io/fits.py2017-04-20 05:15:32.0 +0200 @@ -1,4 +1,5 @@ # Copyright (C) 2013-2014 Science and Technology Facilities Council. +# Copyright (C) 2017 East Asian Observatory. # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by @@ -143,8 +144,12 @@ current_order = None current_cells = [] +# Determine type to use for orders: 32 bit if column type is J, +# otherwise assume we need 64 bit. +moc_type = np.int32 if (hdu.data.formats[0] == 'J') else np.int64 + nuniqs = hdu.data.field(0) -orders = (np.log2(nuniqs / 4) / 2).astype(int) +orders = (np.log2(nuniqs / 4) / 2).astype(moc_type) cells = nuniqs - 4 * (4 ** orders) for (order, cell) in izip(orders, cells): diff -Nru pymoc-0.4.1/lib/pymoc/version.py pymoc-0.4.2/lib/pymoc/version.py --- pymoc-0.4.1/lib/pymoc/version.py2016-09-26 13:05:27.0 +0200 +++ pymoc-0.4.2/lib/pymoc/version.py2017-04-20 05:15:32.0 +0200 @@ -13,4 +13,4 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. -version = '0.4.1' +version = '0.4.2' diff -Nru pymoc-0.4.1/test/test_ascii.py pymoc-0.4.2/test/test_ascii.py --- pymoc-0.4.1/test/test_ascii.py 2016-09-26 13:05:27.0 +0200 +++ pymoc-0.4.2/test/test_ascii.py 2017-04-20 05:15:32.0 +0200 @@ -40,3 +40,28 @@ write_moc_ascii(moc, file=out) self.assertEqual(out.getvalue(), test_ascii_sorted) + +def test_ascii_large(self): +orig = MOC() +orig.add(29, [ +3458700, 3458707, +3458708, 3458709, +]) + +out = StringIO() +write_moc_asci
Bug#832884: ITP: python-gavovotable -- library for Virtual Observatory VOTables
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: python-gavovotable * Version: 0.9.6 * Upstream Author: Markus Demleitner * URL: http://docs.g-vo.org/DaCHS/votable.html * Programming Language: Python * License: GPL-3+ * Description: Command-line TAP Shell for Virtual Observatory The VOTable format is defined by the International Virtual Observatory Alliance to facilitate data exchange in astronomy. This library implements most of the specification and offers simple and advanced, streaming interfaces. Also included is a simple client to query TAP servers. A manual is available at: http://docs.g-vo.org/DaCHS/votable.html Previous packaging working had been done since 2009, and published in the APT repositary documented at: http://soft.g-vo.org/repo http://vo.ari.uni-heidelberg.de/debian Thus, despite this being the initial upload to Debian itself, to likely makes sense to follow the package name and changelog in the existing source packages. -Paul -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFXmylVc444tukM+iQRAneRAKDN3+p173LW8iAM6VhUxqLIVXevnwCdFnqF JrdEwJ6tqKBKMiob/bdx/94= =y1gB -END PGP SIGNATURE-
Bug#832891: ITP: tapsh --- Table Access Protocol Shell for Virtual Observatory
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
package: wnpp
Owner: Paul Sladen
Severity: Wishlist
* Package Name: tapsh
* Version: 0.7.1
* Upstream Author: Markus Demleitner
* URL: http://soft.g-vo.org/tapsh
* Programming Language: Python
* License: GPL-3+
* Description: Command-line TAP Shell for Virtual Observatory
TAP ("Table Access Protocol") is a protocol that lets you send
database queries and data to servers within the Virtual Observatory
(VO). The TAP shell provides a command line interface to querying TAP
servers, complete with metadata management and command line
completion. It allows for sending cross-service queries,
i.e. queries across individual TAP servers.
The tool is built-upon the ARI/GAVO 'votable' library
(python-gavovotable) that is associated with GAVO Dachs Internally the
tool uses the GAVO 'votable' library
A manual is available at:
http://docs.g-vo.org/tapsh-manual.html
Previous packaging working had been done in 2012 and 2014.
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
iD8DBQFXmye2c444tukM+iQRAq9iAJoCYWzPUTp2r/qcW79XDLi1JvpmCwCgiIcV
x/s8f5c1+S4SvtNk7lONxqk=
=6r+h
-END PGP SIGNATURE-
Bug#832918: ITP stilts -- Astronomical Starlink Tables Infrastructure Library Tool Set
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: stilts * Version: v3.0-7 * Upstream Author: Mark Taylor * URL: http://www.star.bris.ac.uk/~mbt/stilts/ * Programming Language: Java * License: GPL, GPL+Classpath, LGPL, Apache, CDDL, * Description: Astronomical Starlink Tables Infrastructure Library Tool Set Stilts (Starlink Tables Infrastructure Library Tool Set) is a command-line and scriptable interface for handling tables within the astronomical observatory. Stilts can handle and convert between a number of astronomical table formats, including FITS, ColFITS, VOTable, NASA CDF, ASCII, CSV/TSV/TST, Gaia DPAC GBIN (additional dependencies required), IPAC, and WDC; viz: http://www.star.bristol.ac.uk/~mbt/topcat/sun253/inFormats.html For packaging, we will need to add a '/usr/bin/stilts' wrapper: http://www.star.bris.ac.uk/~mbt/stilts/sun256/sun256.html#stilts-cmd http://www.star.bris.ac.uk/~mbt/stilts/sun256/sun256.html#invoke Which will probably contain something like: exec java -classpath topcat-lite.jar uk.ac.starlink.ttools.Stilts It is built upon the Starlink Tables Infrastructure Library Tool Set (STIL) and is the command-line counterpart to the graphical 'Topcat' application. From the wrapper command we can see the close relationship to Topcat, although hopefully more modularisation can be achieved so that only 'default-jre-headless' is required. In addition to the large dependency list shared with Topcat, a number of those dependencies in turn come from Starlink which has sub-modules with several licences. The various licences are documented by Upstream in: http://www.star.bris.ac.uk/~mbt/stilts/LICENSE.txt and a fuller-audit will be required about the precise dependency relationship and corresponding licence compatilibity. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFXm3atc444tukM+iQRAp8tAKDgmwjDaLP/yL2fCYqYXhy1xXtcwACg4jmH Pi9EvgNjYQnZfp2AHnu/oeY= =OPUg -END PGP SIGNATURE-
Bug#835071: openbve: New homepage URL?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 tags 835071 + patch easy thanks https://anonscm.debian.org/cgit/pkg-cli-apps/packages/openbve.git/commit/?id=b04d4dc https://anonscm.debian.org/cgit/pkg-cli-apps/packages/openbve-data.git/commit/?id=aba6 A grep shows lots of additional instances, so I'm wondering whether it's worth introducing a patch to update the links in the documentation and user-interface too. -Paul -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFYbPQjc444tukM+iQRAgkkAJ9/rwx3g88cjBNNfOD4Oofd5CwuYQCeKbiv Rs905RejK/qpqw98TGIgl/0= =oZvS -END PGP SIGNATURE-
Bug#839744: ITP: PyMOC -- Python Multi-Order Coverage map library
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 package: wnpp Owner: Paul Sladen Severity: Wishlist * Package Name: pymoc * Version: 0.4.1 * Upstream Author: Graham Bell * URL: https://github.com/grahambell/pymoc * Programming Language: Python * License: GPLv3+ * Description: Python Multi-Order Coverage map library PyMOC manipulates a stack of Multi-Order Coverage maps for exchanging images within the Virtual Observatory (VO). These MOCs can be encoded as FITS, JSON or ASCII formats. Multiple MOCs can be stacked using union, intersection and subtraction the result converted and saved in alternative formats. A 'pymoctool' command-line tool that accesses the library is included. MOCs can be viewed and rendered by other tools, including Aladin. Debian Packaging can be found at: https://anonscm.debian.org/git/debian-astro/packages/pymoc.git -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) iD8DBQFX87Cbc444tukM+iQRAvRYAKCCUIT2UVMZ44YXLj/O26jdVjoXLwCfbEHQ 3aOOXCusDJO7JQStnkxj17g= =B4mj -END PGP SIGNATURE-
Bug#254834: NMU: Change htons()->htonl() to fix multicast (Closes: #254834).
Non-maintainer upload via DELAYED, to fix this bug as I hit it yesterday and it's been open (with fix supplied) for 22months. Thomas: can you cancel this from Delayed and apply it manually if you're not happy with it; debdiff attached. -Paul -- High on a Spanish mountain, surrounded by howling dogs. Southampton, GB diff -u udptunnel-1.1/debian/changelog udptunnel-1.1/debian/changelog --- udptunnel-1.1/debian/changelog +++ udptunnel-1.1/debian/changelog @@ -1,3 +1,9 @@ +udptunnel (1.1-1.1) unstable; urgency=low + + * NMU: Change htons()->htonl() to fix multicast (Closes: #254834). + + -- Paul Sladen <[EMAIL PROTECTED]> Sun, 21 May 2006 15:00:02 +0100 + udptunnel (1.1-1) unstable; urgency=low * Initial Release. --- udptunnel-1.1.orig/udptunnel.c +++ udptunnel-1.1/udptunnel.c @@ -217,7 +217,7 @@ (*relays)[i].udpaddr.sin_port = htons(udpport + i); (*relays)[i].udpaddr.sin_family = AF_INET; (*relays)[i].udp_ttl = udpttl; -(*relays)[i].multicast_udp = IN_MULTICAST(htons(udpaddr.s_addr)); +(*relays)[i].multicast_udp = IN_MULTICAST(htonl(udpaddr.s_addr)); (*relays)[i].tcpaddr.sin_addr = tcpaddr; (*relays)[i].tcpaddr.sin_port = htons(tcpport + i);
Bug#540247: abi-compliance-checker 1.1-1, manpage patches and tweaks
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hello Ryan et al, I ended up packaging this aswell, but you bet me to a submission by a clear 18 hours. I've cloned your git tree of the package and tweaked it; I wonder if you'd be interested in reviewing/accepting the patches. The main one is switching to a hand-written and formatted manpage which I spent a little while writing and so would be interested to see integrated. git shortlog of the changes I'm offering shows: .gitignore: *~ set unreleased 1.1-2 Include handwritten manpage Use install -T 0755 and dh_installdirs debian/watch: Case insentisve match (MediaWiki CamelCasing) debian/copyright: add email address for upstream maintainer Copyedit Description: in control/break up paragraphs expand RAS to Russian Academy of Sciences in manpage keep lintian happy debian/control: further grammar fix debian/changelog: NMU as 1.1-1.1 if required I've kept every little change as separate commits so you can pick and choose easily; on the (small) off-chance that you're happy with the whole lot as-is, the last commit is an NMU boilder plate, allowing it to be signed+sponsored as-is---drop that if you need to cherry pick. http://github.com/sladen/abi-compliance-checker/tree/manpage-et-al git://github.com/sladen/abi-compliance-checker.git Hope it's useful, and thank you for being such a speedy packager! -Paul - -- Why do one side of a triangle when you can do all three. Somewhere, GB. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFKfYB6c444tukM+iQRAjBDAKC5twxl9bYKcEhbeaL7pTLowuA3fgCcDwVV qzoer1s8T5w6BeUcy1IL/To= =tufx -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#481960: openssh-blacklist: compressed with bzip2, breaks upgrade on dpkg <1.10.24
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Package: openssh-blacklist Severity: grave The 'openssh-blacklist' package is currently "questionably useful" on older Debian systems. Issue; 1. With the security update, 'openssh-server' 1:4.3p2-9 now Depends: 'openssh-blacklist'. 2. 'openssh-blacklist' 0.1.1 has been packaged with bzip2 compression (contains data.tar.bz2). 3. Versions of 'dpkg-deb' below 'dpkg' <1.10.24 do not understand bzip2. 4. Because of the hard dependency, a fixed 'openssh-server' does not get configured/restarted. Example; (dpkg=1.10.21) $ sudo apt-get -f install ... Unpacking openssh-blacklist (from .../openssh-blacklist_0.1.1_all.deb) ... dpkg-deb: file `/var/cache/apt/archives/openssh-blacklist_0.1.1_all.deb' contains ununderstood data member data.tar.bz2, giving up dpkg: error processing /var/cache/apt/archives/openssh-blacklist_0.1.1_all.deb (--unpack): subprocess dpkg-deb --fsys-tarfile returned error exit status 2 $ sudo apt-get install openssh-server ... The following packages have unmet dependencies: openssh-server: Depends: openssh-blacklist but it is not going to be installed Background; The 'openssh-blacklist' packages contains approximately 2*100k*80-bit hashes; around ~2MB of raw (effectively random) data. This data is hex-encoded upto a larger uncompressed size of ~4MB (primarily containing '[0-9a-f\n]'). Being a minimal-entropy (although sorted) data-set, there is little to be gained from the choice of compressor (+/-5%; entirely from hex de-expansion); aside from reduced compatibility. Ideally the package would be revved and uploaded using gzip compression, for compatilibity with long-standing previous installations of Debian---which perhaps receive no updates except for critical security issues. -Paul - -- Why do one side of a triangle when you can do all three. Somewhere, GB. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFIMdMpc444tukM+iQRAnHzAKCY4SKopHjjMt3EULyT286Y8BLZAgCeLX+v 06H03pGpfy23P7SoWFyltpA= =tWOz -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#481960: openssh-blacklist: compressed with bzip2, breaks upgrade on dpkg <1.10.24
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Mon, 19 May 2008, Kees Cook wrote: > On Mon, May 19, 2008 at 08:21:04PM +0100, Paul Sladen wrote: > > The 'openssh-blacklist' package is currently "questionably useful" on > That's a very old version of dpkg that doesn't support bzip2. Ubuntu has shipped with a dpkg that did not include bzip2 support[0]; and that was not so very long ago. Sarge was the first version of Debian to ship with a bzip2-capable dpkg, which was *less than 3 years ago*[1]. Put simply. Anyone with a three-year old Debian install was either running pre-release software at install time, or *may[2] have hit this issue*. > The space-savings is non-trivial, The space-savings from bzip2'ing 'openssh-server' and 'openssh-client' are greater than the saving from bzip2'ing 'openssh-blacklist'.[3] > so bzip2 is going to stay [for 'openssh-blacklist'] If 'openssh-server' and 'openssh-client' are deamed important enough to be 'gziped' then any hard dependencies should be treated the same---otherwise the core (gziped) package is uninstallable[4] and the entire dependency tree for 'sshd' might as well be bziped. If new dependencies are introduced [during a security fix] I think an extra 100kB on the mirror is preferable to *zero or more* broke packaging setups on systems.[5] (The installed on-disk size for the user is the same)[6][7]. When packages started to be bziped in Ubuntu (eg. OOo) those packages received 'Pre-Depends: >=1.10.24' for dpkg; I would hope the same is true for more system critical packages. I would rather see something fail cleanly, than break. Leaving both the packaging system in an inconsistent state and no 'sshd' running.[7] Any security fix should ideally be doing the minimum of alteration, and require the minimum of intervention. The OpenSSH update was a /critical/ security update, in response to a /critical/ bug introduced early. I hope that the update can be made available as a distro-signed binary package for as many affected systems as possible, without the further risk of leaving any of the those machines broken. It is not the security fix that has lead to a breakage. It is _new_ code and the packaging choices made with it: 1). A hard 'Depends:' for a non-critical (but good-idea) extension. 2). Use of a non-critical (but mild space-saving) packaging extension. 3). Non-use of a non-critical (but good practice) Pre-Depends: safety net. 4). Installation on a not-the-latest (but well serving) Debian system. It is the unfortunately coincidental combination of all four. I'm happy to supply patches to update any of the three avenues, should you have a preference. -Paul [0] http://changelogs.ubuntu.com/changelogs/pool/main/d/dpkg/dpkg_1.10.22ubuntu2.1/changelog although not an LTS release. [1] Sarge (Debian 3.1), released on 2005-06-06. (buggy) openssl 0.9.8c-1 uploaded on 2006-09-17. [2] Depending on the exact set of intermediate updates/upgrades. [3] Saving of 121639 bytes vs. 117867 bytes (i386 and all respectively). The saving on the binaries packages massively outweighs the saving on the all package when multiplied across 12 architectures. Note: the advdef (in 'advancecomp') deflater beats .bzip2 on the 'openssh-blacklist/data.tar' and is still backwards compatibile with older versions of dpkg. [4] In the situations where it is deamed preferable to have .gz [5] That size even comes up, I fear, is a distraction from the packaging compatibility issue. [6] As an interesting aside, with a raw binary packing of the hash blobs, the .gz, beats no packing, beats .bz2. [7] I no longer have immediate access to the first machine I hit this on; I had closed the ssh session and have not location why 'sshd' had been stopped before the installation failed. Why, I don't know. -- Why do one side of a triangle when you can do all three. Somewhere, GB. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFIMgjTc444tukM+iQRAr22AJ9OcDTf0vYF6NfL3dbNbSD/9YpffwCeLbfM RfysF67ngvoB9pv6GJAgUGs= =sabs -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#481960: openssh-blacklist: compressed with bzip2, breaks upgrade on dpkg <1.10.24
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Tags: patch On Tue, 20 May 2008, Paul Sladen wrote: > I'm happy to supply patches to update any of the three avenues, Two one-lines debdiffs attached, for using '-Zgzip', _or_ 'Pre-Depends'. I would prefer the 'gzip' option, as this maintains zero-delta when backporting to sarge/warty era systems. -Paul - -- Why do one side of a triangle when you can do all three. Somewhere, GB. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFIMvFfc444tukM+iQRAobYAKDWvdwIm2ktKQhIWZZWpuHdstLJTACfbp1R bIsl25AkigRzvCTcjohK77k= =bao9 -END PGP SIGNATURE- diff -Nru openssh-blacklist-0.1.1/debian/changelog openssh-blacklist-0.1.1/debian/changelog --- openssh-blacklist-0.1.1/debian/changelog2008-05-13 17:23:16.0 +0300 +++ openssh-blacklist-0.1.1/debian/changelog2008-05-20 18:19:25.0 +0300 @@ -1,3 +1,11 @@ +openssh-blacklist (0.1.1-1gzip) various; urgency=low + + * Pack .deb with gzip (instead of bzip2) for increased compatibility +as 'openssh-blacklist' is distributed as a hard Depend: of other packages + that are. Closes: #481960. + + -- Paul Sladen <[EMAIL PROTECTED]> Tue, 20 May 2008 16:30:37 +0100 + openssh-blacklist (0.1.1) stable-security; urgency=low * Set Maintainer to Kees for Debian. diff -Nru /tmp/kVSSxbPKZL/openssh-blacklist-0.1.1/debian/rules /tmp/2BgQzW3JeU/openssh-blacklist-0.1.1/debian/rules --- openssh-blacklist-0.1.1/debian/rules2008-05-12 23:20:20.0 +0300 +++ openssh-blacklist-0.1.1/debian/rules2008-05-20 18:18:26.0 +0300 @@ -64,7 +64,7 @@ dh_shlibdeps dh_gencontrol dh_md5sums - dh_builddeb -- -Zbzip2 + dh_builddeb -- -Zgzip binary: binary-indep binary-arch .PHONY: build clean binary-indep binary-arch binary install configure -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQBIMu4Yc444tukM+iQRAm/oAKCkXE2z0BFMzx2Ds757xXXwdoDKjQCfVmjY Feb3yHN9pz3L/p4PBLV4CKk= =t2Bz -END PGP SIGNATURE- diff -Nru openssh-blacklist-0.1.1/debian/changelog openssh-blacklist-0.1.1/debian/changelog --- openssh-blacklist-0.1.1/debian/changelog2008-05-13 17:23:16.0 +0300 +++ openssh-blacklist-0.1.1/debian/changelog2008-05-20 18:08:45.0 +0300 @@ -1,3 +1,15 @@ +openssh-blacklist (0.1.1-1predepend) various; urgency=low + + * Add Pre-Depends: dpkg (>=1.10.24) because of bzip2 use. + - This requirement of -policy was removed on 2007-12-21[0], but is +still important for those maintaining sarge/warty-based systems, where +otherwise a broken packaging situation will be caused by the +failed installation when dpkg-deb encounters the 'data.tar.bz2'. +[0] http://lists.debian.org/debian-project/2007/12/msg00047.html +Closes: #481960. + + -- Paul Sladen <[EMAIL PROTECTED]> Tue, 20 May 2008 16:47:14 +0100 + openssh-blacklist (0.1.1) stable-security; urgency=low * Set Maintainer to Kees for Debian. diff -Nru /tmp/AD2p6dFdZd/openssh-blacklist-0.1.1/debian/control /tmp/WGAm5ah8V9/openssh-blacklist-0.1.1/debian/control --- openssh-blacklist-0.1.1/debian/control 2008-05-13 17:17:57.0 +0300 +++ openssh-blacklist-0.1.1/debian/control 2008-05-20 17:52:23.0 +0300 @@ -8,6 +8,7 @@ Package: openssh-blacklist Architecture: all Recommends: openssh-client +Pre-Depends: dpkg (>= 1.10.24) Description: list of blacklisted OpenSSH RSA and DSA keys Contains the list of known-bad OpenSSH keys, for ssh-vulnkeys to use when examining suspect keys. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQBIMu4hc444tukM+iQRAiWZAKDTKQ3zk8l+CVq8GlIxGgKRbELKSQCfSfth HBh12TH1bt9cPvrxEiVKcCA= =zXwT -END PGP SIGNATURE-
Bug#314334: apt-get on jffs2 / disabling cacheing completely
Hit this "apt relying on shared writeable" issue today with a jffs2 issue. N800/Maemo seem to have got around this by (I think using an older version of apt?); it's all a bit messy and could really do with handling transparently within apt. First messages I found on this were from 2001, over at handhelds.org. Another option may just be to skipping the cacheing completely---it's probably a bit unnecessary on flash devices with (relatively low) seek times. -Paul -- Why do one side of a triangle when you can do all three. Helsinki, FI. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#519549: 'at' introduces unecessary Depends: on exim4
Package: at Version: 3.1.10.2 'at' now has a hard dependency on 'exim4|mail-transport-agent', which is not necessary for the functioning of 'at'. As in the case of butterflies and earthquakes, this can have fallout... the situation having been discovered thusly: during a dist-upgrade, 'at' forced installation of 'exim4-base', which in-turn started running on port 0.0.0.0:25 automatically. Following a later watchdog-effected reboot, the newly installed exim4 on the vserver host "stole" port 25 from all of the vserver instances, thereby rejecting mail for multiple virtual machines. The only immediate "clean" solution was removing both the 'at' and 'exim4-base' packages from the vserver host; which has broken the ability to use 'at' for batches/delayed execution (such as backup runs). Ideally 'at' should only Suggests: the installation of a MTA. For reference, Ubuntu currently ship: Version: 3.1.10.2ubuntu1 Depends: libc6 (>= 2.4), libpam0g (>= 0.99.7.1), lsb-base (>= 3.2-14) Suggests: postfix | mail-transport-agent -Paul -- Why do one side of a triangle when you can do all three. Somewhere, GB. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#476796: usb.USBError: No error, workaround
In the meantime, for anyone else who hits this bug while coding with pyusb
and gets the message:
usb.USBError: No error
I used the following kludge, which likely will also work for you:
...
try:
data = fd.interruptRead(endpoint.address, endpoint.maxPacketSize)
except usb.USBError as e:
if e.args != ('No error',): # http://bugs.debian.org/476796
raise e
...
-Paul
--
Why do one side of a triangle when you can do all three. Somewhere, GB.
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#562572: usbmuxd debian/ directory licence differs from upstream
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Source: usbmuxd Version: 1.0.0-1 usbmuxd and libusbmuxd are distributed under the GPLv2+ and LGPLv2.1+ licences, respectively. Please relicense the debian/ additions under a compatible lowest common denominator licence (LGPLv2.1+, BSD, or Public Domain). This will ensure that software needing to run or link against *usbmuxd on a Debian/Ubuntu system is not subject to differing licensing conditions than could be readily expected when linking against an upstream source bundle, or one found on a another distribution; usbmuxd-1.0.0-1: +Files: * +License: GPL-2+ ... +Files: libusbmuxd/* +License: LGPL-2.1+ ... +Files: debian/* +License: GPL-3+. The copyright holders in question for debian/* are recorded as: (C) 2009 Bradley Smith (C) 2009 Julien Lavergne -Paul - -- Why do one side of a triangle when you can do all three. Somewhere, GB. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFLNYOic444tukM+iQRAh/aAJ9ukqQe/l1ttK5qSheNFvgChnDalgCgj4+P +IFszrz71S7G2QBdNp9eORA= =DoL9 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#562574: openttd.config prompts with db_input high, when not required
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Source: openttd Version: 0.7.3-1 During installation, the 'openttd.config' checks for the existance of proprietary game data files in the directory: /usr/share/games/openttd/data and if the files are not found will run: db_input high openttd/datafiles to display: For its operation, OpenTTD needs the data files from the original Transport Tycoon Deluxe game. See the /usr/share/doc/openttd/README.Debian file for more details about the needed files and their location.p Three are three problems with this solution: (a) The test is incorrect. It is sufficient to merely chdir to any directory (for example, '~/.wine/drive_c/MPS/TTWin95/') and then run 'openttd', to enable the game to operate. (b) The message is set to "high" priority when it is non-essential, and already has a reasonable default. (c) The message is given repeatedly, even during a single upgrade session. Ideally, the prompting should be removed completely, so that it no longer disrupts the automatic update process. Failing that, the message could be sent to "low" priority with success returned always, such that it is only shown once per upgrade session. -Paul - -- Why do one side of a triangle when you can do all three. Somewhere, GB. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFLNa0Mc444tukM+iQRAt7NAJ9MPX2aPC9tQyvDkyU21U1kUdOZ5QCgp/N3 FtUcMzwobF8j565aXHK0y+A= =oSBn -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
