Re: conffile purging and maintainer scripts
On Thu, 23 Mar 2006, Ian Jackson wrote:
> Thomas Hood writes ("Re: conffile purging and maintainer scripts"):
> > If a file /etc/foo was formerly a conffile of the package but no
> > longer is so then /etc/foo should be dealt with in the preinst or
> > postinst.
>
> Regrettably this is currently true. I think this is a bug in dpkg and
> I think I know how to fix it. I'll see what I can do about it in a
> more appropriate venue than debian-devel :-).
Maybe at the same time you could integrate ucf into dpkg... because with
ucf you have a similar problem: the configuration file is no longer a
conffile since it's handled by ucf, and you have to take care to remove it
in maintainer scripts (on purge).
Cheers,
--
Raphaël Hertzog
Premier livre français sur Debian GNU/Linux :
http://www.ouaza.com/livre/admin-debian/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: libopenobex library transition
On Thu, Mar 23, 2006 at 05:16:55PM +0100, Hendrik Sattler wrote: > Am Donnerstag, 23. März 2006 06:36 schrieb Steve Langasek: > > On Wed, Mar 22, 2006 at 04:55:39PM +0100, Hendrik Sattler wrote: > > > Sadly, they changed the soname again. So instead of directly uploading > > > the new upstream version, I would like to wait for the following bugs to > > > be resolved, > > > first: > > > affix: #358279 > > > kdebluetooth: #358275 > > > multisync: #358283 > > > obexftp: needs upload of 0.19 (have to ask my sponsor :) > > First of all, why are there already two libopenobex packages in unstable > > (libopenobex -> libopenobex1, and libopenobex1.0 -> libopenobex1.0-0)? > > Secondly, why do the above bugs need to be fixed before uploading the new > > version of libopenobex? AIUI, they already build-depend only on > > libopenobex-1.0-0-dev, which is the older version of the library. > Right. > > Third, if you're changing the name of the -dev package anyway, why are you > > changing it to libopenobex1-dev instead of libopenobex-dev? Is the API > > expected to change with every soname change? > No but I cannot say that it will never change in an incompatible way (be it > renaming of structs, functions, the whole glue or how the library can be > detected). > I followed libpkg-guide here: > http://www.netfort.gr.jp/~dancer/column/libpkg-guide/libpkg-guide.html#id265760 Yes, please don't follow that recommendation; there's nothing "ideal" about putting sonames in the -dev package names -- having such -dev packages Provide: libpackage-dev is better than nothing, but again it still gives problems if you need to add a versioned build-dep as in this case. Junichi's library packaging guide is indeed pretty good for the most part, but this particular recommendation is a bad one. > Additionally, libopenobex-1.0-0-dev already has a > Provides: libopenobex-dev > which makes this impossible (could stay installed in this case). If all packages are supposed to build-depend on libopenobex-dev (>= $version), then it's again not an issue. > > > When those are in testing, I will request removal of the old binary lib > > > packages (libopenobex-1.0-0*) from testing and sid. > > Well, that answers my first question. It also means that those bugs are > > filed at the wrong severity: these packages currently do *not* FTBFS, > > because you haven't requested removal of the old library yet. > BTW: do they get removed automatically after some time or do I have to > request > that? What about the source packages? I gather that this was addressed in a previous message. > > > The build dependencies of those packages should then be at > > > libopenobex-dev (>= 1.1) > > > > That won't work. libopenobex-dev is a virtual package; you can't have a > > versioned (build-)dep on a virtual package. > Ui, didn't know that. Then it must follow every soname manually... No, there's no reason *at all* to change -dev package names in conjunction with soname changes. You might consider it worthwhile to change -dev package names when the API changes incompatibly, but that's not the same thing as an soname chnage. Cheers, -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. [EMAIL PROTECTED] http://www.debian.org/ signature.asc Description: Digital signature
RE: Australian timezone (australasia) update for this weekend
Hi, No need to answer this one, I found that when running the zic on your file, the update was done to the: /usr/share/zoneinfo/Australia/Sydney File, while the timezone was set to NSW. I moved the NSW file, then made a link from NSW to Sydney, then ran the zdump and it worked fine. Regards, Michael. -Original Message- From: Mansour, Michael Sent: Friday, 24 March 2006 3:23 PM To: [EMAIL PROTECTED]; debian-user@lists.debian.org Cc: debian-devel@lists.debian.org Subject: RE: Australian timezone (australasia) update for this weekend Hi Anibal, I've followed your manual procedure below, and have your "australasia" file in /root. I run: # zic australasia Which returns to the prompt, but then when running zdump it hasn't changed anything. I also note that /etc/localtime is a link to: /usr/share/zoneinfo/Australia/NSW and that file is dated 2003. How else can this be done? Thanks. Michael. -Original Message- From: Aníbal Monsalve Salazar [mailto:[EMAIL PROTECTED] Sent: Thursday, 23 March 2006 10:23 AM To: debian-user@lists.debian.org Cc: debian-devel@lists.debian.org Subject: Australian timezone (australasia) update for this weekend Hello, The Australian timezone change is a couple of days away. If you are running servers depending on acurrate time, you will need to check or update them. If you are running sid (unstable) or etch (testing), to update your Australian timezone, just run 'apt-get update; apt-get install libc6'. For stable (sarge) and oldstable (woody) you may have to update your system mannually. To check/change your timezone, use tzconfig. In my case, tzconfig tells me that my timezone is Australia/Melbourne. To check if your timezone change will happen on either "Sun Mar 26" or "Sun Apr 2", run 'zdump -v Australia/Melbourne | grep 2006'. Replace Australia/Melbourne with your own timezone. If your timezone change will happen on "Sun Apr 2", you're safe. If your timezone change will happen on "Sun Mar 26", you may want to update the Australian timezone. If you cannot update libc6, you could update the Australian timezone mannually, as described in the following lines. 1. Download the timezone (australasia) file. It includes the Australia, Indian and Pacific timezone information. wget -nH -N http://people.debian.org/~anibal/timezone/australasia If you don't want to use my copy of the australasia file, download http://ftp.gnu.org/pub/gnu/glibc/glibc-2.4.tar.bz2 (size is 15,202,445) and extract glibc-2.4/timezone/australasia. Together with my copy of the australasia file, there are files with the md5 checksum and my gpg signature of the australasia file. 2. Backup you zoneinfo directory. tar cf - /usr/share/zoneinfo | bzip2 -9 - > zoneinfo.tar.bz2 3. As root, run: zic australasia 4. Check your timezone change. zdump -v Australia/Melbourne | grep 2006 It should show "Sun Apr 2". Regards, Aníbal Monsalve Salazar -- http://v7w.com/anibal
RE: Australian timezone (australasia) update for this weekend
Hi Anibal, I've followed your manual procedure below, and have your "australasia" file in /root. I run: # zic australasia Which returns to the prompt, but then when running zdump it hasn't changed anything. I also note that /etc/localtime is a link to: /usr/share/zoneinfo/Australia/NSW and that file is dated 2003. How else can this be done? Thanks. Michael. -Original Message- From: Aníbal Monsalve Salazar [mailto:[EMAIL PROTECTED] Sent: Thursday, 23 March 2006 10:23 AM To: debian-user@lists.debian.org Cc: debian-devel@lists.debian.org Subject: Australian timezone (australasia) update for this weekend Hello, The Australian timezone change is a couple of days away. If you are running servers depending on acurrate time, you will need to check or update them. If you are running sid (unstable) or etch (testing), to update your Australian timezone, just run 'apt-get update; apt-get install libc6'. For stable (sarge) and oldstable (woody) you may have to update your system mannually. To check/change your timezone, use tzconfig. In my case, tzconfig tells me that my timezone is Australia/Melbourne. To check if your timezone change will happen on either "Sun Mar 26" or "Sun Apr 2", run 'zdump -v Australia/Melbourne | grep 2006'. Replace Australia/Melbourne with your own timezone. If your timezone change will happen on "Sun Apr 2", you're safe. If your timezone change will happen on "Sun Mar 26", you may want to update the Australian timezone. If you cannot update libc6, you could update the Australian timezone mannually, as described in the following lines. 1. Download the timezone (australasia) file. It includes the Australia, Indian and Pacific timezone information. wget -nH -N http://people.debian.org/~anibal/timezone/australasia If you don't want to use my copy of the australasia file, download http://ftp.gnu.org/pub/gnu/glibc/glibc-2.4.tar.bz2 (size is 15,202,445) and extract glibc-2.4/timezone/australasia. Together with my copy of the australasia file, there are files with the md5 checksum and my gpg signature of the australasia file. 2. Backup you zoneinfo directory. tar cf - /usr/share/zoneinfo | bzip2 -9 - > zoneinfo.tar.bz2 3. As root, run: zic australasia 4. Check your timezone change. zdump -v Australia/Melbourne | grep 2006 It should show "Sun Apr 2". Regards, Aníbal Monsalve Salazar -- http://v7w.com/anibal
[hardware-donation] Please send in your requests for hardware
For a while now we have maintained a web page listing hardware that developers are looking for to help with their Debian development efforts: http://www.debian.org/misc/hardware_wanted Through this page, we have successfully obtained a number of machines, for example some Cobalt boxes used for the MIPS port. If you, as a porter, someone working on debian-installer, or doing other Debian work, need any specific hardware for your work please feel free to send an email to [EMAIL PROTECTED] and I'll add it to the list. Obviously, there's no guarantee we'll find something but it's always worth a try. Also, it's more likely to find older hardware people might be willing to donate, but if you have any other specific requests please get in contact anyway as I might be able to find sponsors. -- Martin Michlmayr http://www.cyrius.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Уникальные семинары!
On Thursday 23 March 2006 08:04, Jacob S wrote: > Except that, as has been discussed many times before... 1) the C-Rs are > coming from uol.com.br Not disputed. > 2) there are some legitimate users that post from > uol.com.br that do not have C-R on their accounts So they should get a real email provider instead of going with the default, if not cut ties with doing business with spammers entirely. > So you would be blocking a lot of innocent users and the problem > account would still be on the list. Challenge-response constitutes spam. If you do business with an ISP that employs challenge-response, you are doing business with spammers. No loss to the net. -- Paul Johnson Email and IM (XMPP & Google Talk): [EMAIL PROTECTED] Jabber: Because it's time to move forward http://ursine.ca/Ursine:Jabber -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Two Common Light Events in particular Spring 2006
Re: Уникальные семинары!
Em Qui 23 Mar 2006 13:04, Jacob S escreveu: > 3) the problem > address is not a uol.com.br account - the problem account has their > mail forwarded to a uol.com.br account, so the listmasters have not > been able to track down the problem account. Just a side note: the last time I wrote to debian-user-portuguese I received several such mails, from different people. I don't know what to do, but uol's CR is a plague for every brasilian list I know. Tiago. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#358695: ITP: latex-utils -- utilities for LaTeX/xfig
Package: wnpp
Severity: wishlist
Owner: Vincent Danjean <[EMAIL PROTECTED]>
* Package name: latex-utils
Version : 2.1.2
Upstream Author : Vincent Danjean <[EMAIL PROTECTED]>
Arnaud Legrand <[EMAIL PROTECTED]>
* URL : http://gforge.inria.fr/projects/latex-utils/
* License : GPL (2 or any later version)
Description : utilities for LaTeX/xfig
This package provides a Makefile to compile LaTeX documents (in ps or pdf),
latex packages to easily include xfig figures in LaTeX documents and various
scripts help the Makefile to correctly and easily handle its job.
.
One great interrest of this package is that it automatically track most of the
dependencies of the LaTeX document. We should just have to create a Makefile
with the single line 'include LaTeX.mk'
.
Homepage: http://gforge.inria.fr/projects/latex-utils/
Additional notes:
-
This software differs from latex-mk (that has recently enter the debian
archive) in two points :
1) it allows to easily manage .fig files from latex documents
(ie \includegrpahics{foo.fig})
2) the Makefile fragment automatically tracks the dependencies (bib
files, included files, figures, ...). There is no need to setup and
maintain Makefile variables.
You can find the current version of the debian package on my web page:
http://dept-info.labri.fr/~danjean/deb.html#latex-utils
Of course, the /usr/share/bug/latex-utils/control will be removed before
the package will be uploaded (this file is here so that people that are
already using the package can use reportbug)
Best regards,
Vincent
PS:
I would appreciate any help to improve my description as I am not a
native english speaker.
-- System Information:
Debian Release: testing/unstable
APT prefers unstable
APT policy: (990, 'unstable'), (500, 'testing'), (1, 'experimental')
Architecture: i386 (i686)
Shell: /bin/sh linked to /bin/bash
Kernel: Linux 2.6.16
Locale: LANG=fr_FR.UTF-8, LC_CTYPE=fr_FR.UTF-8 (charmap=UTF-8)
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#159385: The only real cure for losing we ght
www.sometimeithurtnotfeel.com/ht9/ is the store to dwindle those we ight the prophecy come true.That is exactly what I am trying to say sir, but on a different note. but they are too influenced by other big countries. On the other hand, Poland is large Interpol's urgent global security alert, known as an "orange notice," was issued "because the escape and unknown whereabouts of al-Qaida terrorists constituted a clear and present danger to all countries," the statement said. Crooks is possibly the loneliest character on the ranch along side Curleys wife who is also very majestic world, and bear the palm alone. This helped persuade Brutus to join the faction, and Churchills -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: This account has been closed. Please refer to www.df-ig.com for new contact information.
I continually get this message. Your IMAP service wishes to alert you to the following LOGIN Failed. Your account has been cancelled. Why? Who do I contact? I have version 6.00.2800.1123 Barbara Snyder Barbara Snyder
Re: conffile purging and maintainer scripts
Thomas Hood writes ("Re: conffile purging and maintainer scripts"):
> If a file /etc/foo was formerly a conffile of the package but no
> longer is so then /etc/foo should be dealt with in the preinst or
> postinst.
Regrettably this is currently true. I think this is a bug in dpkg and
I think I know how to fix it. I'll see what I can do about it in a
more appropriate venue than debian-devel :-).
> Roger Leigh wrote:
> > 1) sarge -> etch upgrades
> > -
> >
> > In order to handle upgrades from sarge correctly, maintainers will
> > still have to manually remove conffiles in their maintainer scripts
> > until at least etch+1 by my reckoning. Is this correct?
Some of the folklore recipies I've seen for doing this are
fundamentally broken. They look in /var/lib/dpkg/status (which may or
may not contain the right information) and they use filenames like
*.dpkg-tmp which are reserved for the use of dpkg.
I'm working on grepping archives for affected packages.
Ian.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: conffile purging and maintainer scripts
Anthony DeRobertis writes ("Re: conffile purging and maintainer scripts"):
> 5. Decide you'd rather keep locally installed FOO, purge Debian FOO package.
DDTT
> Solution to this one is that admins should follow the FHS and put their
> config files in /etc/local/ :-D
No, the solution is not to purge something if you want to keep its
configuration !
Ian.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ???????????????????? ????????????????!
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Thu, 23 Mar 2006 11:52:12 -0800 Tony Godshall <[EMAIL PROTECTED]> wrote: > According to Jacob S, > > -BEGIN PGP SIGNED MESSAGE- > > Hash: SHA1 > > > > On Thu, 23 Mar 2006 07:48:38 -0800 > > Paul Johnson <[EMAIL PROTECTED]> wrote: > > > > > On Thursday 23 March 2006 02:41, Henning Makholm wrote: > > > > Scripsit Paul Johnson <[EMAIL PROTECTED]> > > > > > > > > > On Wednesday 22 March 2006 09:45, Henning Makholm wrote: > > > > >> Listmasters have been trying to > > > > >> identify the responsible subscriber with no luck > > > > > > > > > > Why not just 500 all posts from sites known to use > > > > > challenge-response? > > > > > > > > The challenges are send directly from the idiot site to the From > > > > address in the list posting. They do not pass through Debian > > > > machines. > > > > > > However, in the future, people from that site would not be able to > > > confirm their subscription in the first place if their site uses > > > challenge-response. > > > > Except that, as has been discussed many times before... 1) the C-Rs > > are coming from uol.com.br 2) there are some legitimate users that > > post from uol.com.br that do not have C-R on their accounts 3) the > > problem address is not a uol.com.br account - the problem account > > has their mail forwarded to a uol.com.br account, so the > > listmasters have not been able to track down the problem account. > > > > So you would be blocking a lot of innocent users and the problem > > account would still be on the list. > > Do the people on uol.com.br have a choice? Perhaps they > could vote with their wallets? Probably. Except, I have a feeling we'd end up blocking all isps and large businesses and only allow traffic from personally owned domains. Not that that's an entirely bad thing, though... :-) Jacob -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.2.2 (GNU/Linux) iD8DBQFEIwPLkpJ43hY3cTURAtJ/AKC3eU2jgoCHjZUNGdDIKQ3fJziJpQCfd6fk xqSLiw4bUGT1cq3gLsVbCzM= =pPC8 -END PGP SIGNATURE-
Bug#358666: ITP: o2sms -- A perl library and script to send SMS messages using .ie websites
Package: wnpp Severity: wishlist Owner: "Conall O'Brien" <[EMAIL PROTECTED]> * Package name: o2sms Version : 2.0.107 Upstream Author : David McNamara <[EMAIL PROTECTED]> * URL : http://www.mackers.com/projects/o2sms/ * License : Perl/Artistic Description : A perl library and script to send SMS messages using .ie websites o2sms is a library and and script to send SMS messages via the command line using the websites of Irish mobile operators. This is achieved by simulating a web browser's interaction with those websites. This script requires a valid web account with o2 Ireland, Vodafone Ireland or Meteor Ireland. At the time of writing, o2 give 250 free web texts per month, Vodafone 300 and Meteor 300. After sending a message, o2sms will tell you how many you have remaining. Features: * Support for multiple recipients and groups. * Will split long messages over multiple 160-character texts. * Built-in address book. * Config file for storing username, password and other options. * Support for cookies, which are reused between invocations. * 'Remaining messages' report. * HTTP(S) proxy support. * Append a signature to every text. * T9 emulation (currently limited to capitalisation of first letter in sentences). * Allows messages to be split manually (use "\\" in messages). * Support for squeezed messages (suport 4 squez msg). * Designed for easy integration with other programs. -- System Information: Debian Release: 3.1 Architecture: i386 (i686) Kernel: Linux 2.6.8-2-386 Locale: LANG=en_IE.utf8, LC_CTYPE=en_IE.utf8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ???????????????????? ????????????????!
According to Jacob S, > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Thu, 23 Mar 2006 07:48:38 -0800 > Paul Johnson <[EMAIL PROTECTED]> wrote: > > > On Thursday 23 March 2006 02:41, Henning Makholm wrote: > > > Scripsit Paul Johnson <[EMAIL PROTECTED]> > > > > > > > On Wednesday 22 March 2006 09:45, Henning Makholm wrote: > > > >> Listmasters have been trying to > > > >> identify the responsible subscriber with no luck > > > > > > > > Why not just 500 all posts from sites known to use > > > > challenge-response? > > > > > > The challenges are send directly from the idiot site to the From > > > address in the list posting. They do not pass through Debian > > > machines. > > > > However, in the future, people from that site would not be able to > > confirm their subscription in the first place if their site uses > > challenge-response. > > Except that, as has been discussed many times before... 1) the C-Rs are > coming from uol.com.br 2) there are some legitimate users that post from > uol.com.br that do not have C-R on their accounts 3) the problem > address is not a uol.com.br account - the problem account has their > mail forwarded to a uol.com.br account, so the listmasters have not > been able to track down the problem account. > > So you would be blocking a lot of innocent users and the problem > account would still be on the list. Do the people on uol.com.br have a choice? Perhaps they could vote with their wallets? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#358657: ITP: libsl -- memory-efficient generic linked list library
Package: wnpp Severity: wishlist Owner: Kari Pahula <[EMAIL PROTECTED]> * Package name: libsl Version : 0.3.3 Upstream Author : Stig Brautaset <[EMAIL PROTECTED]> * URL : http://brautaset.org/software/sl/ * License : GPLv2 or later Description : memory-efficient generic linked list library sl is a memory-efficient generic linked list library. It doesn't use container nodes. Instead it requires a pointer to the next item directly in the datastructure you want to create lists (or stacks) of. This can give you significant memory savings when creating long lists of small structures. It also allows for fast push and pop operations since there is no need to allocate or free memory for the container nodes. It also means that a push can't fail because memory couldn't be allocated for the container node. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.15.6 Locale: LANG=C, [EMAIL PROTECTED] (charmap=ISO-8859-15) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#358659: ITP: libggtl -- generic game-tree search library
Package: wnpp Severity: wishlist Owner: Kari Pahula <[EMAIL PROTECTED]> * Package name: libggtl Version : 2.1.2 Upstream Author : Stig Brautaset <[EMAIL PROTECTED]> * URL : http://brautaset.org/software/ggtl/ * License : GPLv2 or later Description : generic game-tree search library GGTL is a library designed to make it easier to program games in C. It provides an AI that is able to play most 2 player strategic games. Nim, Tic-Tac-Toe, Reversi (aka Othello), Connect-4 and Chess are all examples of games that can all be implemented using GGTL. The provided Reversi and Nim extensions implement all the game-specific callback functions GGTL's AI needs. Using one of these extensions means you can have a game with a capable AI up and running in next to no time. Doing so incurs no penalty in flexibility, however--you can override any provided callback function with your own. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.15.6 Locale: LANG=C, [EMAIL PROTECTED] (charmap=ISO-8859-15) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: libopenobex library transition
> > - don't wait for the other packages to be updated before uploading the new > > soname, if one is needed > > - don't wait for the other packages to be updated before filing a removal > > request for the old library version. > > Where do I request such a removal? You do this by fileing a bug against ftp.debian.org I would think. Michael -- Michael Meskes Email: Michael at Fam-Meskes dot De, Michael at Meskes dot (De|Com|Net|Org) ICQ: 179140304, AIM/Yahoo: michaelmeskes, Jabber: [EMAIL PROTECTED] Go SF 49ers! Go Rhein Fire! Use Debian GNU/Linux! Use PostgreSQL! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: libopenobex library transition
Am Donnerstag, 23. März 2006 06:36 schrieb Steve Langasek: > On Wed, Mar 22, 2006 at 04:55:39PM +0100, Hendrik Sattler wrote: > > Sadly, they changed the soname again. So instead of directly uploading > > the new upstream version, I would like to wait for the following bugs to > > be resolved, > > first: > > affix: #358279 > > kdebluetooth: #358275 > > multisync: #358283 > > obexftp: needs upload of 0.19 (have to ask my sponsor :) > > First of all, why are there already two libopenobex packages in unstable > (libopenobex -> libopenobex1, and libopenobex1.0 -> libopenobex1.0-0)? > > Secondly, why do the above bugs need to be fixed before uploading the new > version of libopenobex? AIUI, they already build-depend only on > libopenobex-1.0-0-dev, which is the older version of the library. Right. > Third, if you're changing the name of the -dev package anyway, why are you > changing it to libopenobex1-dev instead of libopenobex-dev? Is the API > expected to change with every soname change? No but I cannot say that it will never change in an incompatible way (be it renaming of structs, functions, the whole glue or how the library can be detected). I followed libpkg-guide here: http://www.netfort.gr.jp/~dancer/column/libpkg-guide/libpkg-guide.html#id265760 Additionally, libopenobex-1.0-0-dev already has a Provides: libopenobex-dev which makes this impossible (could stay installed in this case). > > When those are in testing, I will request removal of the old binary lib > > packages (libopenobex-1.0-0*) from testing and sid. > > Well, that answers my first question. It also means that those bugs are > filed at the wrong severity: these packages currently do *not* FTBFS, > because you haven't requested removal of the old library yet. BTW: do they get removed automatically after some time or do I have to request that? What about the source packages? > > The build dependencies of those packages should then be at > > libopenobex-dev (>= 1.1) > > That won't work. libopenobex-dev is a virtual package; you can't have a > versioned (build-)dep on a virtual package. Ui, didn't know that. Then it must follow every soname manually... > But I don't see any good reason why libopenobex-dev shouldn't be the *real* > package name. > > So, please: > > - make the name of the -dev package libopenobex-dev instead of > libopenobex1-dev, because sonames shouldn't be encoded in -dev package > names libpkg-guide says the opposite. > - don't wait for the other packages to be updated before uploading the new > soname, if one is needed > - don't wait for the other packages to be updated before filing a removal > request for the old library version. Where do I request such a removal? HS
Re: Уникальн ые семинары!
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Thu, 23 Mar 2006 07:48:38 -0800 Paul Johnson <[EMAIL PROTECTED]> wrote: > On Thursday 23 March 2006 02:41, Henning Makholm wrote: > > Scripsit Paul Johnson <[EMAIL PROTECTED]> > > > > > On Wednesday 22 March 2006 09:45, Henning Makholm wrote: > > >> Listmasters have been trying to > > >> identify the responsible subscriber with no luck > > > > > > Why not just 500 all posts from sites known to use > > > challenge-response? > > > > The challenges are send directly from the idiot site to the From > > address in the list posting. They do not pass through Debian > > machines. > > However, in the future, people from that site would not be able to > confirm their subscription in the first place if their site uses > challenge-response. Except that, as has been discussed many times before... 1) the C-Rs are coming from uol.com.br 2) there are some legitimate users that post from uol.com.br that do not have C-R on their accounts 3) the problem address is not a uol.com.br account - the problem account has their mail forwarded to a uol.com.br account, so the listmasters have not been able to track down the problem account. So you would be blocking a lot of innocent users and the problem account would still be on the list. Jacob -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.2.2 (GNU/Linux) iD8DBQFEIscmkpJ43hY3cTURAsAIAJwJjIAMOTLJwtT2KRa092aARICRjgCfU4P6 RbB4NiGot40moPv5o9SmhDU= =zcAY -END PGP SIGNATURE-
Re: Уникальные семинары!
On Thursday 23 March 2006 02:41, Henning Makholm wrote: > Scripsit Paul Johnson <[EMAIL PROTECTED]> > > > On Wednesday 22 March 2006 09:45, Henning Makholm wrote: > >> Listmasters have been trying to > >> identify the responsible subscriber with no luck > > > > Why not just 500 all posts from sites known to use challenge-response? > > The challenges are send directly from the idiot site to the From > address in the list posting. They do not pass through Debian machines. However, in the future, people from that site would not be able to confirm their subscription in the first place if their site uses challenge-response. -- Paul Johnson Email and IM (XMPP & Google Talk): [EMAIL PROTECTED] Jabber: Because it's time to move forward http://ursine.ca/Ursine:Jabber -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: for those who care about GNOME^w glib applications.
On 3/22/06, Steve Langasek <[EMAIL PROTECTED]> wrote: > On Wed, Mar 22, 2006 at 11:29:01AM -0300, Gustavo Franco wrote: > > > > There's a workaround that is run the application as follows: > > > > $ G_SLICE=always-malloc application > > > > Thanks for the information. Can you explain (or provide a pointer to) why > > > these bugs are application bugs rather than bugs in glib, and does glib > > > need > > > to be adding << conflicts with those applications in order to avoid broken > > > partial upgrades? > > > Yes, the new GSlice allocator is triggering those[0] bugs[1]. It happens > > due to > > different (but similar) reasons, e.g: In nettool and galculator[2] the > > code should > > call g_object_unref() - that decrease the ref count and frees the memory > > when > > the ref count is zero, instead call g_free() directly. > > Right, I saw the patch posted to the evolution bug, which made perfect sense > -- objects are being allocated via one API and then being freed by another, > unrelated API that happened to work before because they were both > implemented on top of malloc()/free(). Yes, evo was NMUed. > So, unquestionably an application bug. > > > FYI, GMemChunk (old and deprecated by the upstream) was reimplemented to > > use GSlice, so no need to change or rebuild code to be affected due to buggy > > code. I don't know exactly why asking GSlice to force allocate and free > > memory > > through "standard malloc" solves the problem, maybe someone else can > > clarify this point. > > It works because g_free() is a wrapper around free(), so if the address > passed to g_free() doesn't point to memory allocated using malloc(), glibc > will abort as shown. Making sure everything is allocated with malloc() > works around this. Oh, i see. Thanks! > > Talking about the << conflicts thing. The fact that we don't have a list of > > affected applications yet, i'm working on a partial list[3] based on > > the bugs already > > opened. Hopefully it's less grave than you think. > > Ok, seems like you have this under control then. :) Yes, just a few bugs reported and probably two without a patch (classpath and gnome-games) atm. The others were patched through the BTS, NMUed or Ubuntu has a fix. I'll take a look into the classpath, but the bug was already forwarded to the upstream, so there's a possibility that they've a patch. Cheers, -- stratus
Re: bet there are no senior citizen developers
Stephen Gran wrote: > > > This is Jidanni you're talking to. On Thu, Mar 23, 2006 at 12:56:02PM +, Martin Michlmayr wrote: > [Having said that, I think many of his bug reports are valuable and > that developers sometimes close bugs too quickly because of his bad > reputation for filing random bug reports that only waste people's > time; see e.g. #347975] I've had useful reports from Dan too. He certainly seems to have widespread interests as he has reports against a large number of packages. It's just rude and unnecessary to dismiss him as Stephen did. This is a fine example of the hostility problem so many people report Debian to have. Hamish -- Hamish Moffatt VK3SB <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bet there are no senior citizen developers
* MJ Ray <[EMAIL PROTECTED]> [2006-03-23 12:48]: > > This one time, at band camp, Michael Banck said: [...] > > > Or rethink whether your issue needs posting at all. > > This is Jidanni you're talking to. > Please follow Michael Banck's advice before posting more > opaque comments that look like pure personal attacks. To make this less opaque, see for example #357095 (and pay attention to the contents of the control message). [Having said that, I think many of his bug reports are valuable and that developers sometimes close bugs too quickly because of his bad reputation for filing random bug reports that only waste people's time; see e.g. #347975] -- Martin Michlmayr http://www.cyrius.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bet there are no senior citizen developers
Stephen Gran <[EMAIL PROTECTED]> > This one time, at band camp, Michael Banck said: [...] > > Or rethink whether your issue needs posting at all. > > This is Jidanni you're talking to. Please follow Michael Banck's advice before posting more opaque comments that look like pure personal attacks. -- MJR/slef My Opinion Only: see http://people.debian.org/~mjr/ Please follow http://www.uk.debian.org/MailingLists/#codeofconduct -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: automatically install -dev packages
On Wed, Mar 22, 2006 at 10:08:06PM +0200, George Danchev wrote: > In fact in cases like that you will find auto-apt quite useful. It wraps the > build process in its own environment, peeps around for the files access and > propose to install them on demand via apt-get. > E.g. try: auto-apt run ./configure --whatever I'd rather not. I've read about that suggestion before and that way the system installs everything that configure /could/ use. Frankly, I do not need tcl8.0, tcl8.1 and tcl8.2 in addition to tcl8.3 which I have installed. So, for some configure scripts auto-apt can be useful, but the bigger projects often probe for different incarnations of the same program and normally you would not want to install all of them. In short: Be careful, especially when running ./configure with auto-apt! Greetings Torsten signature.asc Description: Digital signature
Re: Уникальные семинары!
Scripsit Paul Johnson <[EMAIL PROTECTED]> > On Wednesday 22 March 2006 09:45, Henning Makholm wrote: >> Listmasters have been trying to >> identify the responsible subscriber with no luck > Why not just 500 all posts from sites known to use challenge-response? The challenges are send directly from the idiot site to the From address in the list posting. They do not pass through Debian machines. -- Henning Makholm "This imposes the restriction on any procedure statement that the kind and type of each actual parameter be compatible with the kind and type of the corresponding formal parameter." -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[RFC] proftpd 1.3.0rc5 in experimental
I just uploaded the new pre-release of the proftpd daemon. It has been completely repackaged to support apache-like DSO and it is now a single binary package (no more proftpd-* flavors, but for a few dummy packages used for upgrading). Feedbacks are very welcome to find upgrading issues and whatever. AFAIK it can upgrade by aptitude current sid packages. I was looking for how managing dependencies for a plain apt-get dist-upgrade path, but failed 'til now... Ideas? Maybe, I need to maintain a dummy package for proftpd-common too, but I hate introducing too many pseudo-packages. And yes, I know that aptitude is the sane choice for dist-upgrading :) so I'm not too worried about. -- Francesco P. Lovergine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bet there are no senior citizen developers
This one time, at band camp, Michael Banck said: > Hi Dan, > > On Thu, Mar 23, 2006 at 06:27:18AM +0800, Dan Jacobson wrote: > > Bet there are also few developers over 45 years old. > > Probably 99% young, male. > > Bet there is no web page with developer age demographics. > > Anyway, at 45 things get fuzzy, at least for me, so I admire > > those older developers. > > Please post non-development issues to debian-project next time. > > Or rethink whether your issue needs posting at all. This is Jidanni you're talking to. -- - | ,''`.Stephen Gran | | : :' :[EMAIL PROTECTED] | | `. `'Debian user, admin, and developer | |`- http://www.debian.org | - signature.asc Description: Digital signature
Re: We want to honestly apoligize to the whole Debian Community
We also want to apoligize specifically to Mark Shuttleworth whoes name we forged in a post. Accepted, no harm done. Elmo is a MUCH better mail forger, though, you could get some useful tips there. And maybe, one day, a statue. Mark signature.asc Description: OpenPGP digital signature
Bug#358558: ITP: byzanz -- Byzanz is a desktop recorder for creating screencasts
Package: wnpp Severity: wishlist Owner: Bastian Venthur <[EMAIL PROTECTED]> * Package name: byzanz Version : 0.1.0 Upstream Author : Benjamin Otte <[EMAIL PROTECTED]> * URL : http://people.freedesktop.org/~company/byzanz * License : GPL Description : Byzanz is a desktop recorder for creating screencasts Byzanz is a Gnome panel applet allowing you to record your desktop session to an animated GIF file. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable'), (1, 'experimental') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.16-1-k7 Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#358003: ITP: ttf-dzongkha -- TrueType fonts for Dzongkha language
On Tue, Mar 21, 2006 at 07:08:02AM +0100, Christian Perrier wrote: > Quoting Drew Parsons ([EMAIL PROTECTED]): >>> This package includes fonts that are suitable for the display of >>> the Dzongkha language. >> May I suggest for rare languages like this that you also mention in >> the description where the language comes from? (Bhutan in this >> case, isn't it?) > Well, I have one very little argument against doing so: why do it > for Dzongkha and why not do it for, say, French...:-) Because "French" is the adjective in English (the language the package description is written in) for "from France". The same, I would not expect it to be done if the language were called "Bhutanese". -- Lionel -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Уникальные семинары!
On Wednesday 22 March 2006 09:45, Henning Makholm wrote: > Listmasters have been trying to > identify the responsible subscriber with no luck Why not just 500 all posts from sites known to use challenge-response? -- Paul Johnson Email and IM (XMPP & Google Talk): [EMAIL PROTECTED] Jabber: Because it's time to move forward http://ursine.ca/Ursine:Jabber -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: We want to honestly apoligize to the whole Debian Community
On Wednesday 22 March 2006 15:02, Gregg Coner wrote: > Phase 1: Gnome is released and you spend the first entire month doing all > the crazy things to the package that should only be done in experimental > with a goal that at the end of the first month the package will be in > Unstable. Sounds great, so long as that "you" is self-referencing and doesn't expect the rest of the group to take such veiled demands seriously, especially after your forgery stunt... -- Paul Johnson Email and IM (XMPP & Google Talk): [EMAIL PROTECTED] Jabber: Because it's time to move forward http://ursine.ca/Ursine:Jabber -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
