The evil build methods of thy libev
Hi developers, I am a non-DD member of the Debian Perl Group. In the process of packaging the libev-perl package, an event loop for Perl, I stumbled upon the fact that it does not link against libev as a shared object, but instead builds it into its own binary. libev has a "feature" that allows you to build it differently - changing some structs to add customized fields - via a #define. Essentially, every included libev is different than the other ones, therefore you can't simply get rid of this by switching to dynamic linking. I e-mailed upstream about this, but he basically considers this a feature and, remembering of the OpenSSL fiasco, warned not to fiddle in upstream code. As per Policy 3.8.0, packages should not include copies of other packages. I think the best way would be to fix the library and provide another way to include additional fields into its data structures, using the classic approach of a pointer to userdata. However, that would require us to patch any application using libev this way. The second option I came up with was creating a libev-source package from the existing libev package and build packages including libev against this one. Bugs in libev would then only require a rebuild of the dependent packages instead of source package fixing. I think this solution is suboptimal but less invasive. AFAIK the only affected package currently in Debian is rxvt-unicode (see #512487). From the README of libev: Examples of programs that embed libev: the EV perl module, rxvt-unicode, gvpe (GNU Virtual Private Ethernet), the Deliantra MMORPG server (http://www.deliantra.net/), Rubinius (a next-generation Ruby VM), the Ebb web server, the Rev event toolkit. I have not checked whether the mentioned software includes libev or builds against the dynamic library properly. Regards, Maximilian Gaß signature.asc Description: Digital signature
Re: The evil build methods of thy libev
Maximilian Gaß wrote: >Examples of programs that embed libev: the EV perl module, >rxvt-unicode, gvpe (GNU Virtual Private Ethernet), the Deliantra MMORPG >server (http://www.deliantra.net/), Rubinius (a next-generation Ruby >VM), the Ebb web server, the Rev event toolkit. > > I have not checked whether the mentioned software includes libev or builds > against the dynamic library properly. Looking at rxvt-unicodes buildlog, it's the former. Grüße/Regards, René -- .''`. René Engelhard -- Debian GNU/Linux Developer : :' : http://www.debian.org | http://people.debian.org/~rene/ `. `' r...@debian.org | GnuPG-Key ID: 248AEB73 `- Fingerprint: 41FA F208 28D4 7CA5 19BB 7AD9 F859 90B0 248A EB73 -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Hosting the Debian/kCygwin port?
Hi. Sjors Gielen wrote: > Now I'm wondering where to host this project. I've been thinking about > three locations: Sourceforge, Debian or Cygwin. I've filed a project > takeover request for Sourceforge, but the original project admin seems > to work against me a little and it doesn't seem "fit" to release > there. debian-cygwin.sf.net is a Debian GNU/w32 port and I see no reason to remove the project page on Sourceforge. Although there are no releases, there has been some work done (see http://lists.debian.org/debian-win32/) and this site should IMHO at least reside for historical reasons. Maybe there is a chance to reanimate this project or at least reuse some of the work that has been done or concepts that have been developed. You called your port Debian/kCygwin, so why don't you prefer debian-kcygwin instead of debian-cygwin as Sourceforge project name? The former is still available. A few comments about the name: - Debian GNU/kCygwin seems to fit better since you use a GNU userland and Debian people tend to share the FSF view about naming operating systems. - The "k" means normally "kernel of". The idea is probably born because Debian GNU/FreeBSD was already taken by a port that used the libc from FreeBSD. Later IIRC Robert Milan had the idea to use the GNU libc for a Debian FreeBSD Port and only the kernel from FreeBSD. This makes porting single packages a lot easier but requires porting glibc first (the FreeBSD specific diff against glibc is IIRC about 1 MB). > Since the largest part of the project would be the packages in their > new Debian source and binary forms, I at least need an apt repository. > The Debian project already has the structure for that ... You can create the repository yourself using eg. reprepro until your port is more integrated into the Debian infrastructure. I don't think that you will get the permission to host your port on debian-ports.org and thus make this port semi-official before you prove that there has been a significant amount of work done since you are yet unknown among the Debian developers and currently no Debian developer is working with you on this. Debian operates an own Sourceforge clone but this is hosted on a single server with a relatively small harddisk, so this is not an option for hosting. I would first host it on Sourceforge using a not already taken project name and after you are able to show first results try to move it to debian-ports.org. Chances that it will be hosted on debian-ports.org and be integrated in the whole Debian infrastructure are way better when you are able to provide some working packages, at least "required", "important" and a subset of "standard" and "optional" (I think about things like perl and build-essential including its dependencies as a useful subset of "standard" and "optional"). The reasons for the relatively few responses to your mail are probably that most Debian people are not that interested in Windows in general and that you did not provide pointers to the work you have done until now. When you look at the subscriber count of debian-win32 you see that there are at least some people that are interested in such things (it has even more subscribers than debian-qa or debian-www). Such a big project is easier to do when you work together with other people, a more detailed mail with pointers to code and/or specifications to debian-wi...@lists.debian.org might be a good way to find those. I wish you much fun and success with your port Carsten -- Please CC me, I don't read this list regularly. -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Hosting the Debian/kCygwin port?
Sjors Gielen wrote: > ... but the Cygwin packages are different from their Debian > counterparts (think patches), and I'm not sure how that happens with > other ports. debian-devel, is this a problem? No, this is no problem. Debian GNU/kFreeBSD also uses additional patches. Carsten -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: The evil build methods of thy libev
On Wed, Jan 21, 2009 at 02:35:15AM -0800, Steve Langasek wrote: > However, there are zero packages in Debian that use libev. Like I said, rxvt-unicode is using libev. > So I think this is the correct solution - first fix the libev package > (in cooperation with upstream) to not require build-time > customization, then work with the upstreams of packages currently > embedding libev to support building against the shared version. It is the correct solution, but I doubt that it's a possible one (upstream cooperation). We can try, however. ;-) signature.asc Description: Digital signature
Re: Hosting the Debian/kCygwin port?
Hello,
Just a comment about the debian-win32 project: I can read from the list
archive that the main issue was that the VFS doesn't allow to remove or
to overwrite through mv a file in use. IIRC that issue got fixed not so
many months ago in cygwin, and I can confirm that it now seems to work:
$ cat test.c
#include
int main(void) {
int fd = open("blip", O_RDONLY);
pause();
}
$ gcc test.c -o test
$ touch blip
$ ./test &
$ rm blip
$ touch blip
$ ./test &
$ touch blop
$ mv blop blip
So maybe debian-win32 could just be awaken since the barring issue seems gone?
Samuel
--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: The evil build methods of thy libev
On Wed, Jan 21, 2009 at 10:15:43AM +0100, Maximilian Gaß wrote: > I e-mailed upstream about this, but he basically considers this a feature > and, remembering of the OpenSSL fiasco, warned not to fiddle in upstream > code. Wow, nice upstream. Maybe you should reconsider whether this code belongs in Debian at all; we have enough hostile upstreams without seeking them out. > As per Policy 3.8.0, packages should not include copies of other packages. > I think the best way would be to fix the library and provide another way > to include additional fields into its data structures, using the classic > approach of a pointer to userdata. However, that would require us to patch > any application using libev this way. However, there are zero packages in Debian that use libev. So I think this is the correct solution - first fix the libev package (in cooperation with upstream) to not require build-time customization, then work with the upstreams of packages currently embedding libev to support building against the shared version. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developerhttp://www.debian.org/ slanga...@ubuntu.com vor...@debian.org -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Hosting the Debian/kCygwin port?
[removing cyg...@cygwin.org] On Wed, Jan 21, 2009 at 11:14:24AM +0100, Samuel Thibault wrote: > So maybe debian-win32 could just be awaken since the barring issue > seems gone? This is what I would try first. I hope you are successful with this. Carsten -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: The evil build methods of thy libev
At 1232534115 time_t, Steve Langasek wrote: > However, there are zero packages in Debian that use libev. awesome does, and by linking against it (and not embedding it). -- Julien Danjou .''`. Debian Developer : :' : http://julien.danjou.info `. `' http://people.debian.org/~acid `- 9A0D 5FD9 EB42 22F6 8974 C95C A462 B51E C2FE E5CD signature.asc Description: Digital signature
Re: The evil build methods of thy libev
On Wed, Jan 21, 2009 at 02:35:15AM -0800, Steve Langasek wrote: > Wow, nice upstream. Maybe you should reconsider whether this code > belongs > in Debian at all; we have enough hostile upstreams without seeking > them out. in the interest of good cooperation with upstream, we should keep in mind that there are a handfull of sometimes conflicting interests, and while making every library a shared one is one of our concerns as a distribution, it might only rank second for other people. that's not exactly "hostile". in the special case of libev there also might be another way: you could use libev's feature to embed a "user data" void* and create a thin wrapper around this libev that can be used as a shared library. this way the user of the library can choose whether she wants to use the shared library version with user data pointers, or use the way envisioned by libev upstream for that extra 0.0001% performance. this might even be a solution that is acceptable to upstream... cu robert ps: i like libev, it's awesome! -- Robert Lemmen http://www.semistable.com signature.asc Description: Digital signature
Name Removal from your website
Hi There,How are you ?I just discovered my name in 2 links of your website.I am not even aware what those things are about and it is falsely posted ??Links :lists.debian.org/debian-devel/2004/01/msg01948.html lists.debian.org/debian-devel/2004/03/msg00091.htmlI would like to request you to PLEASE CLEAR MY NAME FROM THOSE LINKS ASAP.I am falsely named by someone I dont even know.I never looked up my name on google until today and was really shocked.Its really upsetting and embarrassing.Any help would be held in highest regard.I appreciate your help.Regards,Raghav Dayma _ It's simple! Sell your car for just $40 at CarPoint.com.au http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fsecure%2Dau%2Eimrworldwide%2Ecom%2Fcgi%2Dbin%2Fa%2Fci%5F450304%2Fet%5F2%2Fcg%5F801459%2Fpi%5F1004813%2Fai%5F859641&_t=762955845&_r=tig_OCT07&_m=EXT
Name Removal from your website
Hi There,How are you ?I just discovered my name in 2 links of your website.I am not even aware what those things are about and it is falsely posted ??Links :lists.debian.org/debian-devel/2004/01/msg01948.html lists.debian.org/debian-devel/2004/03/msg00091.htmlI would like to request you to PLEASE CLEAR MY NAME FROM THOSE LINKS ASAP.I am falsely named by someone I dont even know.I never looked up my name on google until today and was really shocked.Its really upsetting and embarrassing.Any help would be held in highest regard.I appreciate your help.Regards,Raghav Dayma _ Net yourself a bargain. Find great deals on eBay. http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Frover%2Eebay%2Ecom%2Frover%2F1%2F705%2D10129%2D5668%2D323%2F4%3Fid%3D10&_t=763807330&_r=hotmailTAGLINES&_m=EXT
Name Removal from your website
Hi There,How are you ?I just discovered my name in 2 links of your website.I am not even aware what those things are about and it is falsely posted ??Links :lists.debian.org/debian-devel/2004/01/msg01948.html lists.debian.org/debian-devel/2004/03/msg00091.htmlI would like to request you to PLEASE CLEAR MY NAME FROM THOSE LINKS ASAP.I am falsely named by someone I dont even know.I never looked up my name on google until today and was really shocked.Its really upsetting and embarrassing.Any help would be held in highest regard.I appreciate your help.Regards,Raghav DaymaSell your car for just $40 at CarPoint.com.au It's simple! _ Time for change? Find your ideal job with SEEK. http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau%2F%3Ftracking%3Dsk%3Atl%3Ask%3Anine%3A0%3Ahottag%3Achange&_t=757263783&_r=SEEK_tagline&_m=EXT
Re: Hosting the Debian/kCygwin port?
Hey, Thanks for all your answers! Carsten Hey wrote: On Wed, Jan 21, 2009 at 11:14:24AM +0100, Samuel Thibault wrote: So maybe debian-win32 could just be awaken since the barring issue seems gone? This is what I would try first. I hope you are successful with this. Carsten I was using Cygwin 1.5 when I started working on this. I indeed noticed that in Cygwin CVS this has been fixed already. Carsten Hey wrote: > debian-cygwin.sf.net is a Debian GNU/w32 port and I see no reason to > remove the project page on Sourceforge. Although there are no > releases, there has been some work done (see > http://lists.debian.org/debian-win32/) > and this site should IMHO at least reside for historical reasons. > Maybe there is a chance to reanimate this project or at least reuse > some of the work that has been done or concepts that have been > developed. I started with the idea when I noticed this project. It had been dead for a few years, and I liked the idea and looked into it. I had some contact with the original admin. He added me to the project, created an SVN repository on my request, but refused to make me an admin, which is needed to i.e. modify the project web frontend. When I filed a project takeover, the admin removed my SVN commit access. Everything currently in SVN is mine (and that's not much, but I've got a lot more waiting here, until I can commit it somewhere.) CVS only contains the same old dpkg sources twice, and some binary builds. I'm still waiting for the admin to either accept or reject the project takeover request. I don't have a lot of hope there, he seems to be working against me. I also e-mailed the debian-ports admin, and he said the same as you: there needs to be a base system before it can be hosted there. Base system means gcc, g++, dpkg-dev, etc. For now, I will keep working on the project locally, until I can host everything I have now at Sourceforge. Once I have a base system going, I'll e-mail the debian-ports admin again. Again, thanks for your replies! Sjors -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Name Removal from your website
Raghav Dayma schreef: Hi There, How are you ? I just discovered my name in 2 links of your website. I am not even aware what those things are about and it is falsely posted ?? Links : lists.debian.org/debian-devel/2004/01/msg01948.html lists.debian.org/debian-devel/2004/03/msg00091.html I would like to request you to PLEASE CLEAR MY NAME FROM THOSE LINKS ASAP. I am falsely named by someone I dont even know. I never looked up my name on google until today and was really shocked. Its really upsetting and embarrassing. Any help would be held in highest regard. I appreciate your help. Regards, Raghav Dayma You do not have to write to all Debian lists when something like this happens. Turns out your name was simply used in some five year old spam e-mail. A spam source somehow found your name, and spammed the debian-devel list. An archive of this list is held at lists.debian.org, and that is what you are seeing (note the date). I'm not sure if archived messages can or will be removed. In any case, this name is not linked to you in any way as far as I know. It's just the fact that your name seems to be quite unique. Sjors -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#512518: ITP: jbig2enc -- a JBIG2 encoder
Package: wnpp Severity: wishlist Owner: Jakub Wilk * Package name: jbig2enc Version : 0.27 Upstream Author : Adam Langley * URL : http://www.imperialviolet.org/jbig2.html * License : Apache License, Version 2.0 Programming Lang: C Description : a JBIG2 encoder JBIG2 encodes bitonal (1 bpp) images using a number of clever tricks to get better compression than G4. This encoder can: * generate JBIG2 files, or fragments for embedding in PDFs, * generic region encoding, * perform symbol extraction, classification and text region coding, * perform refinement coding and, * compress multipage documents. -- System Information: Debian Release: 5.0 APT prefers unstable APT policy: (900, 'unstable'), (500, 'experimental') Architecture: i386 (i686) -- Jakub Wilk -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#512519: ITP: grid-packaging-tools -- system to control dependencies between packages of larger projects
Package: wnpp Severity: wishlist Owner: Steffen Moeller * Package name: grid-packaging-tools Version : 3.2 * URL : http://www.globus.org * License : custom Programming Lang: Perl Description : system to control dependencies between packages of larger projects The grid packaging tools (GPT) is a collection of packaging tools built around an XML based packaging data format. This format provides a straight forward way to define complex dependency and compatibility relationships between packages. The tools provide a means for developers to easily define the packaging data and include them as part of their source code distribution. Binary packages can be automatically generated from this data. The packages defined by GPT are compatible with other packages and can be easily converted. . GPT is a key tool for the build process of the Globus grid middleware. -- System Information: Debian Release: 5.0 APT prefers unstable APT policy: (500, 'unstable'), (500, 'testing'), (500, 'stable') Architecture: amd64 (x86_64) -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#512522: ITP: pct-scanner-scripts -- easy scan and create merged djvu and pdf documents for archiving
Package: wnpp Severity: wishlist Owner: Jelle de Jong Package is part of the possible upcomming pct-desktop-environment, I want to start getting my stable packages into the debian repository. Release often, release soon :-D. Request for sponsering will come to. Thanks in advance, Jelle de Jong Package name: pct-scanner-scripts Version : 0.0.2 Upstream Author : Jelle de Jong URL : https://secure.powercraft.nl/svn/packages/trunk/deb/pct-scanner-scripts/ License : (GPLv3) Programming Lang: (BASH) Description : easy scan and create merged djvu and pdf documents for archiving This package contains a script and configuration system to easy scan documents and create a digital archiving. It can create djvu pdf and jpg outputs, in merged and single files. It also has a configuration file system where different configurations can be defined, this makes it easy to switching between configurations. The script is designed for productivity, and the package is maintained by PowerCraft Technology. Take a look at the --help argument for all the useful and simple options of this tool. This program is a helper script to easy scan and automated digital document archiving for normal and adf scanners. You can debug this script by running it as bash -x pct-scanner-script [options]. The script is just a nice wrapper for the scanadf, djvu and pdf tools -- System Information: Debian Release: 5.0 APT prefers unstable APT policy: (990, 'unstable'), (50, 'experimental') Architecture: i386 (i686) -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#512526: ITP: gmm++ -- generic C++ template library for sparse, dense and skyline matrices
Package: wnpp Owner: Ana Guerrero Severity: wishlist * Package name: gmm++ Version : 3.1 Upstream Author : Yves Renard and Julien Pommier * URL : http://home.gna.org/getfem/gmm_intro.html * License : LGPL v2.1 or later Programming Lang: C++ Description : generic C++ template library for sparse, dense and skyline matrices GMM++ is a framework of pre-defined methods for matrix computation. It is built as a set of generic algorithms for any interfaced vector type or matrix type. . It can be viewed as a glue library allowing cooperation between several vector and matrix types. However, basic sparse, dense and skyline matrix/vector types are built-in, hence it can be used as a standalone linear algebra library. Notice that there is already an old ITP for getfem++ that includes gmm++ [1], but in the meanwhile it get packaged or not, gmm++ is being packaged standalone inside the Debian Krap Team (Qt/KDE Team) given it is needed by kdeedu (step) and koffice2. If somebody into this kind of stuff wants to take care of getfem++/gmm++ is totally welcome. [1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=453065 Ana -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: The evil build methods of thy libev
Maximilian Gaß dijo [Wed, Jan 21, 2009 at 10:15:43AM +0100]: > Hi developers, > > I am a non-DD member of the Debian Perl Group. In the process of packaging the > libev-perl package, an event loop for Perl, I stumbled upon the fact that it > does not link against libev as a shared object, but instead builds it into its > own binary. > > libev has a "feature" that allows you to build it differently - changing some > structs to add customized fields - via a #define. Essentially, every included > libev is different than the other ones, therefore you can't simply get rid of > this by switching to dynamic linking. > (...) Humh... This seems somewhat similar to the situation we faced ~1 year ago which led to not being able to package libmath-pari-perl and libnet-ssh-perl-perl (yes, that's the name). For further details, please see #439339, #440527. I exchanged some private mails with Ilya Zakharevich and with Bill Alombert in this regard (I might still have them somewhere, if you want to see any further reasoning)... Sadly, the only answer was to withdraw the ITPs and forget about the issue :-( -- Gunnar Wolf - gw...@gwolf.org - (+52-55)5623-0154 / 1451-2244 PGP key 1024D/8BB527AF 2001-10-23 Fingerprint: 0C79 D2D1 2C4E 9CE4 5973 F800 D80E F35A 8BB5 27AF -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: New package: gpgtools
On Tue, 20 Jan 2009, Franck Joncourt wrote: > I packaged gpgdir: > Description: recursive directory encryption tool using GnuPG > and Jari Aalto gpgwrap: > Description: a passphrase wrapper for GNU privacy guard (gpg) > Does anyone see any objection with this new package, have a better name? > As we have the *signing party*, I do not want to make too much mess. > I think gpgtools is a good idea since the rejected packages does not fit > very well in signing party. These tools should be packaged along with the other pgp related tools, in signing-party. -- | .''`. ** Debian GNU/Linux ** Peter Palfrader | : :' : The universal http://www.palfrader.org/ | `. `' Operating System | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Hosting the Debian/kCygwin port?
On Wed, Jan 21, 2009 at 02:24:15PM +0100, Sjors Gielen wrote: > Tzafrir Cohen schreef: >> On Tue, Jan 20, 2009 at 12:49:29AM +0100, Sjors Gielen wrote: >>>I'm working on a project porting the Debian tools to Cygwin. >>>Currently, my plan is to: - Provide some required patches to the Cygwin >>>team, for example to allow a file in use to be removed or modified, >>>since this is required for dpkg - Re-compile all packages and patches >>>currently available for Cygwin into a Debian .deb - Provide either a >>>way to bootstrap Debian onto an existing Cygwin installation, or >>>provide installers for a pure Debian/kCygwin installation. >>Which libc do you intend to use? Cygwin uses newc, IIRC, but I can't >>find it packaged in Debian so far. > >As far as I understand it now, newlib is packed along with Cygwin >itself in cygwin1.dll. I'm thinking of a package libcygwin1, which >will contain this (and maybe other) dlls. This package will just be >created in the process of getting Debian to work on Cygwin :) I think that no one answered your original question. We are not interested in hosting your project at sourceware.org and discussion of it in this mailing list is not really on-topic for the Cywgin list. The Cygwin mailing list is for discussion about the cygwin release that is hosted at the cygwin web site. Feel free to send a follow-up message with a pointer to the mailing list where you will be continuing any discussion about this topic. I would appreciate it if further discussion was directed there. cgf -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Fwd: dicomscope: open tasks
Hi there,
I am working on the dicomscope package, and I was told there is an
issue with my package:
...
> E: dicomscope: no-shlibs-control-file usr/lib/libjInterface.so
> N:
> N: Although the package includes a shared library, the package does
> not
> N: have a shlibs control file. If this is intentional, please override
> N: this error.
> N:
> N: Refer to Policy Manual, section 8.6 for details.
>
> -> Unfortunately I have not the slightest idea what might went wrong
>here. Perhaps we should ask on debian-{devel,mentors} for help ...
Could someone please let me know:
1. What this means ?
2. What am I supposed to do to fix that issue.
thanks !
--
Mathieu
--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Fwd: dicomscope: open tasks
* Mathieu Malaterre [Wed, 21 Jan 2009 17:38:30 +0100]: > Could someone please let me know: > 1. What this means ? > 2. What am I supposed to do to fix that issue. Please re-post this question to debian-ment...@lists.debian.org. -- Adeodato Simó dato at net.com.org.es Debian Developer adeodato at debian.org The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man. -- George Bernard Shaw -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: dicomscope: open tasks
Hi
Dne Wed, 21 Jan 2009 17:38:30 +0100
Mathieu Malaterre napsal(a):
> Hi there,
>
> I am working on the dicomscope package, and I was told there is an
> issue with my package:
>
> ...
> > E: dicomscope: no-shlibs-control-file usr/lib/libjInterface.so
> > N:
> > N: Although the package includes a shared library, the package does
> > not
> > N: have a shlibs control file. If this is intentional, please override
> > N: this error.
> > N:
> > N: Refer to Policy Manual, section 8.6 for details.
> >
> > -> Unfortunately I have not the slightest idea what might went wrong
> >here. Perhaps we should ask on debian-{devel,mentors} for help ...
>
>
> Could someone please let me know:
> 1. What this means ?
> 2. What am I supposed to do to fix that issue.
Reading what policy says on this topic could be good starting point:
http://www.debian.org/doc/debian-policy/ch-sharedlibs.html
--
Michal Čihař | http://cihar.com | http://blog.cihar.com
signature.asc
Description: PGP signature
Should gnome-libs die for Lenny?
Hi folks, There are only 5 packages left that have reverse depends on gnome-libs packages. They are as follows: digitaldj: Removal request was filed but maintainer claims someone has done a Gtk2 port and saved it for now. gbib: Removal was proposed. I mailed upstream and the response was "gbib is no longer maintained or developed". I'm trying to build it now with Gtk2 but not having much luck so far. gsnes9x: Removal requested. It's been listed as dead on SourceForge since 2004. snes9express possibly a Gtk2 replacement? powershell: Removal was requested but maintainer saved it. This one could be a bit of a pain to port. soundtracker: Dead upstream since 2001 but one of the devs started on a Gtk2 port. Currently it fails to build and has bugs according to the website. I've e-mailed but not received a response. Given this and the fact that gnome-libs removal was suggested (and bugs filed) over a year ago, does it make sense to remove gnome-libs before Lenny releases? For more details, I have been trying to keep the wiki page up to date: http://wiki.debian.org/Gtk1.2ImlibGnome1Removals Thanks, Barry deFreese -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#512551: ITP: faucc -- C compiler generating Intel code for 16bit/32bit CPUs
Package: wnpp Severity: wishlist Owner: Stefan Potyra * Package name: faucc Version : 2009 Upstream Author : Volkmar Sieh * URL : http://www.faumachine.org [1] * License : GPL Programming Lang: C Description : C compiler generating Intel code for 16bit/32bit CPUs FAUcc is a optimizing C compiler, that can generate Intel code for 16bit/32bit CPUS. It is useful, if you are in need to compile real mode code, like a BIOS. . FAUcc uses the GNU binutils for assembling/linking. [1]: yet unreleased, but release work is currently going on. Cheers, Stefan. -- System Information: Debian Release: 4.0 APT prefers stable APT policy: (500, 'stable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.24-etchnhalf.1-686 Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Should gnome-libs die for Lenny?
On Wed Jan 21 12:08, Barry deFreese wrote: > There are only 5 packages left that have reverse depends on gnome-libs > packages. They are as follows: > > gbib: Removal was proposed. I mailed upstream and the response was "gbib > is no longer maintained or developed". I'm trying to build it now with > Gtk2 but not having much luck so far. And, it's not in Lenny anyway... Matt -- Matthew Johnson signature.asc Description: Digital signature
Re: Bug#500616: ITP: solid3d -- Software library for collision detection of geometric objects in 3D space
A Dilluns 29 Setembre 2008, Sune Vuorela va escriure: > On Monday 29 September 2008 20:40:24 Leopold Palomo Avellaneda wrote: Package: wnpp Severity: wishlist Owner: Leopold Palomo Avellaneda Package name: solid3d Version : 3.5.6 Upstream Author : Gino van den Bergen URL : http://solid3d.berlios.de License : GPL Programming Lang: C++ Description : Software library for collision detection of geometric objects in 3D space. > > > > Collision detection is the process of detecting pairs of objects that are > > intersecting or are within a given proximity of each other. SOLID > > contains operations for performing intersection tests and proximity > > queries on a wide variety of shape types, including: deformable triangle > > meshes, boxes, ellipsoids, and convex polyhedra. > > Please note that Solid is also a technology in kde and currently > /usr/lib/libsolid* is in use by the kde packaging. > > So you need to be careful to not interfere with that. Ok, looking the situation and seeing that the project has its last free version, we have created a fork, with the name solid3D so I hope that it will not interfere in any other lib. Regards, Leo -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Help bts-link be a more effective tool
On Sun, Jan 18, 2009 at 03:21, Raphael Geissert wrote: > Hi, > > Sandro Tosi wrote: > [...} >> In recent bts-link runs, we noticed some errors. The log is available >> at [2]: please take the time to give it a look, search for your >> packages and check the situation. There are errors in that log that >> might be ok, but others can refer to broken links, no more active >> BTSes or any other possible situation. > > What about generating a nice report page and making the PTS warn when bts-link > had troubles handling forwarded bugs of a given package. That would let > maintainers know that there's something wrong without having to find out where > in d.o is the error log of bts-link ;) That's a nice idea: for now I only added a TODO item (so anyone can contribute patches :) ). Cheers, -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: New package: gpgtools
Peter Palfrader wrote: > On Tue, 20 Jan 2009, Franck Joncourt wrote: > >> I packaged gpgdir: >> Description: recursive directory encryption tool using GnuPG > >> and Jari Aalto gpgwrap: >> Description: a passphrase wrapper for GNU privacy guard (gpg) > >> Does anyone see any objection with this new package, have a better name? >> As we have the *signing party*, I do not want to make too much mess. >> I think gpgtools is a good idea since the rejected packages does not fit >> very well in signing party. > > These tools should be packaged along with the other pgp related tools, > in signing-party. Ok. I check that :) Thanks, -- Franck Joncourt http://debian.org - http://smhteam.info/wiki/ signature.asc Description: OpenPGP digital signature
Re: Help bts-link be a more effective tool
On Sun, Jan 18, 2009 at 10:12, Emilio Pozuelo Monfort wrote: >> E: pkg=update-manager, bug=415376, msg=Parse error: > [https://bugs.launchpad.net/update-manager/+bug/95168/] No product specified > but > bug affects multiple products > > That looks wrong to me. The error says "No product specified", but there's a > product specified (update-manager) in the URL! If there wasn't one, it would > look like "https://bugs.launchpad.net/bugs/95168";. I could re-forward it to > https://bugs.launchpad.net/ubuntu/+source/update-manager/95168 to maybe > work-around it, but I think this is a bts-link bug. Ah, it can be :) >> Python Applications Packaging Team >>mercurial > > Looks like they're not using Trac anymore... About this: Vincent/Vernon what about mercurial upstream BTS? is it in some known format we can parse? or should we remove bts-link support for mercurial? (you can keep teh reply to bts-link-devel only, I think) Cheers, -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Help bts-link be a more effective tool
On Sun, Jan 18, 2009 at 11:20, Frank Küster wrote: > "Paul Wise" wrote: >> Debian TeX Maintainers >>texlive-bin > > One was a typo in the upstream address. The other one, Thanks for fixing this. > E: pkg=texlive-bin, bug=351672, msg=Does not deals dupes: > [http://sourceforge.net/tracker/index.php?func=detail&aid=1425283&group_id=23164&atid=377580] > > I do not understand. The upstream item exists and has the state Closed, > with the comment that there was already an upstream request for this > which had been closed already. > > What should we do about that? The fact here is that "Resolution: Duplicate" is set, so we are saying "we don't know what other bug report to look at to decide about this one". Forwarding to the "main" bug should fix the situation (sometimes, dups bugs are closed even if the "other" bug is still under invetigation/fis was not released yet, that's why don't check its status). Cheers, -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Help bts-link be a more effective tool
On Sun, Jan 18, 2009 at 19:54, Bastien ROUCARIES wrote: > On Sat, Jan 17, 2009 at 8:36 PM, Sandro Tosi wrote: >> Hello, > >> If you feel something is missing, should be fixed or enhanced, let >> us[4] know; of course, patches are welcome ;) (git repo at [5]). > > I really useful stuff will be to use user tag in order to crossref > another distrib bugzilla. For instance some bug are fixed on redhat > like #506180 but not upstream. > It will allow to automatize retrieval of information. Of course it would help, but how to implement it? do you have an idea yet? Cheers, -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Help bts-link be a more effective tool
Hello Raphael, On Mon, Jan 19, 2009 at 09:17, Raphael Hertzog wrote: > On Sat, 17 Jan 2009, Sandro Tosi wrote: >> In recent bts-link runs, we noticed some errors. The log is available >> at [2]: please take the time to give it a look, search for your >> packages and check the situation. There are errors in that log that >> might be ok, but others can refer to broken links, no more active >> BTSes or any other possible situation. > > E: pkg=acpi-support, bug=388160, msg=Parse error: > [https://bugs.launchpad.net/ubuntu/+source/acpi-support/+bug/18864] No > task affecting product acpi-support > > The URL given is fine… you should really improve bts-link understanding of > Launchpad. The bug affects acpi-support inside Ubuntu and is certainly a > valid way for us to track a bug. We'll try to do it. > Before you decide to push out errors to maintainers via PTS (as I've seen > mentionned), you should really improve the tool so that the only remaining > errors are really user errors. That's exactly what this call for help was about, getting input from you about what to improve in bts-link. -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
ASP Unleashed 09 _ Synapse 2009(DAIICT)
Dear friend, Welcome to ASP Unleashed '09_Synapse '09 . ASP Unleashed is aimed at harnessing the potential of Web 2.0 by making use of ASP.Net. The event is aimed at popularizing the use of ASP.NET and other Microsoft technologies among the participants. This mail is to remind you that End of Phase I (submission of synopsis document) is Jan 25th '09. So, register as soon as possible. If you have any queries regarding ASP Unleashed , feel free to contact us:- (asp.unleashe...@gmail.com or 09714623064 ). For further information visit http://synapse.daiict.ac.in/aspun.html Regards, Team ASP Unleashed Synapse '09 Dhirubhai Ambani Int. Of Info. & Comm. Tech.(DA-IICT) Gandhinagar, Gujarat 382007 -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Help bts-link be a more effective tool
On Mon, Jan 19, 2009 at 09:46, Stefano Zacchiroli wrote: > On Mon, Jan 19, 2009 at 09:17:06AM +0100, Raphael Hertzog wrote: >> Before you decide to push out errors to maintainers via PTS (as I've seen >> mentionned), you should really improve the tool so that the only remaining >> errors are really user errors. > > BTW, when that is done, please submit a wishlist bugreport on the PTS > requesting integration, *together* with a description of the parsing > rules of the error output of bts-link OR maybe a switch to a format > which can be parsed out of the box and is extensible, e.g., yaml. Thanks Zack for the input, we'll evaluate yaml (or some other format). > The PTS is being polluted by tons of micro-parsers for the output of > the tools it monitor, a convergence at least on the surface syntaxes > would be nice to ease future maintenance. Since we'lll have to implement it from scratch, name one and we'll consider to stick to it. Cheers, -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Help bts-link be a more effective tool
On Tue, Jan 20, 2009 at 01:04, Bastien ROUCARIES wrote: > On Mon, Jan 19, 2009 at 10:50 AM, Bastien ROUCARIES > wrote: >>ll need I suppose cooperation from BTS itself but in a second >> time. We need only two user tags by foreign distrib: >> bts-link-foreign-xref-$distrib set to the foregin bugzilla entry >> bts-link-foreign-status-$distrib magically set by bts-link >> >> In a second time BTS interface could provide link near the forwarded >> link with status in other distrib (with a nice icon describing status) >> >> We will need also a means to put in the package describtion link to >> other distrib bugzilla, and we ease to remove duplicate entry :) > > I downloaded git source but I have some problem. How can I test some > stuff on BTS? I does not include a sandbox package (see bug > 435...@bugs.debian.org) tag 435357 wontfix :) > How can we test improvement? did you clone the git repo and read README? I bet no (at least for the second part): there the command to use to local test. be *careful* when running this on your machine: please run it only if you know what you're doing I hope we'll be able to remove file-based parsing really soon. Cheers, -- Sandro Tosi (aka morph, morpheus, matrixhasu) My website: http://matrixhasu.altervista.org/ Me at Debian: http://wiki.debian.org/SandroTosi -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
ASP Unleashed 09 _ Synapse 2009(DAIICT)
Dear friend, Welcome to ASP Unleashed '09_Synapse '09 . ASP Unleashed is aimed at harnessing the potential of Web 2.0 by making use of ASP.Net. The event is aimed at popularizing the use of ASP.NET and other Microsoft technologies among the participants. This mail is to remind you that End of Phase I (submission of synopsis document) is Jan 25th '09. So, register as soon as possible. If you have any queries regarding ASP Unleashed , feel free to contact us:- (asp.unleashe...@gmail.com or 09714623064 ). For further information visit http://synapse.daiict.ac.in/aspun.html Regards, Team ASP Unleashed Synapse '09 Dhirubhai Ambani Int. Of Info. & Comm. Tech.(DA-IICT) Gandhinagar, Gujarat 382007
Bug#512562: ITP: xsettings-kde -- xsettings daemon for KDE
Package: wnpp Severity: wishlist Owner: "Guido Günther" * Package name: xsettings-kde Version : 0.9 * URL : http://svn.mandriva.com/svn/soft/theme/xsettings-kde/trunk * License : GPL Programming Lang: C Description : XSettings daemon for KDE This package provides a XSettings daemon for K Desktop Environment. It allows XSettings aware applications (all GTK+ 2 and GNOME 2 applications) to be informed instantly of changes in KDE configuration, such as theme name, default font and so on. -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: The evil build methods of thy libev
* Maximilian Gaß: > The second option I came up with was creating a libev-source package from the > existing libev package and build packages including libev against this one. > Bugs in libev would then only require a rebuild of the dependent packages > instead of source package fixing. I think this solution is suboptimal but less > invasive. This is comparable to C++ templates, so it's not *that* bad, it's just rather unusual, and somewhat disappointing for something which is quite likely to transparently pick up kernel API improvements (which it has done with epoll, if the library is old enough). -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: The evil build methods of thy libev
On Wed, Jan 21, 2009 at 09:45:58PM +0100, Florian Weimer wrote: > * Maximilian Gaß: > > > The second option I came up with was creating a libev-source package from > > the > > existing libev package and build packages including libev against this one. > > Bugs in libev would then only require a rebuild of the dependent packages > > instead of source package fixing. I think this solution is suboptimal but > > less > > invasive. > > This is comparable to C++ templates, so it's not *that* bad, it's just > rather unusual, and somewhat disappointing for something which is > quite likely to transparently pick up kernel API improvements (which > it has done with epoll, if the library is old enough). I have gained some more insight into the libev internals. According to a colleague of the library's developer, libev can be heavily customized when embedded, producing entirely different kinds of libraries. Due do this news and the massive inconvience of patching everything, I am currently favoring the libev-source solution. signature.asc Description: Digital signature
Bug#512573: ITP: libnumber-range-perl -- manipulate and test ranges of numbers
Package: wnpp Severity: wishlist Owner: Niko Tyni * Package name: libnumber-range-perl Version : 0.07 Upstream Author : Larry Shatzer, Jr., * URL : http://search.cpan.org/dist/Number-Range/ * License : GPL-1+ | Artistic Programming Lang: Perl Description : manipulate and test ranges of numbers Number::Range will take a description of a range, and then allow you to test if a number falls within the range. You can also add and delete from the range. Quoting http://cpanratings.perl.org/dist/Number-Range : This module does a specific job, does it well and comes with clear, succinct documentation. This package will be maintained by the pkg-perl group. -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Bug#512433: ITP: r-cran-vcd -- GNU R Visualizing Categorical Data
On Tue, 20 Jan 2009, Dirk Eddelbuettel wrote: The Policy Draft you reference is somewhat outdated and in need of a refresher. I've thought this because it is quite old from the tome stamp but I failwd to found something more recent. As for the names: On a few of my more recent ITPs for R / CRAN packages, folks suggested to not use the 'short' names. Hence I would suggest r-cran-msm r-cran-sp r-cran-spc r-cran-vcd for binary _and source_ packages and you may as well stick with r-cran-colorspace Fine, I will regard this for these packages. WHat would you suggest for the recently uploaded package plotrix which is currently in new? should I immediately ask ftpmaster to drop this upload and rename the source package as well? as colorspace is so generic. I completely agree. I just added the explicite hint to my ITPs (if I did not forgot it) to ask for comments on the naming scheme because I was not really convinced that it is the best idea these days (at the time of writing it was probably reasonable). Kind regards Andreas. -- http://fam-tille.de -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Bug#512433: ITP: r-cran-vcd -- GNU R Visualizing Categorical Data
On 21 January 2009 at 23:34, Andreas Tille wrote: | On Tue, 20 Jan 2009, Dirk Eddelbuettel wrote: | | > The Policy Draft you reference is somewhat outdated and in need of a | > refresher. | | I've thought this because it is quite old from the tome stamp but | I failwd to found something more recent. | | > As for the names: On a few of my more recent ITPs for R / CRAN | > packages, folks suggested to not use the 'short' names. Hence I would | > suggest | > | > r-cran-msm | > r-cran-sp | > r-cran-spc | > r-cran-vcd | > | > for binary _and source_ packages and you may as well stick with | > | > r-cran-colorspace | | Fine, I will regard this for these packages. WHat would you | suggest for the recently uploaded package plotrix which is currently | in new? should I immediately ask ftpmaster to drop this upload | and rename the source package as well? Whichever way you see fit. I myself have not been entirely consistent. Packages that were likely to clash (having two or three letter names) I upload as r-cran-$foo, but my own Rcpp went as source rcpp and binary r-cran-rcpp. | > as colorspace is so generic. | | I completely agree. I just added the explicite hint to my | ITPs (if I did not forgot it) to ask for comments on the | naming scheme because I was not really convinced that it is | the best idea these days (at the time of writing it was probably | reasonable). Yes, better safe than sorry. Probably better to keep the 'package name space' separated. Dirk -- Three out of two people have difficulties with fractions. -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
