Re: Packaging InVesalius
Hi Thiago, On Sun, Jan 27, 2013 at 05:59:45PM -0200, Thiago Franco de Moraes wrote: > > Now the rename is working. He is renaming the tarball to > invesalius_3.0~b3.orig.tar.gz. To make it work I used uversionmangle instead > of dversionmangle. The folder inside this tarball has the name > "invesalius3-3.0-b3" instead of "invesalius_3.0~b3", is this a problem? The folder name inside the tarball is no problem. If we do not repack the tarball we just take what is inside (sometimes there are quite strange choices of names - we simply leave this as is.) > The Lintian returns two warnings: > > W: invesalius: extra-license-file usr/share/doc/invesalius/LICENSE.txt.gz > > It's because the package has two license files, one in English and other in > Portuguese. The Portuguese version is because a license must be in Portuguese > to be recognized by the Brazilian government. That's a fair reason. I would suggest to rename it LICENSE.pt.txt (or LICENSE.pt.BR.txt at your preference) and to not compress this file (which can be done by override_dh_compress (you will find examples when greping for this string in SVN.) The debian/copyright file must not be compressed and I would consider the same reason for any translated license statement. You can create a lintian-overrides file to suppress the warning and also add a comment for the reason you have given above. > W: invesalius: binary-without-manpage usr/bin/invesalius-3.0 > If you think > it's necessary, I can create a manpage to InVesalius. This would be great. Kind regards and thanks for your work on this Andreas. -- http://fam-tille.de -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130128081722.ga9...@an3as.eu
Re: Flexbar source code?
Hi Andreas, it is completely clear to me that one should use as many packaged libraries as possible. I'm simply not sure if it works with the older SeqAn package apart from modifications. I would have to test this first, and for reasons of consistency I would then go back to this release also on sourceforge. Since I know that a lot changes were introduced in the recent year, especially in the align module and the argument parser, this could cause problems and I also prefer to rely on the recent implementations. Therefore, I propose to wait for the next SeqAn package release. I will then use it also for the Flexbar version on sourceforge and will try to come along without modifications. Since I applied only very slight modifications, I am optimistic that it works out and I will discuss it with SeqAn developers. I definitely prefer to use an official and unmodified release, and see it as an aim for coming versions. Besides from cleaner sources and packaging, it would then also become obsolete to adjust new SeqAn libs to modifications for inclusion in Flexbar. However, I am not sure how fast a new packaged release will be provided. Best regards Johannes On Jan 25, 2013, at 11:08 PM, Andreas Tille wrote: > Hi Johannes, > > many thanks for getting involved. > > On Fri, Jan 25, 2013 at 03:11:23PM +0100, Johannes Röhr wrote: >> Tony asked me to join discussion on Flexbar packaging. I'm not sure if the >> stable SeqAn 1.3.1 package release misses functionality that I rely on, or >> if the behaviour of functions changed even if it compiles with this release. >> For example the new ArgumentParser has been added rather recently and I >> would not be surprised if its functionality changed in the time between the >> SeqAn package release and the revision that I incorporated. >> >> However, I think the question is not only if it compiles somehow using this >> package. It should be exacly the same behaviour for one version of Flexbar. >> This isn't clear if the seqan sources differ in version, not to talk about >> slight modifications I made. > > Thanks for the clarifications. I'd take this for some veto to not use > the Debian packaged seqan library. The question is whether the > modifications you made might be incorporated into upstream seqan. If > this would be feasible we could package a seqan version that would fit > flexbar (and most probably other applications.) Do you think this is > possible? > >> Thank you for the interest to include Flexbar! > > Sure. We try to fit the needs of all bioinformatics tasks. > > Kind regards and thanks again for joining here > >Andreas. > > -- > http://fam-tille.de -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/e1de0773-19d4-48db-ad6a-e66e83a39...@users.sourceforge.net
Re: Flexbar source code?
Hi Johannes, thanks again for your very helpful clarification. After reading this I think it might be the best idea if we try to package flexbar as it is (including seqan copy) for the experimental branch in Debian which helps Toni for the moment and once there might be some new seqan library release we will build flexbar using this new release and move it to main Debian. I guess for the moment this helps those who need flexbar now and is technically OK. Toni, do you consider it reasonable if we do this together at the Debian Med Sprint in Kiel? Johannes, thanks again for sharing your insight and also for providing flexbar as free software Andreas. On Mon, Jan 28, 2013 at 04:09:08PM +0100, Johannes Röhr wrote: > Hi Andreas, > > it is completely clear to me that one should use as many packaged libraries > as possible. I'm simply not sure if it works with the older SeqAn package > apart from modifications. I would have to test this first, and for reasons of > consistency I would then go back to this release also on sourceforge. Since I > know that a lot changes were introduced in the recent year, especially in the > align module and the argument parser, this could cause problems and I also > prefer to rely on the recent implementations. > > Therefore, I propose to wait for the next SeqAn package release. I will then > use it also for the Flexbar version on sourceforge and will try to come along > without modifications. Since I applied only very slight modifications, I am > optimistic that it works out and I will discuss it with SeqAn developers. I > definitely prefer to use an official and unmodified release, and see it as an > aim for coming versions. Besides from cleaner sources and packaging, it would > then also become obsolete to adjust new SeqAn libs to modifications for > inclusion in Flexbar. However, I am not sure how fast a new packaged release > will be provided. > > Best regards > Johannes -- http://fam-tille.de -- To UNSUBSCRIBE, email to debian-med-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130129072954.gb31...@an3as.eu
