Re: ITP: prime-phylo -- Bayesian estimation of gene trees taking the species tree into account
Hi Erik, could you please confirm that you are interested in finding a sponsor for your work? As I said the Debian Med team is really interested. In case you have some time constraints or whatever I'd volunteer to commit your work to our Vcs (please specify whether you prefer Git or SVN) and will check whether it is ready for upload. Kind regards and thanks so far for your work Andreas. On Thu, Nov 29, 2012 at 09:31:45AM +0100, Andreas Tille wrote: > Hi Erik, > > it seems I missed your ITP - otherwise I would have most probably pinged > you whether you intent to do the packaging inside the Debian Med team. We > try to inject medical and microbiological software into Debian and your > ITP seems to fit perfectly into our biology task[1]. > > So I would suggest you join the Debian Med team (I checked whether you > just have an account on alioth.debian.org but failed in searching for > your name - otherwise I would have added you to the team right now). All > steps are described in our team policy[2]. > > Meanwhile I might check your packaging you provided at mentors however, > I would strongly recommend (you might need require) that you choose at > your preference either our SVN or Git repository (see [2]) for your > packaging work to enable effective team maintenance. For instance I > would turn your citation of the scientific paper into a debian/upstream > reference which is way more flexible than simply putting it into the > long description (which is deprecated). > > I hope you like this idea and we would be very happy to welcome you in > our team. > > Kind regards > > Andreas. > > [1] http://debian-med.alioth.debian.org/tasks/bio > [2] http://debian-med.alioth.debian.org/docs/policy.html > > > Package: wnpp > > Severity: wishlist > > Owner: Erik Sjölund > > > > * Package name : prime-phylo > > Version : 1.0.3 > > Upstream Author : Lars Arvestad, Bengt > > Sennblad and others > > * URL : http://prime.sbc.su.se > > * License : GPL-3 > > Programming Lang: C++ > > Description: Bayesian estimation of gene trees taking the species > > tree into account > > > > PrIME (Probabilistic Integrated Models of Evolution) is a package > > supporting inference of evolutionary parameters in a Bayesian framework > > using MCMC. A distinguishing feature of PrIME is that the species tree is > > taken into account when analyzing gene trees. > > -- > 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/20121129083145.gd4...@an3as.eu > > -- 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/20121206151629.ga31...@an3as.eu
[MoM] Status (Was: Second round of advise on packaging python-csb)
Hi Tomás, because we started the MoM project here on the Debian Med list I'm quoting the last message about your MoM package (which actually was on the python-modules list) here and wonder whether you could give some status update. It would be great if you would give some succcess / failure / blocker report. Perhaps we could finally get some first package out even if not all wishlist features (like testing / Python 3.0) might be implemented yet. Thanks for your work on this Andreas. On Fri, Nov 23, 2012 at 04:53:35PM +0100, Tomás Di Domenico wrote: > Sorry for the late reply, and thanks to everyone who contributed to this > thread. > > On 15/11/12 09:01, Andreas Tille wrote: > > > If Jakub wrote *ideally* than he most probably intended to write that we > > should try to do so but there might be some good reasons to derive from > > this ideal situation. I admit, I usually try to rebuild the docs in my > > packages and I do at least verify that I can *reproduce* all the docs. > > However, sometimes there are good reasons to simply use the > > autogenerated docs from upstream. Without having checked the thing > > myself your description sounds as if it could be the case here. > > IMHO the question whether you rebuild the docs from source and the fact > > whether the docs will end up in a separate binary package are > > orthogonal. I'm a fan if separate docs and I think 24MB are some good > > reason to do this. > > Great, seems like a separate package would be the way to go then. > > > I agree with those other to people who answered this part of your mail > > that having the tests packaged and thus ready for testing by the user > > at the installation target (which is simply different from testing at > > upstream side) is a very good idea and should be approached. Please > > also regard Jakub's (?) hint to DEP8. > > Right. I've contacted upstream regarding the pickled files Jakub brought > up in a previous message. I'm also looking at the autopkgtest docs I can > find, trying to figure it out. Could anyone point me to some python > packages using this kind of automated testing? If you'd rather teach me > how to fish, I'd be very interested in learning about the best way to > search packages for specific content in the "debian/*" files. In this > case, for example, search for "XS-Testsuite: autopkgtest" and "python" > or something of the sort. (Wishful thinking, maybe?) > > Tomás > > > -- > To UNSUBSCRIBE, email to debian-python-requ...@lists.debian.org > with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org > Archive: http://lists.debian.org/50af9bff.3030...@tdido.com.ar > > -- 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/20121206153134.gb31...@an3as.eu
Re: Help with lintian error: bad-distribution-in-changes-file
On Wed, Dec 5, 2012 at 8:56 PM, Charles Plessy wrote: > Le Wed, Dec 05, 2012 at 08:49:19PM -0500, Carlos Borroto a écrit : >> >> I recently switch my desktop to Debian Testing and now I get this >> error when trying to pack for Ubuntu: >> E: snappy1.0.3-java changes: bad-distribution-in-changes-file precise > > Hi Carlos, > > not sure it helps, but you can have a look at Lintian's vendor customisation > system. > > http://wiki.debian.org/Lintian/Spec/VendorCustomization Hi Charles, I found references to VendorCustomization and was reading about it, but it wasn't clear to me how to write my profile file. After poking around a little more I found there is actually an Ubuntu profile configuration[1] in the system already. The only thing I needed was to set "LINTIAN_PROFILE=ubuntu". [1]/usr/share/lintian/profiles/ubuntu/main.profile Thanks, Carlos -- 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/CABgGhBL7n=0+ikVNLBmN1v1+-RjatKF8Y2-=u0ysjtxfepv...@mail.gmail.com
Re: Help with lintian error: bad-distribution-in-changes-file
Le 6 déc. 2012 à 02:56, Charles Plessy a écrit : > Le Wed, Dec 05, 2012 at 08:49:19PM -0500, Carlos Borroto a écrit : >> >> I recently switch my desktop to Debian Testing and now I get this >> error when trying to pack for Ubuntu: >> E: snappy1.0.3-java changes: bad-distribution-in-changes-file precise > > Hi Carlos, > > not sure it helps, but you can have a look at Lintian's vendor customisation > system. > > http://wiki.debian.org/Lintian/Spec/VendorCustomization Or use LaunchPad PPA system? Eric -- 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/aeb2ec0e-9e00-403f-9ff6-93e0e1f13...@gmail.com
Re: Help with lintian error: bad-distribution-in-changes-file
On Thu, Dec 6, 2012 at 11:12 AM, Eric Maeker wrote: > > Le 6 déc. 2012 à 02:56, Charles Plessy a écrit : > >> Le Wed, Dec 05, 2012 at 08:49:19PM -0500, Carlos Borroto a écrit : >>> >>> I recently switch my desktop to Debian Testing and now I get this >>> error when trying to pack for Ubuntu: >>> E: snappy1.0.3-java changes: bad-distribution-in-changes-file precise >> >> Hi Carlos, >> >> not sure it helps, but you can have a look at Lintian's vendor customisation >> system. >> >> http://wiki.debian.org/Lintian/Spec/VendorCustomization > > Or use LaunchPad PPA system? > Eric > I do that already, see Debian Med's PPA[1], but I always like to build locally first to be sure I'm not uploading broken packages to LaunchPad. I found it saves tons of time. [1]https://launchpad.net/~debian-med/+archive/ppa Regards, Carlos -- 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/cabgghb+ncozotm3b-ngclvy5uge_spp8cinmb0stbaot00+...@mail.gmail.com
Re: Help with lintian error: bad-distribution-in-changes-file
On Thu, Dec 06, 2012 at 01:54:11PM -0500, Carlos Borroto wrote: > I always like to build > locally first to be sure I'm not uploading broken packages to > LaunchPad. I found it saves tons of time. +1 Kind regards 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/20121206204350.gb4...@an3as.eu
Re: FastQC read for testing (Was: Some (minimal?) Java help needed)
On Thu, Nov 8, 2012 at 8:00 AM, Andreas Tille wrote: > > Any volunteer to test the lintian clean package with some real data > before I upload? Steffen? > Hi Andreas, I'm interested in testing fastqc. This is my second time dealing with a debian package in svn and my second miserably fail. This is what I'm done so far: $ svn co svn://svn.debian.org/debian-med/trunk/ debian-med-trunk $ cd debian-med-trunk/packages/babraham/fastqc/trunk/ $ ./debian/rules get-orig-source $ cd ../ $ tar xvf tarballs/fastqc_0.10.1.orig.tar.gz $ cd FastQC/ $ ln -s ../trunk/debian $ fakeroot ./debian/rules binary dh binary --with javahelper dh_testdir jh_manifest jh_manifest: No jar in fastqc matching usr/share/fastqc/fastqc.jar. make: *** [binary] Error 1 I also tried building with svn-buildpackage but that took me nowhere. I feel I'm missing something here. Could you please help me understand the workflow for building packages in svn? Thanks, Carlos -- 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/CABgGhB++wCoyLJ5LkO=bXa=zwza47mp3_5dkcu88i52v4wh...@mail.gmail.com
[MoM] Packaging of python-csb - Report #4
Greets, Andreas. Sorry for going silent for so long. So, I've finally been able to get a grasp on both DEP8 testing and the Python3 version. I'm sure there are many new details to fix, but everything seems well underway. The tests are executed correctly, and the Python3 version installs and works fine. I went to upstream with the problem with the pickled files that Jakub brought up. It seems like those files get rebuilt when the tests run, if they do not exist. I asked upstream to remove them from the tarball, and they already did. Future versions will exclude those files, which will also make the package way smaller. About this, would it be possible to exclude the pickled files from the current version? Or do I have to wait/ask for a new upstream tarball? Another open issue I have now (at least that I'm aware of) are the docs, or how to go about creating a separate docs package. In particular, what is the structure of a docs-only package? Perhaps there's an example I can check? I'd ask you to check the package and let me know whatever issues you may find. I'd then like to go for a third round of comments in the Python least. Speaking of which, would it be ok to write in one list and CC the other? Thanks! -- 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/50c110db.2020...@tdido.com.ar
Bug#695311: ITP: python3-csb -- Python framework for structural bioinformatics
Package: wnpp Severity: wishlist Owner: Debian Med group * Package name: python3-csb Version : 1.1.0 Upstream Author : Michael Habeck * URL : http://csb.codeplex.com/ * License : MIT Programming Lang: Python Description : Python framework for structural bioinformatics Computational Structural Biology Toolbox (CSB) is a Python class library for reading, storing and analyzing biomolecular structures in a variety of formats, with rich support for statistical analyses. CSB is designed for reusability and extensibility and comes with a clean, well-documented API following good object-oriented engineering practice. -- 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/20121206214854.20121.19018.report...@countach.comp.bio.unipd.it
Re: FastQC read for testing (Was: Some (minimal?) Java help needed)
Hi Carlos, On Thu, Dec 06, 2012 at 04:16:32PM -0500, Carlos Borroto wrote: > > Any volunteer to test the lintian clean package with some real data > > before I upload? Steffen? > > I'm interested in testing fastqc. This is my second time dealing with > a debian package in svn and my second miserably fail. That's sad. BTW, I have no specific feeling whether to use Git or SVN. So if confirm you would like to work on this package you prefer Git and Steffen who has started the packaging does not insist moving it to Git would be no problem at all. > This is what I'm done so far: > $ svn co svn://svn.debian.org/debian-med/trunk/ debian-med-trunk > $ cd debian-med-trunk/packages/babraham/fastqc/trunk/ > $ ./debian/rules get-orig-source May be you trapped into a pitfall I builded unintentionally here. While this does not explain your later problem please note that the get-orig-source target in this specific form requires a not yet packaged uscan. The comments inside the debian/rules file will explain what to do. I worked a bit on a more clever uscan which is able to remove files mentioned in debian/copyright Files-Excluded. The discussion was done on debian-devel list and the response was quite positive. However, we are currently in freeze time - so features like this need to wait a bit. I perfectly agree that this trick is a bit hackish. Ironically you would not have run into this pitfall if we would have used Git with pristine-tar. So either you will use the uscan that is linked to in the comment of debian/rules or I could upload the original tarball I was using. Unfortunately the new queue these days takes some time - I uploaded at 20.11.2012 but it is not yet in unstable. Otherwise you could download the source from there as well. > $ cd ../ > $ tar xvf tarballs/fastqc_0.10.1.orig.tar.gz > $ cd FastQC/ > $ ln -s ../trunk/debian > $ fakeroot ./debian/rules binary > dh binary --with javahelper >dh_testdir > >jh_manifest > jh_manifest: No jar in fastqc matching usr/share/fastqc/fastqc.jar. > make: *** [binary] Error 1 I guess I might need the full log (without the snip). There might be some slight chance that the build target is not createt in case the original jar files remain present. If you are reluctant to use the new uscan you might simply remove all *.jar files from the original tarball and create some fastqc_0.10.1+dfsg.orig.tar.gz (or .xz as I decided to do.) > I also tried building with svn-buildpackage but that took me nowhere. I admit I never seriosely used svn-buildpackage - so no advise from me here, sorry. My workflow is tar xaf cd cp -a /debian . find . -name ".svn" -type d | xargs rm -rf pdebuild > I feel I'm missing something here. Could you please help me understand > the workflow for building packages in svn? In short: 1. use uscan from http://anonscm.debian.org/gitweb/?p=users/tille/devscripts.git;a=blob;f=scripts/uscan.pl;hb=HEAD to get the original tarball I used 2. if this fails please send full build log (or make it accessible somewhere 3. lets consider using Git if problems remain Thanks for working 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/20121206220712.gg4...@an3as.eu
Re: [MoM] Packaging of python-csb - Report #4
Hi Tomás, On Thu, Dec 06, 2012 at 10:40:43PM +0100, Tomás Di Domenico wrote: > So, I've finally been able to get a grasp on both DEP8 testing and the > Python3 version. I'm sure there are many new details to fix, but > everything seems well underway. The tests are executed correctly, and > the Python3 version installs and works fine. Sounds good - I have seen your commits. > I went to upstream with the problem with the pickled files that Jakub > brought up. It seems like those files get rebuilt when the tests run, if > they do not exist. I asked upstream to remove them from the tarball, and > they already did. Future versions will exclude those files, which will > also make the package way smaller. Fine. > About this, would it be possible to exclude the pickled files from the > current version? Or do I have to wait/ask for a new upstream tarball? There is no need to wait. We frequently are observing the need to strip files from upstream source. This is usually done in a get-orig-source target in debian/rules. Some weeks ago I tried to gain for some simplification of this process which is very frequently reinvented and you know geeks do not like to reinvent the wheel over and over. The result is an enhanced uscan script which I just advertised in my previous mail to this list about fastqc. May be you will just have a look at this mail and download the new uscan from http://anonscm.debian.org/gitweb/?p=users/tille/devscripts.git;a=blob;f=scripts/uscan.pl;hb=HEAD (copy it as uscan somewhere in your PATH so that it is executed before /usr/bin/uscan.) If you do so you just need to specify the files to remove in debian/copyright in the Files-Excluded: field and everything should work automatically (if not ping me and I'll help.) If you do not like this not yet official way you might need to dive into the other get-orig-source scripts we do have to a large extend in our Vcs - but IMHO this is learning stuff that will become outdated (the sooner the better.) > Another open issue I have now (at least that I'm aware of) are the docs, > or how to go about creating a separate docs package. In particular, what > is the structure of a docs-only package? Perhaps there's an example I > can check? I guess you can find a lot of examples for instance via apt-cache search python doc | grep "^python[^[:space:]]\+-doc" And jes, guessing from the size of the docs I think having them separated from the python module package makes sense. > I'd ask you to check the package and let me know whatever issues you may > find. I'll do so tomorrow. > I'd then like to go for a third round of comments in the Python > least. Speaking of which, would it be ok to write in one list and CC the > other? I'd like to see list cross-posting as an exception. Nobody will blame you about this in rare cases but in most cases it is better to stick to one list. As I said, I'm reading python-modules as well - it is perfectly fine to stick to this list. I just wanted to have some MoM report showing up here. 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/2012120655.gh4...@an3as.eu
Re: FastQC read for testing (Was: Some (minimal?) Java help needed)
On Thu, Dec 6, 2012 at 5:07 PM, Andreas Tille wrote: > In short: > >1. use uscan from > > http://anonscm.debian.org/gitweb/?p=users/tille/devscripts.git;a=blob;f=scripts/uscan.pl;hb=HEAD > to get the original tarball I used >2. if this fails please send full build log (or make it accessible > somewhere >3. lets consider using Git if problems remain > > Thanks for working on this This did the trick. I was able to build this package after using the new uscan, totally didn't read the comment inside debian/rules before. I'll be testing the resulting binary with real data tomorrow. I'll share the results. Regarding migrating to git, I would love if that's the case. I managed to write down a nice workflow[1] for updating git packages and for submitting then to launchpad. It takes me almost no time in most case. I will be using fastqc often enough that I would be happy to help keeping it updated. [1]https://docs.google.com/document/d/17rlBfHLLrVwL20eptZGyRkrrIZslsvmiZukjZWSqa1M/edit Best regards, Carlos -- 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/CABgGhBLVpWD=ktwzbtkiff75+4swetufgx7wk5t53tuuj_3...@mail.gmail.com
Ten Simple Rules for the Open Development of Scientific Software
Hi, feel free to read http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002802 I just sended the authors the following remark via G+: I like your ten simple rules. What I'm missing in rule 7 and/or 8 is the idea that you should try to get your software included into distributions. In the field of Bioinformatics from a Debian/Ubuntu point of view Debian Med (http://www.debian.org/devel/debian-med) is your partner and also Fedora and SuSE have similar projects. Kind regards 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/20121207071917.ga19...@an3as.eu
Re: FastQC read for testing (Was: Some (minimal?) Java help needed)
[Steffen: Please read last paragraph about moving fastqc from SVN to Git] Hi Carlos, On Thu, Dec 06, 2012 at 06:59:35PM -0500, Carlos Borroto wrote: > On Thu, Dec 6, 2012 at 5:07 PM, Andreas Tille wrote: > > In short: > > > >1. use uscan from > > > > http://anonscm.debian.org/gitweb/?p=users/tille/devscripts.git;a=blob;f=scripts/uscan.pl;hb=HEAD > > to get the original tarball I used > >2. if this fails please send full build log (or make it accessible > > somewhere > >3. lets consider using Git if problems remain > > > > Thanks for working on this > > This did the trick. I was able to build this package after using the > new uscan, totally didn't read the comment inside debian/rules before. That's definitely not your fault. Its actually a bit wacky idea of mine to rely on not yet published features but I somehow feel three arguments on my side: 1. Using the new uscan here makes it more popular 2. By doing so it gets more testers 3. Things become so brain dead simple that it would be a real waste of time to write explicite code to remove stuff from tarballs > I'll be testing the resulting binary with real data tomorrow. I'll > share the results. That's great. > Regarding migrating to git, I would love if that's the case. I managed > to write down a nice workflow[1] for updating git packages and for > submitting then to launchpad. It takes me almost no time in most case. > I will be using fastqc often enough that I would be happy to help > keeping it updated. Steffen, would you agree upon a move of fastqc SVN to Git? Kind regards Andreas. > [1]https://docs.google.com/document/d/17rlBfHLLrVwL20eptZGyRkrrIZslsvmiZukjZWSqa1M/edit -- 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/20121207072543.gb19...@an3as.eu
