Re: Removing tophat (Was: Bug#938677: Please check autopkgtest of (may be failed) attempt of Python3 port (Was: Bug#938677: tophat: Python2 removal in sid/bullseye))
CC-ing Debian Med user list and specifically Tony to raise his
opinion from a BioLinux point of view. Tony, what's your point
about loosing software that can not be portet to Python3?
Kind regards, Andreas.
On Wed, Nov 27, 2019 at 12:03:13PM +0100, Andreas Tille wrote:
> On Wed, Nov 27, 2019 at 10:21:42AM +0100, Alex Mestiashvili wrote:
> > Before removing tophat I guess we need to handle the reverse deps:
> > med-bio
> > cufflinks
> > giira
> > med-cloud
> >
> > med-bio and med-cloud just need to remove the tophat from the list.
>
> We can easily release new med-* metapackages.
>
> > cufflinks doesn't depend on Tophat, just enchances it.
> >
> > So there is only giira which will suffer from tophat's removal, but it is
> > also a 5 years old code never updated by upstream.
>
> My guess it will also not get any update. Grepping the source:
>
> src/geneFinder/ReadInParameters_GeneFinder.java:
> " \n -mT [tophat/bwa/bwasw] : specify desired tool for the read mapping,
> DEFAULT: tophat \n" +
> src/geneFinder/TopHat_Call.java:for(File readFile :
> GeneFinder.rnaFilesWithNames.keySet()){ // if there is more than one read
> file, report them to tophat in a list
> src/geneFinder/TopHat_Call.java:secondExe = "tophat2
> --no-sort-bam " + optionString + "-o " + out_dir + " " + nameRefFile + " " +
> rnaFile;
> src/geneFinder/TopHat_Call.java:secondExe = "tophat2
> --no-sort-bam " + optionString + "-o " + out_dir + " " + nameRefFile + " " +
> fileNames;
> src/geneFinder/GeneFinder.java: TopHat_Call
> tophatStart = new TopHat_Call();
> src/geneFinder/GeneFinder.java:
> tophatStart.callTopHat(nameRef);
>
> It even looks as if we could replace tophat by hisat there ...
> Currently popcon shows 5 votes which is half the number of tophat. Not
> sure whether giira itself is triggering votes for tophat. If we might
> get some explicite vetos for giira here it might be worth to try
> replacing the tophat call by a hisat2 call - otherwise we probably will
> not loose much it it is removed as well.
>
> > It has some popcon number, but again it uses an outdated algorithm provided
> > by tophat which is not maintained anymore.
> >
> > Shall I open bugs against the reverse dependencies asking to remove tophat?
> > Could you please advise with the common procedure for that?
>
> I think we simply wait for some veto for a giira removal here and
> act accordingly. Meanwhile I can re-render debian-med metapackages.
>
> Kind regards
>
>Andreas.
>
> --
> http://fam-tille.de
>
> ___
> Debian-med-packaging mailing list
> debian-med-packag...@alioth-lists.debian.net
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-med-packaging
--
http://fam-tille.de
Re: Removing tophat (Was: Bug#938677: Please check autopkgtest of (may be failed) attempt of Python3 port (Was: Bug#938677: tophat: Python2 removal in sid/bullseye))
On Wed, Nov 27, 2019 at 10:01:18AM +0100, Alex Mestiashvili wrote: > med-bio should exclude tophat from the list of recommended packages. Done https://salsa.debian.org/blends-team/med/commit/97d887604435dda6527a8a213d71ed2ae940562c Kind regards Andreas. -- http://fam-tille.de
Re: Removing tophat (Was: Bug#938677: Please check autopkgtest of (may be failed) attempt of Python3 port (Was: Bug#938677: tophat: Python2 removal in sid/bullseye))
Hi Andreas, On 11/20/19 4:34 PM, Andreas Tille wrote: On Wed, Nov 20, 2019 at 03:33:33PM +0100, Alex Mestiashvili wrote: It is not that trivial to fix tophat and more over there is a successor - HISAT2. It is not maintained upstream since 2016 and one of the co-authors asks to stop using it: Please stop using Tophat. Cole and I developed the method in *2008*. It was greatly improved in TopHat2 then HISAT & HISAT2. There is no reason to use it anymore. I have been saying this for years yet it has more citations this year than last Fine for me. In 2017 we had already a discussion about removing tophat from Debian[0], and now I believe the time has come. I have no problems if you file a ROM. What we might consider for this case or in general: If there is a successor of some software would we want to 1) Use a virtual package name in the successor 2) Create a metapackage depending from the successor and delivering some docs about how to use the successor 3) Make med-bio (or whereever the outdated package was advertised in) conflicting with the outdated software? I think 1 and 2 don't really feet for this case. med-bio should exclude tophat from the list of recommended packages. Just filing a ROM request will not remove the package from user installations. Its a question whether we really want to prevent that users keep that package - but in case we want this the technical means mentioned above came to mind (not sure whether this is a complete list of possibilities). I guess I'll leave it to the user. However once python2 is removed the package will be broken, but I don't see a good way to prevent that and notifying the user is too much effort IMO. So I'll take the easy way and file ROM. Thanks, Alex
Removing tophat (Was: Bug#938677: Please check autopkgtest of (may be failed) attempt of Python3 port (Was: Bug#938677: tophat: Python2 removal in sid/bullseye))
On Wed, Nov 20, 2019 at 03:33:33PM +0100, Alex Mestiashvili wrote: > > It is not that trivial to fix tophat and more over there is a successor > - HISAT2. > It is not maintained upstream since 2016 and one of the co-authors > asks to stop using it: > > > Please stop using Tophat. Cole and I developed the > > method in *2008*. It was greatly improved in TopHat2 then HISAT > > & HISAT2. There is no reason to use it anymore. I have been > > saying this for years yet it has more citations this year than last Fine for me. > In 2017 we had already a discussion about removing tophat from > Debian[0], and now I believe the time has come. I have no problems if you file a ROM. What we might consider for this case or in general: If there is a successor of some software would we want to 1) Use a virtual package name in the successor 2) Create a metapackage depending from the successor and delivering some docs about how to use the successor 3) Make med-bio (or whereever the outdated package was advertised in) conflicting with the outdated software? Just filing a ROM request will not remove the package from user installations. Its a question whether we really want to prevent that users keep that package - but in case we want this the technical means mentioned above came to mind (not sure whether this is a complete list of possibilities). So you are the expert - do whatever you feel is necessary to do. Kind regards Andreas. > [0] https://lists.debian.org/debian-med/2017/12/msg00089.html -- http://fam-tille.de
Re: [Debian-med-packaging] Bug#938677: Please check autopkgtest of (may be failed) attempt of Python3 port (Was: Bug#938677: tophat: Python2 removal in sid/bullseye)
On 10/11/19 11:11 AM, Andreas Tille wrote: > Hi, > > I tried to use 2to3 to port tophat to Python3 in Git[1]. Unfortunately > the autopkgtest fails with: > > > autopkgtest [09:04:13]: test run-unit-test: [--- > > [2019-10-11 09:04:13] Beginning TopHat run (v2.1.1) > --- > Traceback (most recent call last): > File "/usr/bin/tophat", line 4106, in > sys.exit(main()) > File "/usr/bin/tophat", line 3898, in main > run_log = open(logging_dir + "run.log", "w", 0) > ValueError: can't have unbuffered text I/O > autopkgtest [09:04:13]: test run-unit-test: ---] > autopkgtest [09:04:14]: test run-unit-test: - - - - - - - - - - results - - > - - - - - - - - > run-unit-testFAIL non-zero exit status 1 > > > I'd applaude if in beginning of November if I'm back from beeing offline > if this would be fixed. :-) > > Kind regards > > Andreas. > > [1] https://salsa.debian.org/med-team/tophat > It is not that trivial to fix tophat and more over there is a successor - HISAT2. It is not maintained upstream since 2016 and one of the co-authors asks to stop using it: > Please stop using Tophat. Cole and I developed the > method in *2008*. It was greatly improved in TopHat2 then HISAT > & HISAT2. There is no reason to use it anymore. I have been > saying this for years yet it has more citations this year than last In 2017 we had already a discussion about removing tophat from Debian[0], and now I believe the time has come. [0] https://lists.debian.org/debian-med/2017/12/msg00089.html
Please check autopkgtest of (may be failed) attempt of Python3 port (Was: Bug#938677: tophat: Python2 removal in sid/bullseye)
Hi, I tried to use 2to3 to port tophat to Python3 in Git[1]. Unfortunately the autopkgtest fails with: autopkgtest [09:04:13]: test run-unit-test: [--- [2019-10-11 09:04:13] Beginning TopHat run (v2.1.1) --- Traceback (most recent call last): File "/usr/bin/tophat", line 4106, in sys.exit(main()) File "/usr/bin/tophat", line 3898, in main run_log = open(logging_dir + "run.log", "w", 0) ValueError: can't have unbuffered text I/O autopkgtest [09:04:13]: test run-unit-test: ---] autopkgtest [09:04:14]: test run-unit-test: - - - - - - - - - - results - - - - - - - - - - run-unit-testFAIL non-zero exit status 1 I'd applaude if in beginning of November if I'm back from beeing offline if this would be fixed. :-) Kind regards Andreas. [1] https://salsa.debian.org/med-team/tophat -- http://fam-tille.de
