[Debian-med-packaging] Please consider tagging librdp-readseq releases

[Qiong Wang]

Hi Andreas,

I am sorry for the error messages you received during compiling. It was due to missing test files used by a Junit test. Thanks for pointing out the problem. I have updated our GitHub and tagged a new release for the affected repo. 

I encourage you to consider including our entire RDPTools module for its nice collection of tools that should be interesting to your clients. In addition to the well-know RDP Classifier, SeqMatch,  and ProbeMatch, it includes newer tools including FrameBot, a frameshift co rrection and nearest neighbor assignment tool and Xander, our latest gene-targeted metagenomic assembler tool. It also includes many important utility functions useful in microbial ecological analysis. Most of these tools have been recently published and detailed tutorials or workflow are provided on GitHub.    


Qiong

——————————————
Qiong Wang
Information Technologist II
Center for Microbial Ecology
Michigan State University
Ph: 517-353-3842

On Feb 25, 2015, at 3:26 AM, Andreas Tille <andreas at an3as.eu> wrote:

> Hi Qiong,
> 
> thanks for your attempt to support our packaging effort.  I hope you do
> not mind that I took the Debian Med development list into CC since there
> are more people involved than me personally.
> 
> On Tue, Feb 24, 2015 at 04:13:11PM -0500, Qiong Wang wrote:
>> The RDPTools is a module built upon a collection of repos. We expect our users to run git command to do updates. For example, to update to version 2.0.0, run these three commands to update the entire module:
>> git checkout 2.0.0
>> git submodule update --init
>> make
> 
> Also for Debian packaging there is some workflow to create a package
> right from a Git repository.  So in principle we could deliver packages
> also that way.  However, this method is less established and has some
> other side effects.
> 
> In any case in my mail from yesterday I wanted to make you aware that
> something is wrong with the release you did tag yesterday at Github.
> May be this is related to the nature of the project that it consists of
> several separate repositories - but this is wild specualtion since I do
> not have personal experience.  In any case I would recommend to fix the
> download tarball - other users interested in the tarball might be fooled
> by this empty download.  If you want I could throw this question to our
> debian-mentors mailing list where several people with a lot of knowledge
> about things like this are hanging around.
> 
>> If the Classifier is the primary repo that your group is interested,
> 
> Well, this might be a misunderstanding:  We are not a group with some
> specific interest.  Debian Med tries to support any software which is
> interesting in general in the field of biology and medicine.  So in
> principle the whole suite of rdptools perfectly fits our scope and we
> hope to do a service for you as developers of this software as well as
> your users.
> 
>> and you setup only works with tar balls, we can tag the release for the Classifier and the three dependent package (AlignmentTools, ReadSeq and TaxonomyTree).  I have just tagged these packages as a start. 
> 
> Since Tim (in CC) just has prepared the packaging for ReadSeq I started
> with this one just as a test.  We might reconsider our approach to
> tackle the whole RDPTools suite if this might make more sense (which is
> hard to decide for me alone).
> 
> Unfortunately when I tried to build ReadSeq I was stumbling upon a
> problem when running the test suite (which is automatically done at
> package build time):
> 
> ...
> ant -f junit-tmp.xml junit
> Buildfile: /home/andreas/debian-maintain/repack/librdp-readseq-java/ReadSeq-2.0.0/junit-tmp.xml
> 
> junit:
>    [javac] /home/andreas/debian-maintain/repack/librdp-readseq-java/ReadSeq-2.0.0/junit-tmp.xml:7: warning: 'includeantruntime' was not set, defaulting to build.sysclasspath=last; set to false for repeatable builds
>    [javac] Compiling 15 source files
>    [javac] Note: /home/andreas/debian-maintain/repack/librdp-readseq-java/ReadSeq-2.0.0/test/edu/msu/cme/rdp/readseq/utils/gbregion/JoinRegionTest.java uses unchecked or unsafe operations.
>    [javac] Note: Recompile with -Xlint:unchecked for details.
>    [junit] Running edu.msu.cme.rdp.readseq.IndexedSeqReaderTest
>    [junit] Indexed 5 sequences in 1 ms
>    [junit] Tests run: 4, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0,122 sec
>    [junit] Running edu.msu.cme.rdp.readseq.SequenceParserTest
>    [junit] Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0,075 sec
>    [junit] Running edu.msu.cme.rdp.readseq.readers.QSeqReaderTest
>    [junit] Tests run: 1, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0,068 sec
> 
> BUILD FAILED
> /home/andreas/debian-maintain/repack/librdp-readseq-java/ReadSeq-2.0.0/junit-tmp.xml:11: Test edu.msu.cme.rdp.readseq.readers.QSeqReaderTest failed
> 
> 
> Do you have any idea what might went wrong here?
> 
> Kind regards and thanks for your support
> 
>        Andreas.
> 
> 
>> On Feb 24, 2015, at 3:37 PM, Andreas Tille <andreas at an3as.eu> wrote:
>> 
>>> Hi Qiong,
>>> 
>>> thanks for your intend to help.
>>> 
>>> On Tue, Feb 24, 2015 at 11:29:22AM -0500, Qiong Wang wrote:
>>>> Thank you for your inquiry about tagging our releases on GitHub. Since many RDP tools are dependent on other tools from RDP, we created a tool module called RDPTools to make it easy for our users to install and update on their local system. We plan to only tag releases in RDPTools repo. We will follow the major.minor.patch versioning recommended by Git. Please take a look at the current tag 2.0.0 from https://github.com/rdpstaff/RDPTools/releases. Let us know if that works for you. 
>>> 
>>> In principle it would be OK to collect different modules in one main
>>> release tarball.  However, somehow seems to be broken with the creation
>>> of this tarball since the download has only 1710 Bytes and all subdirs
>>> are empty.
>>> 
>>> BTW, our actual target is rdpclassifier anyway since it is used by
>>> Qiime which we are packaging as well.
>>> 
>>> Kind regards
>>> 
>>>    Andreas.
>>> 
>>> -- 
>>> http://fam-tille.de
>> 
>> 
> 
> -- 
> http://fam-tille.de

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/debian-med-packaging/attachments/20150225/f02671b9/attachment.html>