Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    GCC version: 10.1.0
    Michael R. Crusoe
    @mr-c

    With gcc 10.2.0 and `-std=c++17 -fconcepts' (which just entered Debian)

    /build/seqan3-3.0.1+ds/include/seqan3/alphabet/concept.hpp:759:39: error: comparison is always true due to limited range of data type [-Werror=type-limits]

    Internal compiler error: Error reporting routines re-entered.

    Michael R. Crusoe
    @mr-c
    seqan/seqan3#1993 has the details
    Michael R. Crusoe
    @mr-c
    @marehr hmm.. gcc 10.2 is still not happy https://github.com/seqan/seqan3/issues/1993#issuecomment-667137301
    Let me see if I can backport gcc 10.2 to Debian Buster for y'all :-)
    lutfia95
    @lutfia95
    Hi, what is the main changes in minimiser/kmer_hash between seqan3.0 and seqan3.0.2?
    smehringer
    @smehringer

    Hi @lutfia95,
    I think there have been a few edge case fixes for kmer_hash, namely

    • kmer_hash now preserves common_range
    • It now provides a size() member function if the underlying range is sized
    • If the underlying range is too small for a full kmer, the resulting view is empty

    Also note that the kmer_hash may have returned wrong values due to precision errors of std::pow casting its value to double. We fixed that with our own implementation of pow for integrals.

    Minimizers were just recently added so this a a new feature all together. We provide two views:

    • seqan3::views::minimiser
    • seqan3::views::minimiser_hash
      The second does kmer hashing of the forwward and reverse strand, as well as the minimiser computation in one.
    Ahmad lutfi
    @ah_ol_twitter
    Ok thank you very much
    Michael R. Crusoe
    @mr-c
    My dear German colleagues. I'm having a diffult time navigating http://oeffentlicher-dienst.info/tv-l/ ; my goal is to discover what the exact requirements are for each wage classification from E9 and up.
    Any assistance would be very appreciated!
    I found http://oeffentlicher-dienst.info/tv-l/allg/entgeltgruppen.html but it is still vague and imprecise
    Michael R. Crusoe
    @mr-c
    Some progress, I have found the description of the pay grades at http://oeffentlicher-dienst.info/pdf/tv-l/tv-l-nr11-anlage-a.pdf
    2 replies
    Michael R. Crusoe
    @mr-c
    Should I test a new "release candidate" for Debian?
    7 replies
    Michael R. Crusoe
    @mr-c
    Any help with the post-installation test cmake errors with 3.0.2 is appreciated: https://gitter.im/seqan/Lobby?at=5f723fb1b8a99f4519b270a0
    Michael R. Crusoe
    @mr-c
    Ah, seems like copying over the test/cmake directory fixes that
    1 reply
    Hannes Hauswedell
    @h-2
    Hi everyone! Since I am not up to date with current conventions: where do discussion points go that are not immediate tasks? Do I create an issue at seqan/seqan3 for them or are other places designated for this? There are quite a few issues associated with building/running the tests on a machine without internet. I would like to gather these points somewhere.
    3 replies
    Hannes Hauswedell
    @h-2
    Has someone successfully built our test suite with CMake-3.17? Because I am getting A LOT of errors.
    9 replies
    Michael R. Crusoe
    @mr-c
    Debian is having problems with the seqan3 tests timing out on arm64. Right now we compile + run the unit, header, snippet, and performance tests; which of theses groups of tests can we safely skip?
    2 replies
    Hannes Hauswedell
    @h-2
    Hm, why doesn't our I/O use our fast streambuf iterators?
    7 replies
    Lydia Buntrock
    @Irallia
    Do you know a tool with which you can manage references.
    More precisely: I would like to be able to see how often a paper I have read is cited in the other papers I have read. And I want to be able to see which authors are authors of several papers I have read. So that I can have a kind of relevance assessment for myself and my topic.
    I've tried JabRef, but you can't even include the references of the papers.
    rrahn
    @rrahn
    Not sure if it fits your purpose but I am using Mendeley.
    15 replies
    Michael R. Crusoe
    @mr-c
    Maybe someone wants to add seqan3 to https://github.com/rigtorp/awesome-modern-cpp ?
    4 replies
    rrahn
    @rrahn

    :tada: SeqAn 3.0.2 is released :tada:

    We are happy and proud to announce the new release of our modern c++ library for sequence analysis. Find out more about it https://seqan.de/seqan-3-0-2-release/ or check it out right away https://github.com/seqan/seqan3/releases/tag/3.0.2. Many thanks to all contributors and supporters.

    Michael R. Crusoe
    @mr-c
    Seqan 3.0.2 is in Debian, but I'm having issues with the arm64 build timing out. Is there a subset of the tests I could use that isn't so intensive to compile?
    This is the post-installation script that needs to be modified for Debian https://salsa.debian.org/med-team/seqan3/-/blob/master/debian/tests/run-tests#L26
    Michael R. Crusoe
    @mr-c
    Speaking of celebrations, I finally got a normal residence card! https://twitter.com/__crusoe/status/1316364807205990400
    7 replies
    Michael R. Crusoe
    @mr-c
    This message was deleted
    Hannes Hauswedell
    @h-2
    Is someone at FU tomorrow? There is some stuff in my old office that could be picked up then.
    3 replies
    Michael R. Crusoe
    @mr-c
    Has anyone released a seqan3 based app yet? Or is that waiting for Seqan3 3.1.0?
    6 replies
    Michael R. Crusoe
    @mr-c
    FYI, travis-ci.org is closing down next month. Someone needs to go to https://travis-ci.com/organizations/seqan/migrate to migrate seqan3 to travis-ci.com
    Hannes Hauswedell
    @h-2
    seqan.de is down.
    12 replies
    simonsasse
    @simonsasse
    seqan.de is up and running again. There was an issue with a caching plugIn.
    1 reply
    smehringer
    @smehringer
    :tada: Noice
    Michael R. Crusoe
    @mr-c
    I just learned about the German Employee Inventions Act (ArbnErfG) ; seems like a barrier to Free / Open Science. Maybe @h-2 or others have some advice or opinions to share..
    4 replies
    Marcel
    @marehr
    Funny and short lightning talk about refactoring code (https://www.youtube.com/watch?v=fzmjXK9JZ9o); @rrahn right?
    2 replies
    Michael R. Crusoe
    @mr-c
    Lamda2 doesn't build with GCC 10 anymore https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=975142
    2 replies
    Hannes Hauswedell
    @h-2
    Thank you all again for the wonderful digital hat and for being at the defence :)
    5 replies
    Michael R. Crusoe
    @mr-c
    My calendar says there is a CIBI related meeting right now, but I don't have the connection details; did I miss something?
    4 replies
    Michael R. Crusoe
    @mr-c
    @smehringer Do you have time for a spontaneous OpenMS / Seqan3 CLI chat?
    5 replies
    rrahn
    @rrahn
    Here a good 10 minutes read for your next coffee break: https://mtlynch.io/code-review-love/ I really appreciate the points being made in this article.
    Simon Gene Gottlieb
    @SGSSGene

    In the Wiki there a schedule: https://github.com/seqan/seqan3/wiki/Group-Organization
    It says that there are three meetings:

    1. Seqan Team - Monday - 1PM
    2. Seqan Core Dev - Tuesday- 1PM
    3. Seqan External - First Monday of the Month at 3pm.

    I guess these dates are outdated and the new dates are:

    1. Seqan Team - Wednesday 12-noon
    2. Seqan Core Dev - Monday 10am
    3. Seqan External - does this still exists?
    20 replies
    ammaraziz
    @ammaraziz
    Hi Folks, I'm sure this question has been asked a few times. Is there any python package for seqan3?
    8 replies
    Marcel
    @marehr
    A pretty cool article for computer scientists regarding RNA/DNA/amino acids and how they work in a language understandable to them xD https://berthub.eu/articles/posts/reverse-engineering-source-code-of-the-biontech-pfizer-vaccine/
    Knut Reinert
    @kreinert
    @smehringer Kommst du zum Interview?
    Ahmad lutfi
    @ah_ol_twitter
    Hi Folks, is there any implementation of Q-gram index in SeqAn 3 ?
    1 reply
    Ahmad lutfi
    @ah_ol_twitter
    Thanks for the answer, I would like to use q-gram for the similarity search between reference genome and sequencing data. I am trying now if the FM-Index is more efficient than q-gram.
    1 reply
    Michael R. Crusoe
    @mr-c

    I know that Seqan 1.x is basically unmaintained, but we did get a significant looking set of test failures with gcc 10: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=980685 We only have one application in Debian that uses Seqan 1.x (dindel) and it is dead/archived by its authors.

    The likely outcome is that both dindel and seqan 1.4.x will not be in the next release of Debian unless someone wants to fix the test failures.

    4 replies
    Hannes Hauswedell
    @h-2
    Just a thing I stumbled upon: Is there a reason some of the meta info of the arg parser is a vector of strings?
    6 replies