Camelia, the Perl 6 bug

IRC log for #bioperl, 2012-07-03

| Channels | #bioperl index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:44 gtuckerkellogg joined #bioperl
05:20 gtuckerkellogg joined #bioperl
06:02 gtuckerkellogg joined #bioperl
06:11 gtuckerkellogg joined #bioperl
07:57 vickMania joined #bioperl
08:38 kai deafferret: that seems to have helped :)
08:46 necrosis joined #bioperl
08:46 necrosis Hello
08:47 kai hi
08:47 necrosis I have bioinformatic unit now
08:47 necrosis Now we are learning perl
08:48 kai sounds like a good way to start :)
08:49 necrosis We are going to try alignments and so on.
08:49 necrosis it is possible to write pearl code in "R"?
08:50 kai I've never really used R, but as far as I'm aware R uses it's own scripting language
08:52 necrosis rly? R is similar like other scripting languages
08:52 necrosis But i guess, Perl is much more easy like R
08:52 necrosis all right. My course is going ff
09:32 vickMania joined #bioperl
11:10 leont joined #bioperl
11:54 gtuckerkellogg joined #bioperl
12:36 kai is the old bugzilla still visible somewhere? I'm trying to figure out the reason for a change to the genbank seqio module, and the commit message only mentions a bugzilla id
12:36 kai ah, wiki still has a ling
12:37 kai link, too
12:42 leont joined #bioperl
13:42 carandraug joined #bioperl
14:11 deafferret kai: :)
14:14 kai so apart from commenting on a bug about the SeqIO::GenBank module, I also had to open a bug about biopython's Bio.SeqIO writing of GenBank files
14:14 pyrimidine ?
14:15 kai pyrimidine: redmine was being mean to me
14:15 pyrimidine did it taunt you a second time?
14:15 kai I'm good now
14:16 pyrimidine ok
14:16 kai seems like deafferret bullied it into cooperation
14:16 deafferret lol
14:16 deafferret does biopython have an IRC presence?
14:16 pyrimidine no
14:16 pyrimidine bioruby does
14:16 * deafferret boos
14:16 kai now my only problem is that GenBank's specs suck
14:17 pyrimidine genbank has a spec?!?
14:17 kai yeah
14:17 kai it's just very well hidden
14:17 pyrimidine linky-link?
14:17 kai part of the specs are in the release notes...
14:17 kai ftp://ftp.ncbi.nih.gov/genbank/r​elease.notes/gb185.release.notes
14:18 pyrimidine oh yeah, the basic stuff
14:18 kai you're looking for chaper 3 there
14:18 pyrimidine The problem is, not all the annotation is noted
14:19 kai and then for the biopython bug I ran into, about feature qualifiers, the docs are at http://www.insdc.org/docum​ents/feature_table.html#7.3.1
14:19 kai which I managed to figure out after reading over the bioperl parser, and then following the git history to http://bugzilla.open-bio.org/show_bug.cgi?id=1516
14:20 pyrimidine kai: Peter Cock and Brad Chapman are pretty responsive on the biopython list (I collaborate with both)
14:20 kai which then gave me the correct enchantment for google
14:21 kai yeah, peter replied to my bug report within minutes
14:21 kai which reminds me, the list changed as compared to the implementation bioperl has. I guess I'll prepare a patch for that, too
14:22 pyrimidine for some reason, redmine is not assigning our usual default channel for bug report notification (bioperl-guts).  Will have to look into that.  I suppose I could file a bug :)
14:22 pyrimidine ...and then the Higgs Boson would swallow the open source universe...
14:22 pyrimidine sorry, need caffeine
14:25 pyrimidine Re: parsers, the problem is that all the behavior (how to handle each condition, classes for seq/feat/ann obj creation) is baked into the parser.  It's entirely too monolithic
14:25 pyrimidine (that is for both bioperl and biopython)
14:26 pyrimidine we just need to clone deafferret 100-fold, and all his clones can work on it
14:29 kai pyrimidine: the computer scientist in me thinks that this problem can be solved with a further abstraction layer
14:29 pyrimidine kai: exactly
14:29 pyrimidine low-level generic parser
14:30 kai so we basically define a file format using some grammar, and have a code generator that procudes the required perl/python/ruby/java/whatever code
14:30 pyrimidine yep
14:30 * rbuels bets there are perl-yacc or flex thingies out there
14:30 kai I wonder if I can talk our CS dept. into handing that out as a B.Sc. thesis or the like
14:31 pyrimidine that's probably what it would require
14:31 pyrimidine :P
14:31 kai well, sounds more useful than a lot of other topics they have.. ;)
14:31 pyrimidine heh
14:31 pyrimidine true
14:31 pyrimidine but then again, they don't like things that aren't *novel*
14:32 kai well, we just need the right buzzwords... "translational comp-sci" or the like
14:33 pyrimidine rbuels: there are a few things.  the bioruby folk have been looking into a few options
14:33 rbuels pyrimidine, kai:  would be a good MEng project.  those don't have to be 'novel'
14:33 * rbuels knows, has an MEng
14:33 kai and of course they can add a code generator for whatever functional language is hip with the kids^H^H^H^HCS people these days
14:33 pyrimidine rbuels: how do you feel about the bioruby/D project approach?
14:34 rbuels pyrimidine: sounds OK as an idea, the proof is in the pudding
14:34 rbuels i haven't looked at any of their code
14:34 pyrimidine neither have I
14:34 pyrimidine been too busy this year, unfortunately
14:35 pyrimidine but  a good kind of 'busy'
14:35 kai rbuels: hm, all the engineers I know learn how to build cars or stuff for building cars. then again, that's probably what I get for living in an area where people build mercedeces, porsches and 90% of all car electronics
14:36 leont bioruby/D sounds like I know who's behind that :-p
14:39 pyrimidine leont: yep, he has mentioned you before as well
14:39 leont :-)
14:40 pyrimidine about trying to lure you over to the bioruby/D side, or something similar :)
14:41 kai do they have cookies?
14:41 * kai refuses to be lured to any side that doesn't have cookies
14:42 leont D is charming but lacking libraries to do anything useful with it.
14:42 leont BioRuby isn't really convincing to me, since I prefer Perl over Ruby
14:44 pyrimidine leont: I agree
14:44 kai for the Bio* side, I had the impression that all the other languages lack features BioPerl already has. I almost had to write another HMMer3 parser, but then GSoC seems to have saved me the trouble :)
14:45 pyrimidine kai: they're gradually catching up, but not everything has been addressed
14:45 pyrimidine some Bio* are stronger in protein structure, etc
14:45 kai well, it's a lot of work, I'm not blaming anybody
14:46 leont BioPython may have a few things we don't have (I still have to plug in my SFF parser into Bio::SeqIO), but generally we have implemented a lot more.
14:46 leont Perl + R covers pretty much anything
14:46 pyrimidine leont: will be using that one soon :)
14:46 pyrimidine Bio::SFF
14:47 leont :-)
14:49 kai one thing I need to look into soon is biosql, that should make it easy to store data in a way that can be modified using perl or python scripts
14:50 * leont would like to see a BioSpore, but he isn't sure how much SPORE is alive anyway
14:50 leont (it's description format for JSON webservices that can be converted into working code)
14:51 leont A bit like WSDL, but actually understandable
14:51 pyrimidine I recall SPORE coming up a few times before in various discussions.  I think someone mentioned it a few years back at BOSC
14:53 kai a bit hard to google...
14:53 pyrimidine heh, just searched the bioperl archives and found one hit for Spore attributed to... leont
14:54 kai ah, https://github.com/SPORE/specifications
14:55 pyrimidine leont: I do like that idea
14:56 pyrimidine I can start asking around when I go to BOSC next week
14:56 kai that's next week already? have fun..
14:57 kai my boss decided to only send me to ECCB
14:58 pyrimidine Eastern Caribbean Central Bank?
14:58 pyrimidine Oh, European Conference on Computational Biology
14:58 pyrimidine (the former sounds both fun AND boring)
14:59 * leont doesn't even know if someone in his department goes, should have asked before leaving for warmer places
15:00 kai I really need to get a post-doc working on marine microbiology..
15:00 kai the tropical kind, not the polar kind
15:03 kai anyway, time for me to get out of here
15:03 kai here as in work, not IRC..
15:03 pyrimidine kai: o/
15:20 hash_table joined #bioperl
17:18 leont joined #bioperl
17:19 leont Ugh, network outage
19:00 pyrimidine joined #bioperl
19:38 hash_table joined #bioperl
23:57 leprevost joined #bioperl

| Channels | #bioperl index | Today | | Search | Google Search | Plain-Text | summary