Camelia, the Perl 6 bug

IRC log for #bioperl, 2011-03-08

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

All times shown according to UTC.

Time Nick Message
00:09 rdesfo joined #bioperl
00:54 rdesfo left #bioperl
03:54 CIA-57 bioperl6: takadonet master * ra71522e / lib/Bio/PrimarySeq.pm : implemented first version of subseq that takes a start and stop coordinates. Still have to implement a version that takes Bio:LocationI - http://bit.ly/gAr9hJ
03:55 rdesfo joined #bioperl
03:56 rdesfo left #bioperl
04:56 CIA-57 bioperl6: takadonet master * r19e183e / t/PrimarySeq.t : Commented out test for old api for translate - http://bit.ly/hQrAIU
04:56 CIA-57 bioperl6: takadonet master * r4d89f57 / lib/Bio/PrimarySeq.pm : Second version of translate. Implemented as a single method instead of multis. Got find_orf method done as well. Most test are passing. No type checking or exception throwing done - http://bit.ly/gV5eqU
05:01 CIA-57 bioperl6: takadonet master * re8b324c / t/Tools/CodonTable.t : changed test that used old translate API using positional based instead of named based - http://bit.ly/hxiXrs
06:48 bag joined #bioperl
07:39 zenman joined #bioperl
07:46 bag left #bioperl
07:47 zenman left #bioperl
08:13 mariolone joined #bioperl
08:13 mariolone left #bioperl
09:11 genehack left #bioperl
09:34 mariolone joined #bioperl
09:34 mariolone left #bioperl
09:42 cassj joined #bioperl
10:46 genehack joined #bioperl
12:35 _ilbot left #bioperl
12:36 _ilbot joined #bioperl
12:36 Topic for #bioperl is now Be patient!  Sometimes it takes a few minutes for people to notice your question! | http://bioperl.org/wiki/IRC#Getting_help | http://www.bioperl.org/wiki/Using_Git | nopaste to gist.github.com
13:35 CIA-57 bioperl6: takadonet master * r9827df2 / (lib/Bio/PrimarySeq.pm t/PrimarySeq.t): Uncommented test for testing subseq. Added new multi subseq but nyi since waiting on each_Location method to be done - http://bit.ly/eidNFd
14:05 zenman_ joined #bioperl
14:09 zenman_ left #bioperl
14:48 cassj left #bioperl
15:11 pyrimidine rbuels: I'm going through the application now, making small changes if there is something wrong
15:24 pyrimidine rbuels: re: What steps will you take to encourage students to interact with your project's community before, during and after the program?
15:25 pyrimidine I think posting regular blog reports might be a good idea, though that largely depends on how much exposure they get
15:25 pyrimidine (the blogs get)
15:26 pyrimidine otherwise, regular progress reports are good, as long as they describe the good and the bad :)
16:08 mzgrideng joined #bioperl
17:05 CIA-57 Bio-FeatureIO: Robert Buels master * r63daeff / lib/Bio/FeatureIO.pm : add better error message for when there is no file or filehandle - http://bit.ly/euS7DH
17:56 ank left #bioperl
18:05 rbuels pyrimidine: ok, feel free to edit
18:15 pyrimidine rbuels: already did
18:15 rbuels pyrimidine++
18:15 * rbuels is sitting in the gmod meeting
18:16 rbuels actually, the first day of the training session
18:16 rbuels cause i have to stay in durham for an extra day
18:16 * rbuels heckles scott as he talks about the gff3 format
18:16 pyrimidine he likes that!
18:17 pyrimidine rbuels: tell him we're moving the GFF3 spec to github
18:17 pyrimidine :)
18:17 rbuels pyrimidine: ho ho ho
18:18 pyrimidine rbuels: BTW, have you seen this? http://redmine.open-bio.org/
18:18 pyrimidine rbuels: yes, we're switching from bugzilla to redmine
18:19 rbuels saw the discussion going back and forth about it, hadn't seen it yet
18:19 * rbuels doesn't have a strong opinion about it
18:19 pyrimidine it's definitely nicer
18:43 rbuels wow.  it takes a *long* time to talk about how to represent things in GFF3
18:44 pyrimidine a lot of thought went into it
19:26 CIA-57 Bio-FeatureIO: Robert Buels master * rcfc8b44 / t/gff3.t : add failing test for proper gff3 unescaping in parse - http://bit.ly/gGhfVU
19:31 dave_messina joined #bioperl
19:32 CIA-57 Bio-FeatureIO: Robert Buels master * r96e75c1 / t/gff3.t : oops, that failing test was wrong. correct - http://bit.ly/f8JpFF
19:46 dave_messina Hey pyrimidine, is bioperl6 going to be s GSoC project?
19:46 pyrimidine dave_messina: yep
19:46 pyrimidine dave_messina: at least, if someone comes up with a good proposal
19:46 dave_messina Sweet!
19:47 pyrimidine dave_messina: probably biome as well (Moose-based)
19:47 dave_messina I just noticed how much activity there has been lately, by one Philip Mabon.
19:47 pyrimidine yes, aka takadonet
19:47 pyrimidine (he's here)
19:48 dave_messina Ah okay, I figured he'd be at your place. :)
19:48 pyrimidine :)
19:49 CIA-57 Bio-FeatureIO: Chris Fields master * rbd88ea0 / lib/Bio/FeatureIO/gff.pm : <headsmack>unescape, not escape</headsmack> - http://bit.ly/hoMtOM
19:50 pyrimidine rbuels: bug fixed, helps to unescape those (instead of escaping them)
19:50 pyrimidine :)
20:01 rbuels pyrimidine: stop it, i'm fixing stuff!  go away!!
20:02 rbuels MOM!  Chris's bothering me!
20:02 * rbuels yells
20:02 * rbuels chuckles
20:02 rbuels pyrimidine: i'm implementing the gff3 writing
20:02 rbuels pyrimidine: cause i needs it
20:02 rbuels pyrimidine: i'm finding a bunch of other bugs as well, heh
20:03 rbuels as i'm writing this round-tripping test
20:03 takadonet dave_messina: hello sir :)
20:04 dave_messina takadonet:  hiya!
20:04 pyrimidine rbuels: add as many tests as possible
20:04 * rbuels waits for deafferret to say 'stop bothering your brother'
20:05 * pyrimidine shakes deafferret awake
20:05 takadonet dave_messina: are you planning to help out on bioperl6?
20:05 dave_messina takadonet:  nah, just an interested observer. :) I'm glad to see it happening.
20:05 takadonet dave_messina: slow and steady...
20:06 * dave_messina works for me!
20:06 takadonet planning a refactoring when a little bit more is working
20:07 dave_messina btw, pyrimidine, any luck finding somebody to represent BioPerl at BOSC this year? Surely Heikki or Hilmar is going.
20:07 pyrimidine dave_messina: not yet.
20:07 dave_messina takadonet: sounds good.
20:07 pyrimidine dave_messina: though, the reasons for me not making it are pretty good
20:08 pyrimidine rbuels: please feel free to tear FeatureIO a new one
20:08 dave_messina pyrimidine: oh for sure! (assuming you mean your little cloning experiment)
20:09 pyrimidine dave_messina: well, that and I'll likely be relocating
20:09 dave_messina pyrimidine: cool! where to?
20:09 pyrimidine Well, he's contacted you as well :)
20:10 dave_messina Seriously?! That's awesome!
20:10 pyrimidine yes
20:10 pyrimidine the way things are shaping up
20:10 dave_messina Congrats, Chris! That's so great. Really.
20:10 takadonet pyrimidine: congrats
20:10 pyrimidine thanks
20:11 pyrimidine some other interesting things have also popped up locally, but I don't think we want to stay in the midwest after this last winter
20:11 dave_messina oh hell yes. Climate is key.
20:11 dave_messina I hope I can be out in CA one way or another.
20:12 pyrimidine dave_messina: can always look north to San Fran, or south to San Diego.  Lots going on in both areas (academic and tech)
20:12 rbuels pyrimidine: you ok with me introducing a dep on Bio::GFF3::LowLevel?
20:12 pyrimidine rbuels: go for it
20:12 * rbuels will otherwise cut and paste the functions from there into it, lol
20:13 dave_messina That's exactly my plan. So far, I'm hearing most of the action is in the Bay area, but I think I really want to try San Diego. It's cheaper to live there and better weather.
20:13 pyrimidine rbuels: if we abstract out the parser layer it makes things easier; since you've already done that seems like a no-brainer to me
20:15 pyrimidine dave_messina: agreed.  You should contact Scott Markel at Accelrys
20:16 dave_messina Thanks — will do.
20:16 pyrimidine (he was at BOSC and last Jan's GMOD meeting in San Diego
20:16 dave_messina I've met him at ISMB a couple of times, and so I'm hoping he'll remember who I am.
20:19 pyrimidine would be interesting, we could be dramatically increasing the West Coast influence for bioperl; rbuels is Portland, George Hartzell and Chris Mungall are San Fran, Nat Goodman is Seattle I think
20:19 pyrimidine not to mention Jason
20:20 takadonet pyrimidine: takadonet in central Canada :)
20:21 dave_messina pyrimidine: Yeah, that's true! Should be easier to set up hackathons.
20:21 dave_messina takadonet: you'll be welcome to come over and defrost. :)
20:21 takadonet little hard to get travel  approved otherwise of Canada right now
20:22 dave_messina just load up the family truckster and head to Wallyworld.
20:22 pyrimidine takadonet: where in central Canada
20:23 pyrimidine my wife's from Nova Scotia, her brother lives in Ottawa, and her mother now lives in Moose Jaw, Saskatchewan
20:23 takadonet Winnipeg, Manitoba
20:24 pyrimidine cool (literally)
20:24 takadonet at the National Microbiology laboratory ( the Canadian CDC version)
20:25 rbuels dave_messina: portland is a wondrous city, the only down side is the weather
20:25 rbuels dave_messina: well, that and no jobs i guess, heh
20:26 pyrimidine ...and the lack of bioinformatics jobs
20:26 rbuels dave_messina: but this is where young people go to retire
20:26 * rbuels chuckles
20:27 dave_messina rbuels: heh. :) yeah, I don't think I could take the grey and the rain. I've had plenty of that over here. But I think I'd dig the rest of it.
20:32 rbuels pyrimidine: is it really necessary anymore to eschew test description strings?
20:32 rbuels pyrimidine: (in Bio::FeatureIO)?
20:33 pyrimidine rbuels: you mean,  'is($f[0]->primary_tag,'match', 'primary_tag');'?
20:33 pyrimidine vs 'is($f[0]->primary_tag,'match');'?
20:34 pyrimidine I'm pretty neutral on it, actually
20:35 rbuels pyrimidine: that is (was?) a  backcompat thing wasn't it?
20:35 rbuels can't think of any other reason not to have them
20:36 pyrimidine no reason not to have them, but in some cases (like the above) the test description can be a bit redundant, unless you want the descriptions in the verbose TAP output
20:36 * rbuels nods
20:36 rbuels pyrimidine: there is both a Build.PL and a dist.ini, which should i add deps to?
20:37 rbuels both?
20:37 rbuels (well, dzil can autodiscover these, so i will only add to Build.PL)
20:37 pyrimidine ok
20:38 pyrimidine I would like to have a passthrough Build.PL, per our last conversation (similar to Moose)
20:38 rbuels oh yeah ...
20:39 CIA-57 Bio-FeatureIO: Robert Buels master * rec940aa / (lib/Bio/FeatureIO/gff.pm t/gff3.t): fix another little parsing bug, repeated attrs not handled correctly. with tests - http://bit.ly/hm89fM
20:39 CIA-57 Bio-FeatureIO: Robert Buels master * r604d918 / (Build.PL lib/Bio/FeatureIO/gff.pm t/gff3.t): implement GFF3 writing - http://bit.ly/g0zQLc
20:39 pyrimidine we may need to come up with a dzil set of plugins for bioperl at some point
20:39 rbuels there's the gff3 writing
20:40 pyrimidine particularly if we intend on switching everything over to dzil at some point
20:40 pyrimidine rbuels++ # gff3 writing
20:40 pyrimidine now we just need a  karma tracker
20:40 rbuels heh
20:42 takadonet pyrimidine: like #perl6?
20:42 pyrimidine takadonet: yes
21:01 ank joined #bioperl
21:21 sl33v3_ joined #bioperl
21:22 sl33v3 left #bioperl
21:22 sl33v3_ is now known as sl33v3
21:27 dnewkirk joined #bioperl
21:35 CIA-57 bioperl6: Takadonet master * r530047d / (lib/Bio/Tools/CodonTable.pm t/Tools/CodonTable.t): new constructor can be called with or without an id parameter in CodonTable.pm - http://bit.ly/hsITZs
22:01 dave_messina left #bioperl
22:17 takadonet pyrimidine: ping
22:17 pyrimidine pong
22:17 pyrimidine (though I'm here for only a few more minutes)
22:18 takadonet what is your position on positional and named based parameters for methods?
22:18 takadonet currently some methods have both or just one type
22:19 * rbuels thinks it's dumb to have both
22:20 pyrimidine if a method has more that two arguments then probably should be using named args, otherwise positional is fine
22:20 pyrimidine agree with rbuels
22:20 pyrimidine though, with perl6 it's fairly easy to have both
22:22 pyrimidine simple is better, though
22:22 takadonet ya
22:23 * pyrimidine was just looking through the spec to see the current version: http://feather.perl6.nl/syn/S06​.html#Parameters_and_arguments
22:23 takadonet An example would be translate. It had both and i decided to get rid of positional since their is way too many parameters
22:23 pyrimidine right, and in most cases they are optional or have a default
22:23 takadonet ya
22:24 pyrimidine I think these are all mainly implemented in rakudo
22:24 takadonet multi method helps as well but we are not really consistent or do we care to be?
22:25 pyrimidine you can use multis if it makes sense, yes
22:26 takadonet subseq was a good example
22:26 takadonet one for LocationI parameter and another for start,stop
22:26 pyrimidine yep
22:26 takadonet Both are named parameters but looks like we also 'should' be supporting positional based....
22:27 takadonet it would be just two more multi but that would be 4 methods just for subseq
22:27 * rbuels sighs about Bio::Tools::RepeatMasker
22:27 rbuels really bad bug, been around for years.
22:28 pyrimidine no need to have both, it just complicates things
22:28 * rbuels is testing and fixing
22:28 takadonet k
22:28 pyrimidine rbuels: what's wrong?
22:28 takadonet pyrimidine: thanks
22:28 pyrimidine takadonet: np
22:28 rbuels pyrimidine: it messes up the coords and strand for RepeatMasker matches that are on the reverse strand
22:28 rbuels backwards.
22:28 rbuels cause there are no revcom matches in the test data
22:29 * rbuels is fixing this
22:29 rbuels (is making RepeatMasker gff3 today, can you tell?)
22:29 pyrimidine good. lots of examples of that, unfortunately
22:29 * rbuels chuckles
22:29 * rbuels is also drinking GMOD beer
22:29 pyrimidine I recall Bio::SearchIO::psl also having similar problems
22:30 pyrimidine or one of them BLAT-y outputs
22:30 pyrimidine GMOD beer++
22:30 rbuels hehe
22:31 rbuels pyrimidine: if gff.pm gets a feature with Target tag of ['FooSeq',1,2], you think it should recognize that and convert it to 'FooSeq 1 2' ?
22:32 rbuels cause RepeatMasker.pm makes them like that.
22:32 * rbuels is ambivalent
22:32 pyrimidine that seems... hacky
22:32 rbuels yeah, kind of
22:32 rbuels maybe RepeatMasker.pm should make Target tags like 'FooSeq 1 2'
22:33 * rbuels dresses in a flowing robe, puts on a blindfold, holds a sword
22:33 * rbuels left his scale in his other pants
22:33 * takadonet thinks he will make a gff 3 grammar for bioperl6
22:33 pyrimidine yeah, or it needs to be something richer (Bio::Annotation::Target), but then maybe that's overload
22:34 pyrimidine takadonet: good idea, should be fairly simple
22:34 takadonet i worked with the format in p5 already
22:34 pyrimidine most of us do these days :)
22:35 takadonet fasta and genbank are still king
22:35 * pyrimidine => home
22:36 * rbuels does not work much with genbank format
22:38 pyrimidine it's a PITA, b/c it wasn't standardized strangely enough.  Only the feature table was. that's why SeqIO::genbank is such a mess (someone is always screwing with the format)
22:38 dbolser rbuels: can you pay me to work full time on bp?
22:38 rbuels dbolser: hah
22:38 dbolser ;-)
22:38 dbolser it seems worth while that someone do it
22:39 dbolser can we have a wip round on the mailing list?
22:39 dbolser design a choding challenge to decide who gets the job
22:41 rbuels wip round?
22:41 rbuels job?
22:41 CIA-57 bioperl-live: Robert Buels master * r86dbcae / Bio/Tools/RepeatMasker.pm : localize $_ if you are going to modify it in a lib function - http://bit.ly/ggedCC
22:41 CIA-57 bioperl-live: Robert Buels master * r2de6af1 / (3 files in 3 dirs): bugfix to Bio::Tools::RepeatMasker, reverse matches were wrong - http://bit.ly/fO93PV
22:42 rbuels good .... bad ..... i'm the guy with the gun.
22:42 rbuels dave clements basically called me an evil imp at the gmod meeting today
22:42 rbuels was amusing.
22:42 * rbuels reflects that he should be talking about this in #gmod
22:44 dbolser you do have imp like qualities
22:44 rbuels true
22:45 rbuels rbuels the evil code imp
22:45 * rbuels chuckles
22:45 CIA-57 bioperl6: Takadonet master * rb8b5a69 / (lib/Bio/PrimarySeq.pm lib/Bio/Role/Location/Simple.pm): Subseq can now support Bio::Role::Location::Simple obj as a parameter. Implemented simple version of each_Location. - http://bit.ly/haCRGE
22:45 dbolser does a bdb adaptor get built the first time I try to connect to the GFF?
22:45 * rbuels looks at the code
22:46 * dbolser goes to the gbrowse adaptor tutorial
22:46 rbuels dbolser: say this in #gmod!
22:46 rbuels dbolser: keep the channel healthy!
22:46 dbolser ok
23:51 bag joined #bioperl

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