Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-02-19

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

All times shown according to UTC.

Time Nick Message
00:00 deafferret rbuels++
00:00 rbuels karma rbuels
00:00 rbuels em...
00:00 rbuels ENOKARMATRACK
00:04 deafferret BioScala repo is 1 document file from 2007   :)
00:05 rbuels o_O
00:06 deafferret I bet ohloh.net is impressed
00:06 deafferret what was I doing?   oh ya, moose branch
00:08 rbuels no, class::mop branch
00:08 deafferret yy. looks good to me. tell them that I said it's fine  :D
00:09 rbuels why don't you tell them
00:09 rbuels lol
00:13 deafferret I have <= pull there than you do. one of these days I'm gonna get the key to hand out keys... that's be a glorious day
00:14 rbuels i have that key for gmod now  :-D
00:14 * rbuels rubs hands evilly
00:15 deafferret yes..... Yess.....!  YESSS!!!!
00:50 brunov joined #bioperl
01:02 brunov joined #bioperl
01:25 brunov_ joined #bioperl
04:37 balin joined #bioperl
05:27 driveby_bot joined #bioperl
05:27 driveby_bot /home/svn-repositories/bioperl: r16852 (cjfields) : test
05:27 driveby_bot Diff: http://tinyurl.com/yklhfsv
05:28 rbuels test?
05:28 rbuels test.
07:07 bag joined #bioperl
13:26 brunov joined #bioperl
14:14 brandi joined #bioperl
15:08 faceface test?
15:08 faceface rbuels...
15:09 faceface I've been asked to organize a workshop for 'potato breeders' describing the post-genomic resources...
15:09 faceface can we talk about hijacking SGN for potatos?
15:09 faceface you'd be doing all the work of course ;-)
15:09 faceface only joking, but I'd like to discuss fesability
16:31 rbuels faceface: no need to hijack SGN, we want to do potato very muhc
16:31 rbuels faceface: we just don't have many developers and need help
16:32 rbuels faceface: so we need either some money thrown our way, or some open-source-style development.
16:32 * rbuels thinks the latter more likely
17:21 driveby_bot joined #bioperl
17:21 driveby_bot /home/svn-repositories/bioperl: r16853 (cjfields) : silence warning
17:21 driveby_bot Diff: http://tinyurl.com/y9cx44n
17:22 driveby_bot joined #bioperl
17:22 driveby_bot /home/svn-repositories/bioperl: r16854 (cjfields) : silence warning (uninit value)
17:22 driveby_bot Diff: http://tinyurl.com/yfqwp2g
17:29 driveby_bot joined #bioperl
17:29 driveby_bot /home/svn-repositories/bioperl: r16855 (cjfields) : rollback accidental blast_pull commit
17:29 driveby_bot Diff: http://tinyurl.com/ygkhzgt
17:36 driveby_bot joined #bioperl
17:36 driveby_bot /home/svn-repositories/bioperl: r16856 (cjfields) : test count changed (volatile remote data)
17:36 driveby_bot Diff: http://tinyurl.com/yeqgshf
17:37 pyrimidine joined #bioperl
17:37 pyrimidine http://github.com/bioperl
17:38 brunov \o/
17:38 rbuels pyrimidine: did my message to the list about yapc come through
17:38 brunov pyrimidine, is that going to be the official repo?
17:38 rbuels i haven't gotten it yet
17:38 rbuels brunov: ha. ha.
17:38 rbuels brunov: you're funny.
17:39 pyrimidine nope; read-only, and trunk only ftm
17:39 brunov ouch
17:39 perl_splut yapc message came through
17:39 rbuels ok
17:39 brunov rbuels, one can dream
17:39 rbuels oh it'll get there
17:39 rbuels at SGN, we're probably going to put all our code on github soon
17:40 rbuels just hang all our dirty laundry out for everybody to see
17:40 rbuels lol
17:40 brunov what license do you guys use?
17:40 rbuels perl
17:40 rbuels or something
17:40 pyrimidine in my case, I just got sick of code.open-bio.org going down at inconvenient times
17:40 rbuels or maybe it's straight public-domain
17:40 rbuels we haven't thought much about it
17:40 brunov rbuels, nice
17:40 rbuels (licensing)
17:40 rbuels actually, we don't much care.
17:41 rbuels so public domain, i think
17:41 pyrimidine rbuels: http://www.ifosslr.org/ifosslr/article/view/26/35
17:41 pyrimidine nice article on licenses, btw they don't like artistic
17:41 rbuels hmmm
17:42 pyrimidine nice summary on groklaw: http://www.groklaw.net/articl​e.php?story=20100204170037353
17:43 brunov pyrimidine, so you'll keep the mirror repo updated manually or with some cron jobs or .. ?
17:43 pyrimidine cron job, using git-svn, sync every 15 minutes
17:44 brunov cool, so it'll be pretty safe to, for instance, patch against your mirror then
17:44 pyrimidine yes, it should be
17:45 brunov great, thanks. I never learned svn properly, and have to overthink every time I use it. This will ease things up for me.
17:46 rbuels wow, something really bad must have happened.  just saw 2 coast guard sea knight helos go in and out of OHSU in quick succession.
17:46 rbuels those things are loud.
17:47 pyrimidine brunov: You'll still need to submit patches through the regular channels, though.  This is meant to be read-only.
17:48 brunov pyrimidine, right. But it'd still be easier for me to do a local clone, make changes, create a patch, and submit that to bugzilla this way
17:48 pyrimidine brunov: yes ;>
17:48 brunov no pushing nor pull-requesting :)
17:48 pyrimidine right
18:00 * deafferret starts a YAPC::NA::2010 wiki page
18:02 rbuels good idea
18:06 deafferret http://www.bioperl.org/wiki/YAPC
18:06 deafferret post that to your mailing list, and smoke it
18:10 deafferret nm, I'll post it
18:10 rbuels oh, much more sensible.
18:11 rbuels do try to put it on gmod-devel@lists.sourceforge.net as well?
18:11 deafferret hm. not a member of that cult. how bout you cross post?
18:11 * deafferret cargo cults some more (whatever that means)
18:12 rbuels oh bollocks, it's the same stuff.  there are like 4 people at any given time who actually do open-source bioinformatics software.
18:12 * rbuels is exaggerating
18:12 deafferret hmm? my untrained eye detects ~6 active bioperl committers recently
18:12 deafferret PUT ME IN COACH!
18:13 rbuels the miracle of exaggeration!
18:13 rbuels aha! presto!
18:13 brandi left #bioperl
18:13 deafferret and only 1 "you fucked everything up!" thread recently  :D
18:13 deafferret :o)
18:13 rbuels ho ho ho.
18:13 deafferret lol
18:13 rbuels ok, need to relocate.
18:13 rbuels put car into shop.
18:13 rbuels or attempt to.
18:13 rbuels &
18:13 deafferret aren't these corporial bodies a hassle?
18:13 rbuels no kidding.
18:14 * deafferret chomps more of his jar of blue pills
18:15 * pyrimidine banging head against the wall
18:15 pyrimidine trunk is a mess
18:15 * deafferret suddenly feels good about not contributing squat
18:17 pyrimidine deafferret: yes, but most sane people would run tests prior to commits.  Now I'm getting stupid isa_ok fails all over the place, even after Jason reverted things
18:18 pyrimidine I mean, come on, shit like:
18:18 pyrimidine #   Failed test 'checking returned stream'
18:18 pyrimidine #   at t/RemoteDB/HIV/HIV.t line 91.
18:18 pyrimidine #          got: 'Bio::PrimarySeq=HASH(0x33776a8)'
18:18 pyrimidine #     expected: 'K03455'
18:18 deafferret have the perpetrators been caned?
18:19 * deafferret is guilty of only running those tests that he, in his wildest dreams, think are within the score of small file changes
18:19 deafferret scope
18:19 pyrimidine deafferret: publicly humiliated, sorta (and an apology).  but I can't wait around for someone else to revert their bad code.
18:20 * deafferret has never fiddled anything with broad impact
18:20 deafferret k.  /me sends good vibes to pyrimidine, goes back to worrying about $thisdeadlineoverhere
18:21 pyrimidine np, just growsing...
18:22 perl_splut you might not have, intentionally, but BioPerl does have a bit of spaghetti to its modules
18:23 pyrimidine yes, which makes splitting up things a bit tricky.  Doable, but tricky.
18:25 perl_splut not sure why, but seems odd that Bio::Seq would be a SeqI object, but Bio::PrimarySeq isn't... would make more sense for them to inherit off the same chain... heheh
18:26 deafferret is there a class hierarchy chart anywhere?
18:26 pyrimidine not based on the original design.  Bio::Seq is-a Bio::PrimarySeqI, but only in the manner of retaining the interface methods.
18:26 * deafferret fingers his Graphviz trigger finger
18:27 pyrimidine Bio::Seq delegates to the (internal) Bio::PrimarySeq
18:27 pyrimidine SeqFeatures will be attached to the PrimarySeq, not the Bio::Seq itself.
18:28 pyrimidine Bio::Seq should probably be renamed Bio::SeqRecord, if one wanted to be more specific
18:29 pyrimidine as it contains the sequence (PrimarySeq), features (FeatureHolder), and annotations (AnnotationCollection)
18:30 perl_splut Well, I was thinking of the recent issues caused by the change to SeqIO to default to PrimarySeq instead of Bio::Seq which broke a lot of tests :)
18:31 pyrimidine Right.  Bio::PrimarySeq shouldn't be a Bio::SeqI; it is only supposed to represent the methods and attributes specific to the sequence alone, not the sequence record as a whole.
18:32 pyrimidine separation of concerns
18:33 * pyrimidine seeing :P whenever I type Bio::PrimarySeq
18:33 brunov $mind->blown
18:33 pyrimidine it's like the class is ridiculing me
18:34 brunov great. I cannot unsee it now.
18:37 pyrimidine I think the point about this design is that, if you just want a sequence w/o features and annotation, use a primaryseq.  Once you start adding more to it, you need a larger container class.  The fact that Bio::Seq is-a PrimarySeqI just means that it's delegating to the proper instance.
18:38 pyrimidine In Moose terms, I would use a 'handles' to delegate appropriately to a primaryseq attribute.
18:39 brunov or consume the HasSeq role
18:40 pyrimidine right
18:42 pyrimidine left #bioperl
18:42 pyrimidine joined #bioperl
18:45 pyrimidine though that would probably be more along the lines of a HasPrimarySeq role.
18:47 pyrimidine Anyway, the problem with the current design is there is no obvious distinction between a Bio::Seq and a Bio::PrimarySeq
18:48 pyrimidine b/c inheriting the interface obfuscates the fact that we're actually delegating to an internal primaryseq.
18:50 pyrimidine whereas a role deals with this nicely; does('HasPrimarySeq) vs. isa(PrimarySeq).
18:51 pyrimidine okay, babbled on enough, btw
19:33 driveby_bot joined #bioperl
19:33 driveby_bot /home/svn-repositories/bioperl: r16857 (cjfields) : silence warnings about using next w/o a loop
19:33 driveby_bot Diff: http://tinyurl.com/yk8jrkm
19:38 deafferret rbuels: I posted that wiki page to bioperl-l. you'll have to do the gmod one
19:38 rbuels deafferret: ok i'll have a look
19:46 dnewkirk joined #bioperl
19:48 spekkio joined #bioperl
19:55 driveby_bot joined #bioperl
19:55 driveby_bot /home/svn-repositories/bioperl: r16858 (cjfields) : defaults to Bio::Seq, so unnecessary
19:55 driveby_bot Diff: http://tinyurl.com/ylakq7x
20:25 * deafferret sighs
20:25 * deafferret will eventually learn that ACCCGUA ne ACCCGTA
20:26 deafferret unless of course, when you want it to be
20:28 deafferret \o/   WOOT! finally
20:28 deafferret 2 days of  grrrr
20:28 deafferret 9.9M reasons to be happy finally
20:28 * deafferret boogies
20:45 * pyrimidine loving svn-bisect atm
20:56 * deafferret has never had a real-world case yet...
21:05 driveby_bot joined #bioperl
21:05 driveby_bot /home/svn-repositories/bioperl: r16859 (cjfields) : really rollback florent's commit, was some cruft hanging around causing fasta to break
21:05 driveby_bot Diff: http://tinyurl.com/y94v8ma
21:05 driveby_bot joined #bioperl
21:05 driveby_bot /home/svn-repositories/bioperl: r16860 (cjfields) : really rollback florent's commit, was some cruft hanging around causing fasta to break
21:05 driveby_bot Diff: http://tinyurl.com/yjs5v3u
21:14 deafferret ow. citric acid under fingernail makes typing hurt. stupid grapefruit
21:14 * deafferret goes back to brisquet-only diet
21:16 driveby_bot joined #bioperl
21:16 driveby_bot /home/svn-repositories/bioperl: r16861 (cjfields) : fix bad test file
21:16 driveby_bot Diff: http://tinyurl.com/yjlhlfn
21:21 deafferret pyrimidine: would git bisect help with your rollback? Or are you using it somehow?
21:21 deafferret hard to find the bad commits?
21:21 pyrimidine hard to nail down exactly which commit was causing the problem.
21:22 pyrimidine but I resolved it, just reverted the files to just prior to the problematic commit after I went through with bisect to confirm that commit was bad.
21:23 deafferret using git-svn?
21:24 pyrimidine nope, svn anf App::SVN::Bisect
21:24 pyrimidine s/anf/and/
21:24 deafferret ooo   /me looks
21:25 pyrimidine of course, one could use git-bisect, but since we're still with svn
21:28 deafferret I didn't know there was such a thing.  so   svn-bisect run t/sometest.t  ?
21:28 pyrimidine you would run a clean co.
21:29 pyrimidine Then 'svn-bisect start'
21:30 pyrimidine which starts the bisect.  Run tests, and if good, 'svn-bisect good', if bad 'svn-bisect bad'
21:30 deafferret svn-bisect run   does the good/bad for you, doesn't it?
21:30 pyrimidine rinse and repeat (of course, clean up between builds)
21:30 pyrimidine yes, it can.
21:32 pyrimidine the last batch of tests would have been more problematic to fix.
21:32 pyrimidine s/fix./fix that way/
21:36 driveby_bot joined #bioperl
21:36 driveby_bot /home/svn-repositories/bioperl: r16862 (cjfields) : deal with empty buffer warnings
21:36 driveby_bot Diff: http://tinyurl.com/ybu8p73
21:42 bag joined #bioperl
22:27 deafferret SQLite++ # one fast mo-fo
22:34 pyrimidine left #bioperl
22:38 driveby_bot joined #bioperl
22:38 driveby_bot /home/svn-repositories/bioperl: r16863 (cjfields) : squash warning
22:38 driveby_bot Diff: http://tinyurl.com/yjo2q76
22:51 driveby_bot joined #bioperl
22:51 driveby_bot /home/svn-repositories/bioperl: r16864 (cjfields) : squash warnings
22:51 driveby_bot Diff: http://tinyurl.com/y8ppdev
22:56 driveby_bot joined #bioperl
22:56 driveby_bot /home/svn-repositories/bioperl: r16865 (cjfields) : squash warning
22:56 driveby_bot Diff: http://tinyurl.com/yf266g7
23:01 driveby_bot joined #bioperl
23:01 driveby_bot /home/svn-repositories/bioperl: r16866 (cjfields) : squish verbose output from tests
23:01 driveby_bot Diff: http://tinyurl.com/ya7uurc
23:36 deafferret ha... taking this out back to shoot it   http://clab.ist.unomaha.edu/svn_co​/CLAB/RT922/graphviz/overview.png

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