Camelia, the Perl 6 bug

IRC log for #bioperl, 2011-04-27

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

All times shown according to UTC.

Time Nick Message
03:04 svaksha left #bioperl
03:05 svaksha joined #bioperl
03:15 genehack pyrimidine, rbuels: there's a 'git submodules foreach' command that makes it pretty trivial to update all the submodules to whatever their HEAD is
03:16 genehack (FWIW)
03:40 barbae_inpexae left #bioperl
08:39 flu left #bioperl
08:42 flu joined #bioperl
11:51 svaksha left #bioperl
12:08 svaksha joined #bioperl
12:43 Lynx_ left #bioperl
12:53 zenman_ joined #bioperl
13:01 Lynx_ joined #bioperl
13:15 zenman_ left #bioperl
13:50 pyrimidine genehack: that's worth bringing up.  The difficulty we'll have is getting the devs who are still new to git to use 'git submodule' and not bork the works, so to speak
13:52 genehack pyrimidine: the nice thing is if you're in a submodule directory, it's pretty much just being like a checkout of that repo -- worse case you have to 'git co master' to get onto the current HEAD
13:52 genehack and i can't ever remember the syntax for the foreach stuff so i tend to have a 'bin/update-submodules' script as a crutch
13:55 pyrimidine genehack: right, I agree.  And I think it's worth exploring (particularly in this case, as it's a way of bundling related-but-separate code).  It's just getting the other devs on board
13:56 pyrimidine genehack: believe me, a few of them still think git is a fancier or more complex SVN.  It took a little discussion at a hackathon last fall to show them the light, so to speak
13:56 pyrimidine and even then I think they were still a little lost
13:58 pyrimidine not to mention we have a few devs who seem to hate git with the white-hot passion of a thousand suns
14:00 genehack yeah, getting change in basic tools like that can be hard. 8^/
14:00 pyrimidine (a phrase which, out of curiosity I just googled and found the Hindu origin of, made popular via Robert Oppenheimer)
14:01 pyrimidine yes, well, we did actually post about these changes quite a bit prior to the actual move (and we hate sourceforge with a similar white-hot passion...)
14:04 genehack yeah, i would be in the "oh god not sourceforge" camp. happy to not need to use that other than the occasional download
14:05 ank left #bioperl
15:56 carandraug joined #bioperl
16:13 svaksha left #bioperl
16:33 svaksha joined #bioperl
16:40 dbolser rbuels: is there any milage to be had in coercing Bio::GFF3::LowLevel into outputting something that can be fed into Bio::SeqFeature::Generic directly?
16:41 dbolser your manpage says 'These functions do no validation, do not reconstruct feature hierarchies, or anything like that. If you want that, use Bio::FeatureIO.', which is fair enough, but FeatureIO says:
16:41 dbolser cpan slow..
16:42 dbolser while we're waiting, my usecase is this: 1) parse GFF, 2) create a Bio::SeqFeature::Generic or in fact a Bio::Location::Simple, 3) feed into a Bio::Coordinate::Collection representing my AGP, 4) write out a GFF of the feature
16:43 dbolser thought I'd check out Bio::GFF3::LowLevel for reading writing GFF, but of course I can't pass a 'naked' hashref to Bio::SeqFeature::Generic ...
16:43 dbolser do I just need to bless it as such?
16:45 dbolser here goes...
16:46 rbuels dbolser: if you want to write gff3, use the new Bio::FeatureIO
16:46 rbuels dbolser: (if you want to write gff3 using bioperl objects)
16:46 dbolser MSG: I need a Bio::Location, not [Bio::SeqFeature::Generic=HASH(0x96a718c)]
16:46 dbolser NOTE: This is a highly experimental I/O iterator subsystem for genomic sequence features. It is not complete, and is now undergoing a significant refactoring in a separate branch of BioPerl to address specific issues regarding the current implementation. Use at your own risk.
16:46 rbuels dbolser: https://github.com/bioperl/Bio-FeatureIO
16:46 dbolser is that just to ward off newbs?
16:47 rbuels dbolser: stop using that, and use the split-off new version on github
16:47 dbolser ok
16:47 dbolser nice
16:47 dbolser and this uses GFF::LowLevel under the hub?
16:47 rbuels yes, actually
16:47 * dbolser seems to remember...
16:47 dbolser nice
16:47 dbolser thanks
16:48 * dbolser awards himself the most bioperly discussion of the week award
16:48 * dbolser reads the backscroll and strips his award
16:49 dbolser rbuels: have I told you you're my hero?
16:49 dbolser Bio-FeatureIO++
16:49 rbuels dbolser: for what?  fixing the gff3 writing in that module?
16:50 dbolser for fixing 'feature' handling in bioperl (or at least starting that process)
16:50 dbolser 12 seconds left on my work timer
16:50 dbolser night
16:50 rbuels dbolser: yeah, it won't really be fixed until things are split off to the point where this can be swapped in
16:50 dbolser yup
16:50 rbuels dbolser: ok night
16:51 dbolser if I follow the INSTALLATION in the README, will it take precidence over the BioPerl version?
16:52 barbae_inpexae joined #bioperl
16:52 dbolser !  perl (5.8.8) is installed, but we need version >= 5.10.0
16:52 * dbolser cries all the way home
16:52 barbae_inpexae left #bioperl
17:30 takadonet left #bioperl
17:43 * genehack gives dbolser a cookie and a copy of perlbrew
17:43 genehack it's going to be ok.
19:42 pyrimidine it's amazing what a little modern perl (perlbrew or local::lib in this case) can acomplish
19:42 pyrimidine *accomplish
19:56 genehack indeed -- gugod++ # perlbrew rocks
20:04 pyrimidine would go as far as saying, using perlbrew or local::lib is essential for anyone using perl in an environment where they aren't admin (cpanm also helps)\
20:28 carandraug left #bioperl

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