Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-05-25

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

All times shown according to UTC.

Time Nick Message
00:02 slackaholic joined #bioperl
06:56 kai joined #bioperl
08:27 kai morning folks
08:29 kai how do I get the stuff I print out in a parser using $self->debug() to show up during a test run?
08:43 kyanardag_ joined #bioperl
10:53 vinnana joined #bioperl
13:01 CIA-95 bioperl-live: Chris Fields master * ra32ef7c / (MANIFEST MANIFEST.SKIP): not sure why these aren't in trunk, but add default MANIFEST, MANIFEST.SKIP - http://bit.ly/bqPgMB
13:01 CIA-95 bioperl-live: Chris Fields master * rfbc764a / (11 files in 6 dirs): Merge branch 'master' of github.com:bioperl/bioperl-live - http://bit.ly/bTSgd8
13:59 pyrimidine joined #bioperl
14:00 pyrimidine kai: regarding debug(), you have to set verbose() to 1.
14:01 kai pyrimidine: in the code?
14:01 pyrimidine when you create an instance:
14:02 pyrimidine my $in = Bio::SeqIO->new(-format => 'fasta', -file => 'foo', -verbose => 1);
14:04 kai ah, perfect
14:04 kai thanks :)
14:04 pyrimidine I think you can set global debugging output using something like $Bio::Root::Root::DEBUG = 1
14:04 pyrimidine haven't really tried that
14:05 pyrimidine global will help if you have a factory-like class, which creates lots of other instances
14:05 pyrimidine (so each would have to have -verbose set to the parent)
14:06 pyrimidine prepend that last with 'otherwise, so ...'
14:07 pyrimidine back on in a bit.
14:10 CIA-95 bioperl-live: Chris Fields master * r46f4269 / Build.PL : add no_index meta for PrimarySeqI example (should probably be removed); this will prevent docs in these dirs from being indexed and added to CPAN docs - http://bit.ly/9l92Do
15:41 deafferret Oo.
15:59 ffoertter joined #bioperl
16:32 pyrimidine joined #bioperl
17:48 bag_ joined #bioperl
18:02 rbuels JunY: you need to link your blog/status page to the open-bio wiki
18:03 rbuels JunY: also, on the bioperl wiki
18:03 deafferret and the jays.net wiki
18:04 rbuels well, that's optional
18:06 deafferret :p
18:09 pyrimidine looks like hmmer3 is in it's own repo now.
18:09 pyrimidine http://github.com/bioperl/bioperl-hmmer3
18:14 rbuels pyrimidine: yeah i was wondering about that
18:14 rbuels pyrimidine: kblin wants to work on hmmer3 too
18:14 rbuels pyrimidine: but sharpton isn't on IRC
18:15 rbuels pyrimidine: who created that repo?  does thomas sharpton have the bioperl account password?
18:15 pyrimidine no
18:15 pyrimidine I did it
18:15 rbuels oh, ok
18:16 pyrimidine kblin is added
18:16 rbuels kblin: ping?
18:16 * pyrimidine now thinks bioperl-dev is obsolete
18:16 pyrimidine oh well
18:17 rbuels yes it is.
18:18 pyrimidine the problem we have run into (and part of the reason for it's being) is that we tend to lose things on branches
18:18 pyrimidine jason's array-based SF::Lite for one
18:18 rbuels heh, only when they aren't merged
18:19 pyrimidine exactly
18:19 pyrimidine development will progress to a certain point then stall w/o a merge
18:19 rbuels it's a problem of getting-stuff-finalized, really
18:19 pyrimidine yes
18:20 pyrimidine I will look at some of the code I have in there (mainly primer3-related stuff), will move it out to it's own repo
18:21 pyrimidine probably should do something along these lines with whatever still works in bioperl-ext, at some point
18:21 rbuels yes ....
18:21 pyrimidine (though the io_lib stuff will prob migrate to using Pjotr's biolib bindings)
18:22 pyrimidine ah, if I only had a full year to do all that ;>
18:23 rbuels hah
18:23 rbuels regarding the Big Split, do you think git could do merges right if that happened in a branch of bioperl-live?
18:24 rbuels as in, branch bioperl-live, git mv things around for a while
18:24 pyrimidine don't see why it wouldn't
18:24 rbuels question is, could git merge in from master ...
18:24 rbuels i guess we could just try it and see
18:25 pyrimidine oh, are you thinking a merge from master to a splitting-off branch would create copies on the branch?
18:26 pyrimidine (by 'create copies,' recreate the original files)
18:27 rbuels pyrimidine: i'm not sure if it would or not.  i guess the only way would be to try it
18:28 rbuels pyrimidine: actually i have been thinking that the big split could be semi-automated with some clever scripting
18:28 rbuels pyrimidine: maybe something that would iteratively assist in generating a master spec (some datastructure) describing the split
18:29 rbuels pyrimidine: and that could also do the split when the time comes
18:29 pyrimidine wish there were a way to submodule those, but I don't think that's tenable
18:29 rbuels submodule what
18:29 pyrimidine from what I've been reading, submodules are tricky
18:30 kblin rbuels: pong
18:30 pyrimidine submodule the individual splits
18:30 rbuels oh.
18:30 pyrimidine I think that would blow up in our face
18:30 rbuels i don't know anything about submodules
18:30 rbuels kblin: are you coordinating with thomas
18:30 pyrimidine we still have too many devs getting used to git
18:30 rbuels kblin: there is a bioperl-hmmer3 repo now in bioperl/ apparently
18:31 pyrimidine submodules are pretty advanced
18:31 kblin yeah, so I see
18:31 rbuels kblin: since you guys want the same thing, maybe you should do a teleconference and divide up the work
18:32 rbuels kblin: or maybe you could get him onto irc
18:34 pyrimidine kblin: I've added you to the hmmer3 repo as a collaborator, but it might be better to fork it and do pull requests if you are uncertain of things
18:34 pyrimidine 'things' being 'git'
18:34 kblin yeah, I'll have to talk to thomas about his plans
18:34 kblin I've been using git for years, I'm not too worried about that one
18:35 kblin but still, there's some decisions I don't want to force on the original author
18:36 kblin like I disagree on which field of the output is the correct score for the profile :)
18:49 pyrimidine well, regarding changes, it's always easy to run code review off github
18:49 pyrimidine easier if the changes are via pull request
18:49 deafferret hmm... why is bioperl/bioperl-hmmer3 not a branch of bioperl/bioperl-live ?
18:49 kblin yeah
18:49 pyrimidine deafferret: b/c we're trying very hard not to add new code to core ATM
18:49 deafferret orly?
18:50 pyrimidine specifically, code that can be packaged as standalone
18:50 kblin pyrimidine: arguably this makes development a bit less convenient
18:50 deafferret um... as Bio::HMMER ?
18:50 deafferret or Bio::SearchIO ?
18:50 pyrimidine well, that's the problem
18:51 pyrimidine how do we split these up
18:51 pyrimidine (by namespace or by app?)
18:51 deafferret Bio::Stuff13712 ?  :)
18:52 pyrimidine kblin: by less convenient, what do you mean?
18:52 pyrimidine splitting up, or forking?
18:52 kblin well, if I just branch of the main repository (as I did for my tests at work earlier today) I've got all the stuff I need in one place
18:52 pyrimidine right
18:53 pyrimidine however, that's the current problem with core right now
18:53 pyrimidine everything (all 800 or so modules) are in one place
18:54 kblin so how would I build and deploy the hmmer3 repository right now?
18:54 rbuels i kind of think developing this into bioperl-live isn't a problem, since it's not adding modules, is it?
18:54 kblin my current plan was to rsync over the changes into bioperl-live and then do testing and building there
18:55 pyrimidine if it's a branch of bioperl-live, with the thought of merging it in, yes, it would be adding module
18:55 pyrimidine *modules
18:56 pyrimidine is there a specific problem with using PERL5LIB, or some other means of development?
18:57 kblin I guess not, just different development philosophies clashing, I guess. I'll play along :)
18:57 Lynx_ joined #bioperl
18:58 kblin I see the point of wanting to point people who just need Bio::SearchIO to a separate repository at some point
18:59 pyrimidine yes, that's the point
18:59 pyrimidine well, eventually
18:59 pyrimidine I'm completely open to hearing of alternatives, just not so much if it means lumping new code back into core
19:00 pyrimidine if it truly belongs in core, we can put it there
19:00 pyrimidine but if it is pretty self-contained (and it seems it is)
19:00 pyrimidine then maybe it belongs in it's own repo
19:00 pyrimidine has it's own VERSIOn
19:00 pyrimidine etc
19:01 kblin I guess I don't know enough about packages in perl to judge this
19:01 rbuels well i thought the plan for this one was just to enhance the existing hmmer.pm?
19:01 pyrimidine (and, really, we could just as easily move hmmer2 code into it, and call it Bio::HMMER)
19:01 rbuels or is thomas thinking a new module?
19:02 pyrimidine I think it's a new module
19:02 kblin it's a new module so far
19:02 rbuels so that's another question that needs to be resolved, whether this needs a new module
19:02 rbuels that's between kblin and thomas ....
19:02 rbuels but the answer affects whether it needs to go into a separate email
19:02 rbuels er, repo
19:03 rbuels sorry, wires crosses
19:03 rbuels d
19:03 * rbuels does about 8 things at the same time
19:03 pyrimidine same here
19:03 pyrimidine working on DBIC at the same time
19:04 kblin well, while I had trouble getting my tests to pass with the hmmer3 code, I played around with adding hmmer3 support to the existing Bio::SearchIO::hmmer parser
19:04 kblin I got the initial things to parse, but the slight differences between the formats certainly make for ugly regexes
19:05 kblin or for two if ( /pattern/ ) { dostuff } blocks right after each other pretty much doing the same thing
19:05 pyrimidine make it pluggable
19:06 pyrimidine I do something like that with infernal
19:06 pyrimidine (though it could be prettier)
19:07 pyrimidine something like: 'my $searchio = Bio::SearchIO->new(-format => 'hmmer', -version => 3, -file => 'foo');'
19:07 deafferret if this enhancement is going to be a step towards killing the monolith, then it seems to me that either (1) all things HMMER should break off of bioperl-live, or (2) all things SearchIO should break off of bioperl-live
19:09 deafferret personally, I see this as "enhance the existing".   but if I had to choose (1) or (2), i'd say (2) for the sake of maintaining code consistency
19:09 pyrimidine When initializing, call the correct parser based on the version (defaulting to some version), either from within next_result or loading it dynamically
19:09 pyrimidine deafferret: that's a possibility (it would be Search and SearchIO)
19:09 kblin pyrimidine: yeah, looking at the SearchIO::infernal stuff right now
19:10 kblin that should be doable
19:10 pyrimidine like i said, it could be prettier
19:10 kblin it can always be prettier ;)
19:10 pyrimidine :)
19:10 pyrimidine also, look at the way blastxml works
19:11 pyrimidine it calls the proper handler based on whether it is regular or psi-blast-based XML
19:11 pyrimidine (there is a subtle difference btwn the two in the way results are created)
19:21 kblin hm, I can't run the tests on the bioperl-hmmer3 repository, though
19:22 kblin prove complains about being unable to locate Test/Warn.pm
19:23 kblin http://codepad.org/uIBMWzpf
19:23 pyrimidine do you have it installed locally?
19:24 kblin PERL5LIB is set to /data/local/lib/perl5 and /data/local was the install-base for the Build.PL call
19:25 pyrimidine If you don't have it installed locally, Bio::Root::Test looks for Test::Warn in the local dist
19:25 pyrimidine (though I think it looks for it in a weird spot, probably should be ./inc or somesuch)
19:25 pyrimidine oh
19:26 pyrimidine that repo thomas checked in is lacking a few key things
19:26 kblin I guess so
19:26 pyrimidine such as Build.PL, proper tests, etc.
19:26 pyrimidine you know, all the things that would make it self-contained
19:27 kblin well, I've got a "proper test" that follows the bioperl-live conventions
19:27 pyrimidine well, see how far you get with integration into current hmmer code in core.  we can work it in there, and (if needed) worry about splitting SearchIO out later
19:28 pyrimidine kblin:  that's good
19:28 kblin ok, hang on, I'll push my updates so far to github
19:28 kblin to my fork
19:28 pyrimidine Bio::Root::Test is only a convenient wrapper around Test::More/Warn/Exception
19:29 kblin http://github.com/kblin/bioperl-hmmer3
19:30 kblin mainly the stuff in t/
19:33 pyrimidine went ahead and applied it
19:34 kblin thanks
19:35 pyrimidine I set it up with CIA and dev mail hooks, so commits will show up here
19:36 kblin nice
19:43 CIA-95 bioperl-live: Adam Sjøgren master * rfcd90e0 / (4 files in 3 dirs): Return features as the same type of object as the sequence they come from. - http://bit.ly/agvwsF
19:43 pyrimidine um, that wasn't supposed to happen
19:44 pyrimidine Adam's code was a fundamental change
19:48 rbuels pyrimidine: i can revert and move it to a branch
19:49 rbuels pyrimidine: or you can
19:49 rbuels pyrimidine: or maybe it works ...?
19:49 pyrimidine I'm sure it works.
19:50 pyrimidine the problem is, it's automatically using self instead of the contained primaryseq
19:50 pyrimidine (so automatically getting a circular ref)
19:50 pyrimidine yes, we weaken it, but then if we go down this path, this makes primaryseq obsolete
19:51 pyrimidine or at least, redundant
19:51 pyrimidine should clarify
19:51 pyrimidine if you attach a SeqFeature in Bio::Seq
19:51 pyrimidine it attaches the Bio::Seq's contained PrimarySeq by default
19:52 pyrimidine this change makes the default behavior attach the Bio::Seq to the SeqFeature
19:52 pyrimidine Bio::Seq <-> SF (circ ref)
19:53 pyrimidine so, that's a possibly fundamental change
19:56 pyrimidine dave saw my comment, is backing it out
19:59 CIA-95 bioperl-live: Adam Sjøgren topic/adsj-seqobj-return * rfcd90e0 / (4 files in 3 dirs): Return features as the same type of object as the sequence they come from. - http://bit.ly/agvwsF
20:00 rbuels pyrimidine: backing it out like that is not good
20:00 rbuels pyrimidine: why isn't he on irc
20:01 pyrimidine not sure
20:03 pyrimidine rbuels: we need to document how to do this the 'correct' way
20:04 pyrimidine otherwise we'll run into this again
20:04 pyrimidine should just be git revert HEAD, correct?
20:04 rbuels if HEAD is that commit, yes
20:04 rbuels otherwise, git revert <that sha>
20:04 pyrimidine right
20:07 pyrimidine one of the problems with asking someone to go to IRC is many people don't like it
20:07 pyrimidine requires a specific mindset
20:14 _ilbot2 joined #bioperl
20:14 Topic for #bioperlis now http://www.bioperl.org - You can get help here, but you must be patient.  Paste to http://codepad.org/
20:15 CIA-95 bioperl-live: DaveMessina revert-fcd90e0 * r6a3d9f4 / (4 files in 3 dirs):
20:15 CIA-95 bioperl-live: Revert "Return features as the same type of object as the sequence they come from."
20:15 CIA-95 bioperl-live: This reverts commit fcd90e0f2fa94b61ff8351157129678417c32991. - http://bit.ly/ciPzZC
20:15 pyrimidine \o/
20:16 CIA-95 bioperl-live: DaveMessina master * r6a3d9f4 / (4 files in 3 dirs):
20:16 CIA-95 bioperl-live: Revert "Return features as the same type of object as the sequence they come from."
20:16 CIA-95 bioperl-live: This reverts commit fcd90e0f2fa94b61ff8351157129678417c32991. - http://bit.ly/ciPzZC
20:17 pyrimidine and done on a branch, to boot
20:17 rbuels those cia messages are not perfect.
20:18 dave_messina joined #bioperl
20:19 rbuels ah, well there he is
20:19 dave_messina Hey guys, thanks for your help.
20:19 rbuels dave_messina: no problem
20:19 dave_messina Sorry, forgot about IRC
20:19 rbuels oh, it's great fun.
20:19 dave_messina :)
20:19 rbuels lots going on in bioperl today actually
20:20 rbuels lots of discussion about the hmmer3 thing that kblin and thomas sharpton are working on for example
20:20 dave_messina Yeah yeah, I'm excited about that. I expect to be playing with H3 soon.
20:24 rbuels the adsj-etc branch, you might have a little bit of an interesting time merging that
20:24 rbuels because git is going to think it's already merged
20:25 rbuels i think
20:25 rbuels because the commit that's in it is already in master
20:25 dave_messina Well, that's not good.
20:25 * rbuels tests this hypothesis
20:25 rbuels well the solution is probably just to pull from master (to get the revert commit) and then revert the revert
20:25 rbuels lol
20:25 rbuels but maybe it's fine ...
20:25 * rbuels tests
20:25 pyrimidine live and learn
20:26 pyrimidine I've added a bit to the using git page on reverting commits
20:26 dave_messina and it sounded so convenient in the little tutorial I was following...
20:26 * dave_messina saw that chris, thanks.
20:27 pyrimidine np
20:27 rbuels well it's convenient locally, with history rewriting and all that ...
20:27 dave_messina btw, chris, did you happen to look at that Bio:Species unclassified seq thing that Charles Tilford reported a few days ago?
20:29 dave_messina I ask because when I glanced at the Taxonomy code, it sure looked like there was support for that already.
20:29 pyrimidine dave_messina: no
20:29 pyrimidine it should be reported as a bug, though, if it pans out
20:30 pyrimidine just between us here (and, of course anyone who reads the backlog)
20:30 pyrimidine I hate Bio::Species
20:30 CIA-95 bioperl-live: Robert Buels master * r10fff64 / examples/longorf.pl : docpatch to examples/longorf.pl - http://bit.ly/9VOEU8
20:30 CIA-95 bioperl-live: Robert Buels master * r4073c12 / t/Seq/PrimarySeq.t : test fix: if this translate() is supposed to die, it should check $@ either way! - http://bit.ly/9durWS
20:31 dave_messina pyrimidine: me too.
20:31 pyrimidine It has been the source of more headaches
20:31 pyrimidine and it's now integrated with Bio::Tree
20:31 pyrimidine (which makes splitting that code out tricky)
20:32 dave_messina ooh, that's right.
20:32 CIA-95 bioperl-live: DaveMessina topic/adsj-seqobj-return * r6a3d9f4 / (4 files in 3 dirs):
20:32 CIA-95 bioperl-live: Revert "Return features as the same type of object as the sequence they come from."
20:32 CIA-95 bioperl-live: This reverts commit fcd90e0f2fa94b61ff8351157129678417c32991. - http://bit.ly/ciPzZC
20:32 CIA-95 bioperl-live: Robert Buels topic/adsj-seqobj-return * r69fb662 / (4 files in 3 dirs):
20:32 CIA-95 bioperl-live: Revert the revert, so this branch will merge easily eventually.
20:32 CIA-95 bioperl-live: This reverts commit 6a3d9f48d56fb928e3ca908e6113f1bfa4ed2405. - http://bit.ly/bzhEiF
20:32 rbuels ok, the adsj-* branch should be good to go now, it is now mergable.
20:32 rbuels mergeable?
20:32 rbuels mergeabilee?
20:33 * rbuels chuckles
20:33 dave_messina mergetastic
20:33 dave_messina thanks rob.
20:33 pyrimidine mergealicious
20:33 rbuels dave_messina: np
20:33 pyrimidine wow, deafferret disappears when work gets done around here ;-)
20:33 dave_messina Hey, do you guys know if it's possible to join the mailing list from the Google groups side?
20:34 pyrimidine should be
20:34 * rbuels chuckles
20:35 kblin dave_messina: the current hmmer3 code should be useable. at least my test case seems to run ok with it
20:35 dave_messina hmm, that's what I thought. I got an email from somebody who tried to post from there but got bounced into the moderator queue.
20:35 dave_messina Is Mark the ferret?
20:35 pyrimidine oh, that's possible.  I haven't received any notice about it.
20:35 dave_messina Funny, I never pictured him that way.
20:35 pyrimidine nah, deafferret is Jay
20:36 kblin I'm just running on real data, will probably be able to tell if it worked in a day or two
20:36 dave_messina Chris: ah, okay, I'll ask him to go through mailman then.
20:37 dave_messina kblin: great! I saw your commits. thanks for putting it through its paces.
20:37 pyrimidine does the hmmer3 code include wrappers?
20:37 pyrimidine or do the current ones work?
20:37 pyrimidine *one
20:37 kblin pyrimidine: I think I saw one in Thomas code, haven't tried that, though
20:37 pyrimidine ok
20:38 dave_messina Yeah, I think there are some changes to the interface between H2 and H3.
20:38 kblin $dayjob has a fixed pipeline with shell scripts that run the actual commands
20:38 pyrimidine oh, that's nothing
20:38 kblin dave_messina: half the commands are named differently
20:38 pyrimidine you shuold see the difference btwn primer3 v1 and v2.
20:38 kblin dave_messina: a bit like blast -> blast+
20:39 dave_messina aha, gotcha.
20:39 pyrimidine yeah, that's about right
20:39 dave_messina blech, primer3. worst. interface. ever.
20:39 pyrimidine yep
20:39 dave_messina more config options than the cpan shell.
20:39 pyrimidine heh
20:39 kblin that said, the blast+ command line options are much less cryptic
20:40 dave_messina agreed. they're like wu-blast. :)
20:40 dave_messina or blast-1.4, if you prefer.
20:40 pyrimidine less cryptic, and a lot longer
20:40 dave_messina I gotta tell ya, I love that --remote option in blast-plus.
20:40 kblin pyrimidine: usually one implies the other, until people find getopt
20:41 dave_messina the future is abstracting away local and remote computing.
20:41 kblin dave_messina: downside on hmmer3 is that the output format is different enough to break the parser
20:41 pyrimidine kblin: true.  Sometimes can't have it both ways.
20:41 kblin dave_messina: bonus point is that hmmer3 is fast..
20:42 pyrimidine kblin: that was the same problem with infernal for a while
20:42 dave_messina kblin: breaks the H2 parser?
20:42 pyrimidine until it stabilized at v.1, after a few years or so of development
20:42 kblin dave_messina: well, not enough to make it crash or give decent feedback. it just silently finds no hits :)
20:43 dave_messina right, but the H3 output should be stable now, right?
20:43 kblin yeah
20:44 dave_messina ah, good. But yeah, silent failure is not so good.
20:44 kblin and the output of hmmscan (used to be hmmpfam) and hmmsearch are much closer together now
20:44 kblin s/output/output formats/
20:45 dave_messina that's good.
20:45 scott___ joined #bioperl
20:46 kblin and I'm really impressed by the speedup
20:46 scott___ howdy--where can I find info about the pre bosc hackathon?  Didn't see anything on the wiki.
20:46 kblin my test sequence used to churn in hmmer2 for ~30 minutes, with hmmer3 it's done in about a minute
20:46 kblin or ~3 minutes if I run it against both the Pfam-A and Pfam-B profiles
20:48 kblin and looking at the top output, it'd probably be faster if I had a faster hard drive
20:48 dave_messina Awesome. Yeah, Sean Eddy's goal is for H3 to replace Blast, and they are now more or less equivalent in speed.
20:48 dave_messina (or so say the benchmarks I've seen, anyway.)
20:49 kblin I'm planning to try that once I've pushed a release of my toolchain out of the door with blast+ and hmmer3 support
20:49 pyrimidine scott___: http://www.open-bio.org/wiki/Codefest_2010
20:50 scott___ Ah, I was looking at the wrong wiki--thanks!
20:56 pyrimidine np
21:04 deafferret wow! look at all that backlog!   i'm missing the party  :(
21:05 dave_messina Yo DF!
21:09 deafferret LOL   :p    $work[0] rollout  :(   <<@pyrimidine> wow, deafferret disappears when work gets done around here ;-)
21:10 pyrimidine :)
21:16 deafferret dave_messina: nice to see ya! you should hang out more. i have 'jokes'   :)
21:16 dave_messina good! then I will 'laugh'. :)
21:17 pyrimidine then i will 'work
21:17 pyrimidine bah
21:17 pyrimidine *'work'
21:17 * pyrimidine having a fat-finger day
21:38 pyrimidine left #bioperl
22:13 spekki01 whats the easiest way to split a protein db file in fasta format into a smaller chunk say the first 1000 sequences? I'm thinking use bio::alignIO read it in and write it out in the same format but for 1000 iterations?
22:14 kblin Use SeqIO, and run the loop 1000 times?
22:15 spekki01 kk
22:16 kblin not sure if that's the best solution, but it sure works :)
22:24 spekki01 hmm im not sure what im doing wrong here but is this the general idea http://codepad.org/30ZWgLMI
22:25 rbuels spekki01: i think that will work, as long as your input has more than 1000 seqs ..
22:26 rbuels spekki01: are you getting any error messages?
22:26 spekki01 it does, im running it now and im getting http://codepad.org/urH9N9T5
22:27 rbuels spekki01: missing semicolon after next_seq
22:27 spekki01 tx :)
22:27 rbuels put on the daddy pants, don't get stumped by syntax errors
22:28 rbuels they're not the boss of you
22:28 rbuels where's deafferret, my silliness is distinctly second-rate
22:28 spekki01 somedays thought you just don't catch them and it pulls you down lol
22:28 spekki01 lol
22:39 buelsbot joined #bioperl
22:41 deafferret rbuels: you confused 'silly' with 'condescending jackass'
22:41 deafferret beginner mistake. you'll get better.
22:41 rbuels dammit i was trying to be encouraging!  like a scantily-clad cheerleader!
22:42 * rbuels sighs
22:42 deafferret oh. hmm... been 15 years since I was in such a relationship, so I missed it.
22:42 deafferret apologies
22:43 rbuels accepted!!!
22:43 * rbuels flounces off
22:44 deafferret $wife has completely ruined my flirty/sassy radar
22:44 deafferret she's all business
22:45 spekki01 k im havving some major issues using hmmer3, im trying to do a basic hmmsearch but the database file im doing the search on is around 6gigs and its taking way to long it ran for 4 days and was still searching, ive tested this all on a smaller version of the same database and it works fine so, is there anyway to get this hmmsearch to work on such a big db file?
22:45 rbuels spekki01: are you running out of memory?
22:45 rbuels spekki01: is the machine swapping?
22:45 spekki01 not that im aware of
22:45 deafferret vmstat 3
22:47 * deafferret heads back to topic/rbuels-CONFIDENTIAL
22:47 * rbuels is suspicious
22:51 spekki01 hmm sry just trying to find a similar command for osx
22:52 spekki01 im running mac servers with darwin
22:52 spekki01 not my choice...
22:52 * rbuels tries to keep a straight face
22:52 deafferret Mac pride! woot! looking...
22:52 deafferret (um... i only use it for desktop tho)
22:53 deafferret spekki01: run   top    and watch VM:  pageins, pageouts
22:53 deafferret you want 0
22:53 deafferret (or at least numbers that aren't moving)
22:53 spekki01 kk
22:54 spekki01 gimme a moment
22:54 deafferret dunno what 58949(0) means
22:55 deafferret hmm, this adobe meeting gizmo in firefox is burning 20% of my CPU even while idle   :/
22:55 deafferret spekki01: curse you! you've made me hate my mac!
22:55 spekki01 lol
22:56 deafferret soon my 'manic' phase will return
22:56 deafferret then,  hoo-boy!  brace yourselves !   hilarity will ensue!
22:56 spekki01 VM: 4336M + 127M   33770(0) pageins, 6912(0) pageouts
22:57 deafferret are the pageins, pageouts numbers moving around?
22:57 spekki01 that mean anything?
22:58 spekki01 doesnt look like it there still the same as what i posted
22:58 spekki01 86728 hmmsearch  186.2%  5:59.85   3    33     70   68M-  196K    68M+   86M
22:59 spekki01 thats the hmmsearch line
23:00 deafferret have they written a parallel version of that thing?
23:00 * deafferret bats his eyelashes at eddy
23:00 spekki01 it runs on multiple cpu's i know
23:00 deafferret ya but can you throw it at a server cluster as 1000 separate jobs?
23:00 spekki01 well i gotta get going my ride home is here I'll try figure this out tomorrow, thanks deafferret
23:01 deafferret sure
23:01 deafferret try some aloe on the Mac. they tend to like that
23:01 deafferret a happy Apple is a better worker bee

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