Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-08-12

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

All times shown according to UTC.

Time Nick Message
00:01 carandraug joined #bioperl
01:44 Yksi joined #bioperl
05:50 bag_ joined #bioperl
09:14 Yksi joined #bioperl
09:48 kblin hm, anybody still got hmmer 2.x around?
10:09 dbolser not sure
10:09 dbolser FATAL: No such option "--version".
10:30 kblin hehe
10:30 kblin do you have a hmmpfam command?
10:32 kblin that only exists in hmmer 2.x, in hmmer3.x it's called hmmscan
12:08 kblin but it looks like the cli didn't actually change miuch
12:27 Yksi joined #bioperl
12:32 Jun joined #bioperl
12:32 Jun Hi, I got a problem in running bioperl test
12:33 Jun I want to do ./Build.PL test, but it says "cannot creat '_build/cleanup', permission denied"
12:35 Jun Then I tried "sudo ./Build.PL test", it says "too early to specify a build action 'test'. Do 'Build test' instead"
12:35 kblin hehe, great...
12:36 Jun Do you know what is wrong here?
12:39 brandi joined #bioperl
12:39 brandi left #bioperl
12:49 kblin ok, first, try and remove all generated files owned by root
12:50 kblin ./Build.Pl should only be run to create the "Build" script
12:50 kblin then, to run the tests, call "./Build test"
12:52 Jun ok, seems complicated...
12:52 kblin I've seen worse :)
12:52 kblin you run "perl Build.PL"
12:53 kblin once
12:53 kblin and then you only run "./Build" afterwards
12:53 Jun ok, i will do that
12:53 Jun Do I need to install all optional external modules?
12:55 kblin no, unless you need any of them for the modules you want to work on
12:55 Jun ok, i selected all default setting. now "perl Build.PL" failed. Because "use of uninitialized value $mods in split at Bio/Root/Build.pm"
12:56 Jun but anyway, it works now ...
12:56 Jun Thx a million, Kai
12:59 kblin no problem. how's the gsoc coming along?
12:59 Jun It is really good. It is just I am running out of time...
13:00 kblin right, it's over next monday, irght?
13:00 Jun So what I can do now is to summarize what I've done so far, e.g. making tests, documentaton, and leave what is not done in September.
13:01 Jun Yep, next monday is the hard pencil down deadline
13:44 paul__ joined #bioperl
14:09 paul__ hi...I posted a pastebin concerning a problem with BlastPlus: http://pastebin.com/E9nPRNAc
14:11 kblin does it work from the command line?
14:12 kblin if you call blastn directly?
14:18 paul__ lemme check
14:25 paul__ blastn.exe -db nt -remote -query testquery.fa
14:25 paul__ from my command line works fine
14:25 kblin can you, for kicks try...
14:25 kblin blastn.exe -db ./nt -remote -query testquery.fa
14:25 paul__ i think the problem is that BioPerl is sending the value './nt' to NCBI as the database when it should be 'nt'
14:26 kblin yeah
14:26 paul__ yeah, i get the error when using your command line
14:26 paul__ "BLAST Database error: Nucleotide BLAST database './nt' does not exist in the NCB I servers"
14:26 kblin ok, so that looks like a bioperl bug
14:27 paul__ seems like that should have been caught
14:27 paul__ by somebody already
14:27 paul__ since that is the most basic blast you can do
14:28 kblin well, people might not be using remote blast
14:28 paul__ hmm...i guess so
14:28 paul__ how do I fix it?
14:28 kblin currently looking at the code
14:28 kblin how did you install bioperl-run?
14:29 paul__ from git
14:29 paul__ i just did a git pull this morning
14:30 paul__ so code is up to date
14:30 kblin ok
14:30 paul__ what files/subroutines are you looking at so I can learn too?
14:30 kblin lib/Bio/Tools/SandAloneBlastPlus.pm right now
14:32 kblin something there fiddles with the db path
14:35 paul__ yeah, db_path is ./nt
14:35 paul__ probably don't need to build a db path if the blast is remote
14:35 kblin yeah
14:36 kblin I'm guessing line 533ff
14:37 paul__ yeah, it is the if/else on line 540 that is trying to build a path to the database
14:37 kblin I don't have blastplus installed on this machine.. can you rerun your script using -verbose => 1in the new?
14:37 kblin "-verbose => 1"
14:38 paul__ ok
14:38 paul__ output looks the same
14:39 kblin yeah, I just noticed there's no debug statements in the interesting code :)
14:39 paul__ http://pastebin.com/7cbbc47g
14:39 paul__ i c
14:42 kblin ok, let's go to lib/Bio/Tools/StandAloneBlastPlus/BlastMethods.pm
14:43 kblin in line 251, $blast_args{-db} is set
14:43 paul__ ok i'm there
14:44 paul__ need to set to db if remote is true, correct?
14:44 kblin the plan is to remove a leading ./ if is_remote is true
14:44 paul__ I mean db_name
14:44 paul__ ah
14:45 kblin not the best way to do it, but should be pretty straightforward
14:46 paul__ can you update BlastMethods.pm with a fix?
14:47 kblin I'll do a patch and let you test it, if that works for you
14:47 paul__ ok
14:47 Yksi1 joined #bioperl
14:50 kblin man, my perl-fu still sucks :)
14:52 kblin there we go
14:55 paul__ i did a git pull but nothing happened
14:56 kblin didn't push anything yet
14:56 paul__ oh sorry
14:56 kblin still waiting for the pastebin to load, lousy connection here
14:58 kblin hm, wait, I can do better than that
14:59 paul__ shouldn't I test it first before pushing? can you paste the code fix? I'll change my BlastMethods file and test it
15:00 kblin yeah
15:00 kblin http://pastey.net/139508
15:01 kblin use the download&save link, and then apply the patch using "git am < path/to/patch"
15:03 Yksi joined #bioperl
15:04 paul__ I typed "git am < ~/desktop/pastey-139508.diff" and i got errors
15:04 paul__ I'm in the StandAloneBlastPlus directory
15:05 paul__ should I post the output to pastebin?
15:05 kblin ah, I think you need to be in the top level dir
15:05 kblin in factm the output should say that
15:07 paul__ ok i'm in bioperl-run directory
15:07 paul__ got similar error
15:08 paul__ http://pastebin.com/U615fh0q
15:14 paul__ i think i figured it out.....
15:15 paul__ the line endings were dos and I'm working in Cygwin so I changed line endings to UNIX and the patch seems to have worked
15:15 paul__ shoot, blastmethods.pm didn't update
15:15 paul__ i didn't get errors when running the 'git am'
15:16 paul__ I got this though: previous rebase directory /cygdrive/c/cygwin/home/Pau​l/downloads/bioperl/bioperl -run/.git/rebase-apply still exists but mbox given.
15:20 kblin ah
15:20 kblin you need to do git am --abort
15:20 paul__ i manually updated
15:20 kblin that aborts.. ah, or that
15:20 paul__ but am getting a non-lvalue error
15:21 kblin hmm
15:21 paul__ "Can't modify non-lvalue subroutine call"
15:21 kblin ah, that's a call?
15:21 kblin dammit :)
15:21 paul__ $self->db_path =~ s#./##g
15:21 kblin hang on
15:21 paul__ how about this....
15:21 paul__ my $db = $self->db_path; $db =~ s#^./## if $self->is_remote; $blast_args{-db} = $db;
15:23 paul__ my code above seems to work
15:23 paul__ is my code good enough to use for updating BlastMethods.pm?
15:25 paul__ git diff at http://pastebin.com/HdqD2w1K
15:26 kblin currently rfixing my patch
15:26 kblin http://pastey.net/139510
15:28 kblin my wireless keeps dying on me
15:29 kblin yeah, same difference
15:29 kblin if your patch works for you, keep it
15:29 paul__ currently testing your code now
15:30 paul__ it works!
15:30 paul__ are you going to do a push?
15:30 kblin ok, one last change..
15:31 kblin can you make the regex say s#^\./##
15:32 kblin I'm still a bit new to this perl stuff
15:33 paul__ hold on a min ... bbiab
15:33 kblin sure, let me know when you've got time
15:37 paul__ ok ...testing $db =~ s#^\./##; now
15:37 paul__ it works
15:38 kblin great
15:38 kblin I'll push that
15:41 kblin want to give me your name so I can credit you in the commit message?
15:41 kblin <paul__> from IRC always looks a bit weird, I think
15:41 paul__ Paul Cantalupo
15:41 paul__ thank you for all your help
15:41 paul__ lol yeah
15:43 CIA-95 bioperl-run: Kai Blin master * re817288 / lib/Bio/Tools/Run/StandAlon​eBlastPlus/BlastMethods.pm :
15:43 CIA-95 bioperl-run: StandaloneBlastPlus: don't automatically add ./ to remote database names
15:43 CIA-95 bioperl-run: Thanks to Paul Cantalupo for reporting and testing this. - http://bit.ly/alj7M6
15:43 kblin there we go
15:44 * kblin gives CIA-95 a cookie
15:46 pyrimidine joined #bioperl
15:49 pyrimidine Jun: that uninit warning you saw during 'perl Build.PL' is fixed on master in main bioperl repo
15:51 kblin hey pyrimidine
15:53 kblin pyrimidine: I realized that I'd like BioPerl better if it wasn't Perl
15:53 pyrimidine :)
15:53 pyrimidine how about if it were Moose?
15:53 pyrimidine or Perl 6?
15:54 kblin not sure
15:54 kblin haven't seen those really
15:55 pyrimidine what do you routinely use (lang-wise)?
15:55 kblin I've got a couple of years of C under my belt, and I've worked with python alot during my thesis
15:56 pyrimidine looked at biopython?
15:56 kblin yeah, used it. but bioperl supports more formats and features
15:57 kblin and the existing toolchain in my currentjob is build in perl, so not really my choice :)
15:58 kblin I'll cope
15:59 kblin code-wise, bioperl uses less magic than other code I've seen
16:12 deafferret ....0\
16:42 rbuels o hai
16:43 rbuels kblin: moose.
16:43 rbuels kblin: i have not used python enough to be able to really evaluate it myself, but i have heard it says that Moose is a lot better than python's object system
16:44 rbuels kblin: and i certainly like Moose a lot and use it for everything.
16:44 rbuels kblin: http://www.iinteractive.com/moose/
16:45 rbuels kblin: http://search.cpan.org/~drolsky/​Moose-1.09/lib/Moose/Manual.pod
16:45 rbuels kblin: but unfortunately, bioperl predates moose by a lot, and the current dist structure doesn't allow incorporating code written with moose
16:46 rbuels kblin: which is one of the reasons bioperl is going to split into a bunch of smaller distributions for the next release
16:52 deafferret over my apathetic body!!
16:53 deafferret -ponder-  'otherwise scheduled brain cycles' might be more accurate
17:16 spekki0 joined #bioperl
17:21 spekki0 I have a pretty straightforward subroutine that does some epost and efetch stuff, here is the code http://codepad.org/QYB3YhdW It takes in the arguments of a file to write all the genbank records to, an email address for the eutilities email field, and an array of uid's to retreive records for. It then posts the records in sets of 100, retreives the history object, does an efetch on that history object, then does a get_response to retreive
17:21 spekki0 the records. But for some reason on line 81 i keep getting the error "can't call method "set_parameters" on an undefined value at line 81" Anyone wanna have a quick look and see if im doing womething wrong?
17:25 rbuels spekki0: you are not using $factory correctly.
17:26 rbuels spekki0: you have two 'my $factory' lines.
17:26 rbuels spekki0: the one inside the eval is not going to assign to the one outside
17:26 rbuels spekki0: thus when you get to line 81, $factory is undef, because it was never assigned
17:27 spekki0 ah bugger, can't belive i missed that. thanks man :)
17:28 rbuels spekki0: http://perldoc.perl.org/perlsub​.html#Private-Variables-via-my()
17:28 * rbuels awards himself the "Pendantic Doc Linking" merit badge
17:28 spekki0 lol
17:34 * deafferret awards rbuels the 'Incorrectly Spelling Pedantic' merit badge.
17:34 * deafferret has a whole room full of these damn badges
17:35 rbuels yeah, my typing on irc really sucks
17:35 deafferret y'all on Google Latitude? I'm deafferret@gmail.com
17:43 kblin rbuels: dunno, I guess I have to really use Moose to tell
17:43 kblin I don't do a lot of OOP in python
17:43 pyrimidine kblin: it's very nice.
17:44 pyrimidine and it's based on perl 6 OO
17:44 kblin I find that there's many real-world problems that plain old procedural programming can solve cleaner than OOP
17:45 pyrimidine true
17:46 kblin of course then you don't have factory patterns, and all of that GoF coolness
17:46 dukeleto pyrimidine: any more ideas about topic/leto_interpro_bug ?
17:46 pyrimidine dukeleto: haven't check yet.  will look now
17:47 kblin I'll still have a look at Moose
17:48 dukeleto pyrimidine: i am thinking that maybe, before throwing a "relationship already exists" exception, we see if the relationship is the same as the one that exists. Maybe then we can just throw a warning like "redundant relationship" instead of blowing up
17:48 deafferret pyrimidine: I accepted your Latitude, but it doesn't seem to know where you are
17:49 pyrimidine I am everywhere, and yet nowhere
17:49 deafferret is your phone also omnipresent? that makes Latitude less interesting to stare at
17:51 pyrimidine Sorry, I have an env3 w/o a data plan
17:51 pyrimidine :(
17:51 deafferret oh! ha! why did you invite me if you don't have it?   stalker
17:51 pyrimidine :P
17:52 deafferret :)
17:52 pyrimidine latitude will give a specific location for the account if set, based on the google profile
17:52 pyrimidine (or it is supposed to)
17:53 deafferret uhh... but the other people I follow I watch their location update automatically every couple of minutes
17:53 pyrimidine who's the stalker now?
17:53 pyrimidine :)
17:53 deafferret so they commute home and what not and I see their icon move on the map
17:53 kblin why would anybody want google to know that?
17:53 deafferret it's kinda neato.  not nearly accurate or timely enough for good stalking though
17:54 deafferret it's just dead wrong 20% of the time, and way out of date a lot
17:54 pyrimidine kblin: b/c google must know all
17:54 kblin they're evil.. their verizon deal proves that
17:55 deafferret apathy. I like shiny toys of the greater metro area and where all my friends (probably) are right now
17:55 deafferret errr... maybe might be
17:56 pyrimidine 'do no evil'
17:56 pyrimidine maybe steve jobs was right?
17:57 kblin pyrimidine: it's "do no evil (unless it makes our shareholders richer"
17:57 kblin s/richer/richer)/
17:57 deafferret takes rbuels' Pendatic[sic] award and gives it to kblin
17:58 rbuels deafferret++ # good use of sic
17:58 * deafferret pukes a little
18:06 pyrimidine dukeleto: looks as if the actual relationship name changes
18:07 pyrimidine will commit a small change to demonstrate that
18:07 CIA-95 bioperl-live: Chris Fields topic/leto_interpro_relationship * r21fabdf / Bio/Ontology/SimpleOntologyEngine.pm : small change to exception, demonstrate name change - http://bit.ly/a3Y43W
18:08 pyrimidine so, the question is, if the original relationship is a stub, maybe there needs to be a way to denote that or allow replacements
18:09 paul__ joined #bioperl
18:09 pyrimidine (almost seems like the same could be said for the term stuff from that last fix)
18:09 paul__ hello I have a question about Bio::SeqIO::genbank
18:09 dukeleto pyrimidine: hmmm. i am not sure if the original is a stub, it seems that the interpro xml file has redundant relationships
18:09 pyrimidine it's an odd change
18:10 pyrimidine MSG: relationship CONTAINS between IPR000086 and IPR000059 already defined as FOUND_IN
18:10 paul__ Bio::SeqIO fails when trying to parse NC_011532. It says MSG: Unrecognized molecule type:cRNA
18:10 pyrimidine dukeleto: so, maybe it's defining the inverse relationship?
18:11 paul__ can we add "cRNA" to the VALID_MOLTYPE hash in Bio::SeqIO::genbank.pm?
18:11 pyrimidine paul__: that's a recent addition, will fix
18:11 dukeleto pyrimidine: yeah, seems like it
18:11 paul__ thank you
18:12 pyrimidine paul__: I'll probably remove that check, as it's more a PITA than it's worth
18:12 pyrimidine or at least convert it to a warning
18:13 paul__ just found another unrecognized molecule type "ms-DNA"
18:13 paul__ yeah , sounds like a good idea
18:13 pyrimidine that's a bit frustrating, actually
18:13 paul__ i'm currently parsing through a bunch of genbank records and I'm running into these errors
18:14 pyrimidine I'll revert that, then
18:14 pyrimidine NCBI is supposed to have a defined list of these types
18:14 pyrimidine apparently, that's not up to date
18:14 pyrimidine nor is it very surprising
18:19 CIA-95 bioperl-live: Chris Fields master * rf7d3663 / Bio/SeqIO/genbank.pm : change exception on unknown moltype to debugging statement; at some point we can use this for validation, but until then it's a PITA w/o NCBI having a valid list of valid molecule types available - http://bit.ly/9Jh95t
18:19 pyrimidine paul__: done
18:22 paul__ thank you
18:32 kblin pyrimidine: that's another reason for bioperl, btw, quick turnover for the easy bugs
18:33 pyrimidine and decades for the hard ones
18:34 kblin well, if you still care after a year or two, you don't have an excuse to fix it yourself
18:36 kblin or you could hire somebody to fix it
18:36 kblin or make it a thesis project or whatever
18:37 kblin same as everywhere else
18:38 pyrimidine hmm, not sure I want bioperl code to be someone's thesis
18:39 pyrimidine I've seen what happens there COUGH*phylonetwork*COUGH
18:40 deafferret hmm? Bio::PhyloNetwork was someone's thesis and it didn't go well?
18:41 deafferret & food
18:42 pyrimidine no, it was someone's thesis that got added to core (should have been a separate CPAN release).
18:42 pyrimidine why should we maintain someone's thesis code?
19:14 kblin ok, I can see that
19:14 kblin but bugfixing bioperl sounds like a good software engineering thesis
19:15 kblin real-world application and all
19:20 deafferret if it's useful software then we should be happy it's there?  (though preferably not in core. is it tightly bound?)  if it's not useful, delete it?
19:21 pyrimidine deafferret: we'll likely split it out
19:22 pyrimidine lots of code that could follow that path
19:27 pyrimidine rbuels, dukeleto: looks like Interpro is describing relationships between terms in both directions
19:28 pyrimidine IPR000059 is FOUND_IN IPR000086; IPR000086 CONTAINS IPR000059
19:30 pyrimidine looking at add_relationship() in SimpleOntologyEngine, it looks as if it is adding terms both ways automatically
19:36 pyrimidine No, take that back
19:37 pyrimidine it has two relationship stores (normal and inverted), and is storing the same term in each, but inverting it in the inverted store
19:37 * pyrimidine seriously confused and scared
19:41 rbuels pyrimidine: Bio::Ontology has both normal and inverted stores?
19:41 pyrimidine yes
19:41 rbuels pyrimidine: that seems reasonable, for lookups ...  assuming lookups are important
19:41 pyrimidine right
19:41 * rbuels has not used any of this in anger, that is dukeleto
19:42 pyrimidine so, I'm wondering if the problem is when the call is being made from the handler to the engine, if the IDs need to be inverted under some conditions
19:49 pyrimidine dukeleto: fixed it.
19:53 CIA-95 bioperl-live: Chris Fields topic/leto_interpro_relationship * raaba18e / Bio/OntologyIO/Handlers/InterProHandler.pm : correct erroneous assumption that terms must be flipped in relationship, but this needs further tests - http://bit.ly/aRs9rD
19:54 pyrimidine so, maybe additional tests there to make sure the relationships for those terms (IPR000059, IPR000086) are properly defined
19:59 pyrimidine left #bioperl
19:59 pyrimidine joined #bioperl
20:35 lauanana joined #bioperl
20:46 pyrimidine rbuels, dukeleto: digging a bit more, I think the interpro stuff is going to take some work
20:46 pyrimidine unfortunately
20:46 * rbuels wonders where dukeleto is
20:46 * dukeleto is here
20:47 dukeleto pyrimidine: what is up?
20:47 * dukeleto was ingesting foodstuffs
20:47 * rbuels points dukeleto at the backscroll
20:47 pyrimidine so, the problem is that the relations don't indicate which is parent, and which is child
20:47 pyrimidine even though the actual terms (FOUND_IN, CONTAINS) make that obvious
20:48 pyrimidine so, with the commit I made, you get the correct connection, but the problem is $obj1 becomes parent of $obj2
20:48 pyrimidine and then (later on) vice versa
20:48 pyrimidine bam, deep recursion
20:49 pyrimidine when using something like get_ancestor_terms
20:49 pyrimidine so, one could just punt and not add the second term in (and warn)
20:50 pyrimidine or overwrite the relationship
20:50 pyrimidine (now I say it out loud, the fix isn't so bad :)
20:51 pyrimidine so, maybe just skip if the relationship is already defined?
20:51 dukeleto pyrimidine: i think that is reasonable
20:52 dukeleto pyrimidine: with maybe a warning. but i am not sure if the warning is for the people who generate the InterPro XML or those that use it
20:53 dukeleto pyrimidine: i guess they store the relationship in both directions so that each chunk of XML knows all of the relations. but does that make any sense?
20:53 pyrimidine dukeleto: not sure, but it's odd
20:53 pyrimidine an ontology is a DAG, so defining the relationship both ways is redundant
20:56 dukeleto pyrimidine: it seems that the InterPro XML file is what has the bug, but I am not sure if they would consider it so
20:57 pyrimidine it's possible there are frameworks that have relationship types that (when specifying parent relationships) use the inverse of the normal type name
20:58 dukeleto pyrimidine: should it be an error to redefine a relationship between two terms? If we allowed the redefinition of relationships, this bug would go away, right?
20:59 pyrimidine yes
20:59 * rbuels wonders if somebody in the interpro group should be brought in on this
20:59 pyrimidine rbuels: probably
21:01 pyrimidine I think the problem lies in a deficiency in Bio::Ontology::Relationship
21:01 pyrimidine the relationship is defined in only one direction
21:02 pyrimidine but doesn't allow for defining the inverse
21:02 pyrimidine within that same relationship instance
21:02 Jun Hi, pyrimidine. Do you know howo to catch the output from ./Build test
21:02 Jun Now I can pass 179/199 of the old tests of SimpleAlign
21:03 pyrimidine Jun: redirect STDERR and STDOUT
21:03 pyrimidine nice!
21:04 pyrimidine back on ontology
21:04 pyrimidine so, if there is a relationship between two terms (FOO is FOUND_IN
21:04 pyrimidine BAR)
21:05 pyrimidine then the inverse (BAR CONTAINS FOO) should be true
21:06 pyrimidine so, seems to me there needs to be a way of defining the inverse of the relationship
21:06 dukeleto Jun: ./Build test &> test.log & tail -f test.log
21:07 dukeleto pyrimidine: seems reasonable
21:09 pyrimidine could be as simple as having the handler check if a specific relationship already exists and (if so) makes sure it is the inverse type
21:10 dukeleto pyrimidine: yes, i am thinking that the handler should know that A is FOUND_IN B is equivalent to B CONTAINS A, and not warn in that case
21:10 pyrimidine yes
21:11 Jun yep, just found it. Cheers
21:11 pyrimidine if not, allow it to go ahead and die (b.c the relationship is already defined in one way or the other)
21:13 dukeleto pyrimidine: so we just need to keep a small list of which relationships are equivalent when the terms are inverted. If a relationship redefinition fits into one of those, ignore the relationship, because it is redundant. Otherwise, the current behavior of throwing an exception
21:13 pyrimidine yep
21:14 dukeleto pyrimidine: where should this check actually go?
21:15 pyrimidine wondering if it should go in start_element in InterProHandler
21:20 dukeleto pyrimidine: seems like it needs to be in add_relationship on the ontology object
21:23 pyrimidine dukeleto: only worry I have with a change there is how it would affect relationship types defined in other ontologies
21:24 dukeleto pyrimidine: is it possible to have an interpro-specific add_relationship ?
21:25 pyrimidine I think it's using a generic ontology class; only the terms are interpro-specific
21:26 pyrimidine in InterProHandler, in start_element (line 531) there is a check for 'contains' and 'found_in'
21:30 dukeleto pyrimidine: so maybe the check should go into _create_relationship in the interpro handler
21:31 melic_ joined #bioperl
21:33 pyrimidine yes, that works, I think
21:34 dukeleto pyrimidine: are you still thinking that the code you commented out in my branch should stay that way, or are you going to revert that?
21:36 CIA-95 bioperl-live: Chris Fields topic/leto_interpro_relationship * r4308669 / Bio/OntologyIO/Handlers/InterProHandler.pm :
21:36 CIA-95 bioperl-live: Revert "correct erroneous assumption that terms must be flipped in relationship, but this needs further tests"
21:36 CIA-95 bioperl-live: This reverts commit aaba18edcecc99eacb72fa923e6eecf3fc21b211. - http://bit.ly/aDcIQ9
21:39 melic_ joined #bioperl
21:39 pyrimidine joined #bioperl
21:39 spekki0 joined #bioperl
21:39 Yksi joined #bioperl
21:39 kai joined #bioperl
21:39 Lynx_ joined #bioperl
21:39 deejoe joined #bioperl
21:39 driley joined #bioperl
21:39 genehack joined #bioperl
21:39 flu_ joined #bioperl
21:39 deafferret joined #bioperl
21:39 ChanServ joined #bioperl
21:41 bag_ joined #bioperl
21:42 pyrimidine netsplit
21:50 pyrimidine rbuels: see jrockway on #catalyst talking about the new book?
21:51 pyrimidine (03:45:46 PM) jrockway: it's my book, but with occasional "see the last paragraph? that's old so don't do that." and then it just moves on
21:51 pyrimidine (03:45:56 PM) jrockway: no offense, but that is not how you "update" a book
21:51 pyrimidine heh
21:52 rbuels pyrimidine: hahaaaha
21:52 rbuels from the horse's mouth.
21:52 pyrimidine yep
21:52 * rbuels is glad to see confirmation of his initial evaluation
21:52 pyrimidine (03:49:24 PM) jrockway: i know there's no money in books, but they should at least pretend to try
21:52 pyrimidine (03:49:36 PM) jrockway: that is why i didn't want to update my book
21:52 pyrimidine (03:49:38 PM) jrockway: it's *all* outdated
21:52 pyrimidine (03:49:43 PM) jrockway: the examples are useless
21:52 pyrimidine (03:49:46 PM) jrockway: kill with fire, start over
21:53 pyrimidine so, there you have it
21:53 pyrimidine s'okay, have the Definitive Guide in my hands now
21:58 pyrimidine will be back on tomorrow.  I may add a simple method to SimpleOntologyEngine to do a quick lookup for relationships between one or two terms.  Seems sorely lacking there.
21:59 pyrimidine commuting
21:59 pyrimidine left #bioperl
21:59 rbuels O\
22:42 deafferret o7
23:15 driley_ joined #bioperl
23:19 driley_ joined #bioperl
23:19 driley_ joined #bioperl
23:19 bag_ joined #bioperl
23:20 driley_ joined #bioperl
23:20 driley__ joined #bioperl
23:22 driley joined #bioperl
23:24 driley__ joined #bioperl

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