Camelia, the Perl 6 bug

IRC log for #bioclipse, 2008-09-12

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

All times shown according to UTC.

Time Nick Message
01:52 m4r1in joined #bioclipse
05:22 CIA-31 joined #bioclipse
06:47 malick joined #bioclipse
06:52 rojasm joined #bioclipse
07:25 Gpox joined #bioclipse
07:54 jonalv joined #bioclipse
07:57 jonalv moin
07:57 zarah hi jonalv
08:02 masak joined #bioclipse
08:08 olas joined #bioclipse
08:08 olas hi
08:08 zarah hi olas
08:08 jonalv hi olas
08:14 shk3 joined #bioclipse
09:24 jonalv hi shk3
09:24 jonalv I jave aquestion for you
09:24 jonalv You gave me a big sdf file
09:24 jonalv from nmrshiftdb
09:25 jonalv Do you know if CML is supposed to be creatable for all those molecules?
09:25 jonalv I am having some issues while trying to create CML for them...
09:30 shk3 cml should be possible for all
09:30 shk3 can't see why it shoudn't
09:30 jonalv org.xmlcml.cml.base.CMLRuntimeException: bond already in array b30/a20__a18/m1
09:32 shk3 sounds like a bug, I would say
09:32 shk3 can you file a cdk bug?
09:32 shk3 and attach the mol file
09:32 shk3 you are using cdk, aren't you?
09:33 jonalv yea I should take a look at it. It's one of the entries in the sdf file...
09:33 shk3 if you extract the mdl, I will look at it
09:33 shk3 just fixing a problem in cml plugin
09:33 jonalv I will try to figure out which one after I have eaten. It's just to extract the entry from the sdf and savde it as .mol right?
09:34 shk3 yes
09:34 jonalv k
09:34 shk3 everyting between the $$$$
09:41 maclean joined #bioclipse
09:44 shk3 question:
09:44 shk3 in the cml plugin, I need to pass a URL to a method in jumbo
09:44 shk3 the url needs to point to a file inside the plugin
09:44 shk3 is this possible?
09:45 shk3 the plugin may be a jar, so is a url possible?
09:45 shk3 if I use FileLocator.find, the url looks like bundle://2591/file, which confused jumbo
09:46 steinbeck joined #bioclipse
10:00 olas shk3: you could convert to local fileurl I suppose
10:00 olas then it will start with file://
10:00 olas bundle:// is an OSGI thingy so jumbo will not understand it
10:01 maclean We're trying 'Platform.resolve()'
10:03 olas ok
10:04 shk3 seems to work
10:04 shk3 it's FileLocator.resolve now
10:04 shk3 at least it works when running from eclipse
10:04 shk3 assume it will work in exported product as well
10:04 olas ack
10:30 olas shk3: try to export the product and test
10:30 olas should be straightforward
10:38 shk3 export?
10:38 shk3 straightforward?
10:38 shk3 you are joking
10:45 olas have you tried it in bc2?
10:45 olas select the features and click "export"
10:46 olas if the features are stable, it's straightforward
10:46 olas if your plugins have faults in their build .properties or include excluded classes from classpath, it fails.
10:57 shk3 hm, my spectrum stuff is not yet in a feature, I think
10:57 shk3 we might need a spectrum feature
10:57 shk3 but I will try to get everything working first
10:58 thomas_ku joined #bioclipse
10:59 olas ok
11:15 jonalv shk3: you solved your problems?
11:20 shk3 yes
11:23 jonalv good
11:57 jonalv shk3: I can not create cml for this mol file: http://pastebin.com/f1dd5bcbb
12:01 shk3 jonalv: will look at it right now
12:01 jonalv shk3: great
12:01 shk3 well, going to the loo first
12:02 maclean joined #bioclipse
12:03 jonalv hi maclean
12:04 maclean hi.
12:07 maclean I'm not sure what my answer would be to the 100 atom smiles question.
12:08 maclean I think it is better to prevent the user from doing the wrong thing rather than allowing them to and telling them they did wrong.
12:08 maclean LBYL rather than EAFTP :)
12:11 jonalv maclean: sounds like a good general principle, how dows it apply to this special case? :)
12:12 maclean Disable the "generate smiles" option if the molecule is larger than 100?
12:12 maclean (if that's possible)
12:12 jonalv I don't see where this 100 atom limit comes from
12:12 jonalv to me it sounds like soemthing someone just made up
12:13 maclean Probably, but Chemistry has to stop somewhere so that Biology can begin.
12:14 masak the biology department agrees fully
12:14 maclean http://www.rcsb.org/pdb/static.do?p​=general_information/news_publicati​ons/news/news_2008.html#20080902b
12:16 maclean More seriously, I wouldn't want bioclipse to generate the smiles for, e.g. the ribosome.
12:16 jonalv But what about just trating atom containers that are biopolymers as biology?
12:16 shk3 i can't see why smiles should stop somewhere
12:16 shk3 100 isn't that much
12:16 shk3 (at least not including Hs)
12:19 maclean but you can have 'bio'polymers in Chemistry. Small peptides, for example, are interesting to both disciplines.
12:21 jonalv okey I don't know what to do when I am importing a SDF file into the databse and a molecule contains more than 100 atoms and the cdk manager refuses to calcualte smiles for me...
12:21 olas smiels should never include Hs IMO
12:22 maclean hmmm. I see now the problem, jonalv.
12:23 maclean I thought this was generating smiles from a menu-option.
12:23 jonalv maclean: nothing ever is that easy in Bioclispe these days. Everything is an API that should be runnable from JavaScript
12:24 maclean this is true. There's probably an xml file involved somewhere, too.
12:24 olas can we add a flag force=true?
12:25 shk3 jonalv: the problem is in the mdl file - it contains a bond twice
12:25 shk3 this is because the bond is twice in the nmrshiftdb db
12:25 shk3 meaning in mysql
12:25 shk3 it probably is in there, since it was painted twice in jcp
12:25 jonalv shk3: so the file is broken?
12:25 shk3 well, sort of
12:25 jonalv (I have the same problem with another file too)
12:26 shk3 I consider the file broken
12:26 shk3 i can change it in the database
12:26 jonalv okey
12:27 shk3 the question is a bit how the bond got in twice
12:27 shk3 I try this in jcp
12:27 jonalv I would love a new one since I use that file for breaking my database
12:27 jonalv oh did I say breaking?
12:27 jonalv I off course meant testing
12:27 shk3 yes
12:27 jonalv :)
12:27 shk3 best would be to change the sdf directly
12:27 shk3 fastest for you
12:28 shk3 go to that entry,
12:28 shk3 make the fourth line
12:28 shk3 46 48  0  0  0  0  0  0  0  0999 V2000
12:28 shk3 and delete the line  18 20  1  0  0  0  0
12:30 jonalv that's it?
12:30 olas is it a corrupt file?
12:30 olas or is it correct?
12:30 olas jonalv: if that part of the sdfile is corrupt, skip it. Also report the indices of molecules that were skipped after the import.
12:30 jonalv olas: Yea I need to do something like that yea...
12:30 shk3 well, not sure if it is exactly illegal
12:31 jonalv I am testing with this now
12:31 shk3 but in any case it's ambiguous
12:31 shk3 since there are two bonds from 18 to 20 with different stereo wedges
12:36 jonalv shk3: What does net.cb.bibtex do? :)
12:37 shk3 includes jabref
12:37 maclean references?
12:37 shk3 and enables parsing of bibtex files
12:37 shk3 will be used in specmol
12:37 shk3 but I not yet ported that bit from bc1
12:37 jonalv okey
12:37 jonalv MAy I wonder what it is used for?
12:38 jonalv What's in bibtex files?
12:38 shk3 literature references
12:38 maclean references
12:43 masak references to literature.
12:44 shk3 jonalv: if you have other entries with that problem, please report.
12:44 shk3 thanks
12:44 jonalv shk3: will do
12:44 jonalv shk3: I am trying to slurp that file :)
12:44 jonalv it takes a while...
12:48 shk3 in an editor?
12:55 jonalv no, into structredb
12:56 jonalv but I am making fingerprints, smiles and CML for all of them...
12:56 masak no CIA bot today :(
12:56 masak maybe we should start letting zarah do the svn log reporting.
12:56 masak at least she's here more often.
12:57 jonalv masak: how do you do that?
12:57 jonalv polling svn log?
12:57 masak jonalv: good question. periodic pulling, I guess.
12:57 masak aye.
12:58 jonalv that's... thats...
12:58 masak doing svn up regularly
12:58 jonalv oh well...
12:58 masak and if it updates new stuff, report the latest svn logs
12:59 masak would be even better if there were an RSS/Atom feed somewhere, though
13:00 maclean what about this friendfeed thingy?
13:00 maclean http://beta.friendfeed.com/ro​oms/chemistry-development-kit
13:00 jonalv is ist tracking commits?
13:00 maclean for cdk, yes, I think.
13:01 masak maclean: that link took me to an info page, not to the cdk room
13:01 olas cool
13:01 maclean I don't pretend to understand what it actually IS.
13:02 maclean I would categorise it as a "web-based thing".
13:02 jonalv maclean: I don't see any commit tracking on that page
13:03 maclean Well, I think that you have to make an account to see it. And, on closer inspection it might be listening to a cia bot?
13:04 maclean I really don't know anything about any of this stuff - please to ignore :)
13:26 olas joined #bioclipse
13:27 olas hi again
13:42 shk3 hi again
13:50 steinbeck joined #bioclipse
14:00 steinbeck joined #bioclipse
14:16 jonalv hi again
14:16 zarah hi again
14:17 jonalv :)
14:17 olas hi again
14:17 olas :?
14:17 olas jonalv-specific command?
14:18 stain stop this hi game!
14:18 olas :-)
14:18 olas hi stain
14:18 masak hi stain
14:18 jonalv hi stain
14:18 zarah hi stain
14:18 olas what are you up to a friday afternoon like this?
14:18 olas afternoon tea?
14:19 olas afternoon beer?
14:19 masak afternoon soup?
14:19 olas or afternoon-taverna-hacking?
14:19 stain I'm hacking Taverna and Maven and Raven and Axis and WSS4j and classpaths
14:19 masak .oO( afternoon space walk )
14:19 stain very fun :)
14:19 masak Raven?
14:20 olas Like Maven but with an R
14:20 masak olas: no shit, sherlock
14:20 jonalv What's R doing in Maven?
14:20 olas jonalv: taverna
14:21 masak R is doing Taverna in Maven?
14:21 * masak confused
14:21 jonalv olas: is taverna a verb?
14:21 olas uses it for runtime dependency management
14:21 olas (AFAIK)
14:21 olas where is our commit bot?
14:21 olas early weekend?
14:22 jonalv CIA is probably out on some mission...
14:22 stain jonalv: it's Tavernas plugin system built on Maven
14:22 masak stain: ah.
14:22 jonalv intresting
14:22 stain no, stay away ..
14:22 stain hihi
14:22 jonalv so it's buit for Taverna?
14:22 jonalv Okey I will
14:22 stain it's built for anything
14:23 jonalv I will stick to OSGI
14:23 stain here's some info.. http://www.mygrid.org.uk/dev/blog/2008/04/raven/
14:23 stain yeah, I wish we could as well
14:23 stain which I bet would just give lots of other problems, but then we wouldn't be alone!
14:23 olas lol
14:24 * olas and Raven are not good friends
14:24 olas last time we met Raven bit me
14:24 stain olas: he does that to everyone
14:24 stain didn't you wear gloves?
14:24 olas haha
14:24 olas no
14:24 olas almost lost a finger
14:25 olas "get back into the cage!"
14:25 * jonalv doesn't like maven and will stay away from raven
14:25 stain well but Maven is kind.. just a bit mysterious
14:26 masak if ( /[mr]aven/i ) { print "stay away!" }
14:26 jonalv but I actually (propbably fro the first time) managed to build soemthing using maven the other day
14:27 masak jonalv++
14:28 * olas likes EMF
14:28 olas EMF++
14:28 olas karma
14:28 zarah olas has a karma of 21
14:28 jonalv maven--
14:29 olas raven--
14:29 jonalv karma
14:29 zarah jonalv has a karma of 33
14:29 olas :-(
14:29 jonalv here it comes...
14:29 * olas used to have >76
14:29 jonalv yea there it came...
14:29 olas what were the reasons again to clear my karma?
14:29 masak olas: usually, people are very quiet about having cheated
14:29 masak you're the exception.
14:30 * olas doesn't understand
14:30 masak of course you don't.
14:30 stain karma
14:30 zarah stain has a karma of 0
14:30 stain :((
14:30 olas I... did not cheat that bot... zarah...
14:30 jonalv stain++
14:30 masak stain++ # welcome
14:31 stain thank you.. do I need to say good things about CDK now?
14:31 olas no
14:31 olas Bioclipse is enough
14:31 masak :)
14:31 jonalv just don't say bad things about Bioclipse :)
14:31 olas cdk is OK I guess
14:32 masak patches are ok too
14:32 olas but sometimes makes our lives miserable, like yesterday
14:32 olas we should work against a CDK jar
14:32 jonalv no we should work for a CDK jar
14:33 olas haha
14:33 olas smartass
14:33 olas much friday now...
14:35 jonalv very much fridahy
14:38 shk3 karma
14:38 zarah shk3 has a karma of 4
14:48 jonalv stain: where are you?
14:56 maclean joined #bioclipse
15:14 CIA-30 joined #bioclipse
15:14 maclean CIA! You're back!
15:20 olas nice :-)
15:22 CIA-30 bioclipse: ospjuth * r6963 /bioclipse2/trunk/plugins/net.bioc​lipse.qsar.ui/src/net/bioclipse/qs​ar/ui/editors/ResponsesPage.java: Response page now creates responses for molecules and displays them in a table.
15:24 CIA-30 bioclipse: ospjuth * r6964 /bioclipse2/trunk/plugins/net.bioclip​se.qsar.ui/src/net/bioclipse/qsar/ui/​editors/DescriptorLabelProvider.java: Implements ILabelProvider to enable sorting in viewer using the non-overridden ViewerSorter. Also display text for Responses.
15:24 olas anyone who knows how to define NaN as default to a float attribute in XML Schema?
15:26 olas guess I can do it programmatically
15:29 masak olas: Only the data types BINARY_DOUBLE and BINARY_FLOAT support the NaN value.
15:36 olas thx
15:37 CIA-30 bioclipse: ospjuth * r6965 /bioclipse2/trunk/plugins/net.bioc​lipse.qsar.ui/src/net/bioclipse/qs​ar/ui/editors/ResponsesPage.java: Set default value to NaN for responses.
15:41 olas have a nice weekend all
15:41 * olas is off to buy heaps of cheese
15:42 olas bye
15:42 olas left #bioclipse
16:19 malick left #bioclipse
19:44 rojasm joined #bioclipse
21:02 steinbeck joined #bioclipse

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