Camelia, the Perl 6 bug

IRC log for #bioclipse, 2010-03-10

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

All times shown according to UTC.

Time Nick Message
03:42 sneumann joined #bioclipse
07:18 annzi joined #bioclipse
07:33 Gpox joined #bioclipse
08:03 egonw joined #bioclipse
08:53 jonalv joined #bioclipse
09:17 olass joined #bioclipse
09:19 olass moin
09:19 zarah hello olass, you fantastic person you
09:19 * olass blushes
09:34 jkutzer joined #bioclipse
09:48 olass @pz 1915
09:48 zarah http://pele.farmbio.uu.se/cgi-bi​n/bugzilla/show_bug.cgi?id=1915
09:59 olass jonalv: ping
10:00 olass jag inspekterar vad jag trodde var en ICDKMolecule och ser att den är en 'CDKMolecule$$EnhancerByCGLIB$$1118f06f "molecule"= CDKMolecule$$EnhancerByCGLIB$$1118f06f  (id=183)'
10:00 olass vad betyder det?
10:00 olass iaf så falerar den testet: molecule.getClass().isAssign​ableFrom(CDKMolecule.class) i CDKManager.saveSDF(...)
10:01 olass woops, sorry, english here
10:01 olass jonalv: can you explain this?
10:02 jonalv olass: All BioObjects in Bioclipse are wrapped in proxies for recording purposes
10:02 jonalv olass: what you are seeing is the proxy class
10:02 olass how do I get the ICDKMolecule?
10:02 jonalv olass: you should program against interaces
10:02 Gpox CDKMolecule.class.isAssagnab​leFrom(molecule.getClass())
10:03 olass Gpox, jonalv: please see line 1953 in CDKManager
10:03 jonalv molecule.getClass().isAssign​ableFrom(CDKMolecule.class) should probably be molecule.getClass().isAssign​ableFrom(ICDKMolecule.class)
10:03 olass ok, I can try...
10:04 olass not that I wrote that code...
10:06 * egonw is back from fika
10:06 egonw and reading up right now
10:07 jonalv olass: I think that line should read: if(molecule instaceof ICDKMolecule) {
10:08 egonw jonalv:
10:08 Gpox olass: and then ac = ((ICDKMolecule)moleucle).getAtomContainer()
10:08 egonw molecule.isAssignableFrom(ICDKMolecule.class) I think
10:08 egonw Gpox: not?
10:08 Gpox why not instanceof?
10:08 egonw you convinced me of assignable instead of instanceof
10:08 egonw don't remember the arguments though
10:08 egonw it's all the same?
10:09 jonalv egonw: why are you not using instaceof?
10:09 masak joined #bioclipse
10:09 jonalv oh haj masak
10:09 egonw jonalv: will try to remember the argument...
10:09 olass o/
10:09 masak oh haj jonalv.
10:10 olass ok, chaning to instanceof solved that poroblem
10:12 masak "Every problem in computer science can be solved by another layer of instanceof. Except the problem of too many instanceofs."
10:13 jonalv masak: is there sucha thing as too many instanceof?
10:13 jonalv :)
10:17 masak jonalv: search your heart. you know it to be true.
10:18 * jonalv searches for his heart
10:20 jonalv masak: okej you probably have a point
10:21 egonw jonalv: replied to your dbmolecule/inchi comment...
10:21 egonw jonalv: I think it might work if we put the ICDKMolecule in a separate plugin
10:21 egonw olass: any objections against one more plugin?
10:21 jonalv egonw: hangon while I read your answer
10:22 jonalv egonw: well a blocker is a strong word... ...but on the other hand it certainly cripples StructureDB..
10:23 egonw criplles sounds like a blocker to me
10:24 jonalv egonw: but ICDKMolecule is in the domain plugin too
10:24 egonw too? like what is there too?
10:24 jonalv egonw: why would you need to depend on cdk.business in order to use the ICDKMoleucle instead of the CDKMoleucle?
10:24 egonw what's the domain plugin?
10:25 egonw jonalv: ah, sorry...
10:25 egonw .cdk.domain depends on .inchi
10:25 jonalv egonw: I am confused...
10:25 egonw so, cannot do it the other way around
10:25 egonw but if we move ICDKMolecule out of cdk.domain into cdk.model
10:26 egonw or CDKMolecule into cdk.business perhaps
10:26 jonalv egonw: hang on, Gpox is showing me the code...
10:26 olass maybe a good idea to separate ICDKMolecule from CDKMolecule
10:27 egonw jonalv: there is no cdk.domain plugin
10:27 jonalv egonw: it turns out the error message on line 79 of the InchiManager is less than awesome
10:27 Gpox just make DBMolecule adatable to IAtomContainer
10:28 jonalv egonw: the real problem is that it can't get the Atomcontainer from by adapter. Not that the molecule is not a CDKMolecule
10:28 olass Gpox++ //Adapters is the future
10:42 egonw jonalv: you mean you cannot use the adapter trick?
10:43 jonalv egonw: I can, but the exception didn't hint at the problem being with adapters
10:47 jonalv olass: ping
10:47 * olass is talking to egon in skype
10:47 olass ponging later
10:47 jonalv olass: I am looking at row 358 in CDKMoleuclePropertySource. What is your defence?
10:48 jonalv hm maybe it's not really your fault.
10:49 jonalv GAH, the properties sytem we have now is _so_ borken
10:52 annzi joined #bioclipse
11:02 olass jonalv: pong
11:02 olass you called?
11:03 jonalv olass: yea Gpox helped me to sort things out. I am going to rewrite that code of yours now to fix the probl;wm...
11:03 jonalv Gpox++
11:03 olass jonalv++
11:06 CIA-57 bioclipse.cheminformatics: Ola Spjuth master * r7db14bc / (2 files):
11:06 CIA-57 bioclipse.cheminformatics: Adapted saveSDFile to support proxied parameters.
11:06 CIA-57 bioclipse.cheminformatics: Also, generalized parameters to <? extends IMolecule> - http://bit.ly/99r6bD
11:07 * egonw is very pleased with the ChEMBL/SPARQL applications
11:07 egonw it triggers quite a few wishes, bug reports, etc...
11:07 egonw seems a very good use case for Bioclipse
11:10 jonalv egonw: what applications are you talking about?
11:10 Gpox olass: I assume the testCreateSDFile_String_IMoleculeArray passes now :)
11:10 egonw those for the paper I am composing on Bioclipse-RDF wrt to QSAR/PCM/cheminformatics
11:10 jonalv egonw: ah right :)
11:12 olass Gpox++
11:14 annzi egonw: speaking of chembl/sparql, I will start making more questions for the chembl plug-in now, ok with you?
11:16 CIA-57 bioclipse.qsar: Ola Spjuth master * r6c33da7 / (2 files):
11:16 CIA-57 bioclipse.qsar: * Initialize structure- and responselist if empty.
11:16 CIA-57 bioclipse.qsar: * Also removed projectFromChEMBL from manager.
11:16 CIA-57 bioclipse.qsar: * Cleaned up for 80 cols - http://bit.ly/aQL9JS
11:16 CIA-57 bioclipse.qsar: Ola Spjuth master * r2b54d71 / (2 files): Refactored out reading/writing of QsarModel to/from file. - http://bit.ly/bkMYDn
11:16 CIA-57 bioclipse.qsar: Ola Spjuth master * rf4aecfc / features/net.bioclipse.qsar_feature/feature.xml : Removed statistics plugin from qsar feature. - http://bit.ly/d9HUR5
11:16 CIA-57 bioclipse.qsar: Ola Spjuth master * r5ac7efa / plugins/net.bioclipse.qsar.ui/META-INF/MANIFEST.MF : Added chembl plugin as dep. - http://bit.ly/bWC2IK
11:16 CIA-57 bioclipse.qsar: Ola Spjuth master * rf2202bc / (4 files in 3 dirs): Implemented a wizard for QSARProject From ChEMBL. - http://bit.ly/dikxNc
11:17 olass ok, time 4 lunch
11:23 egonw annzi: yes
11:36 sneumann joined #bioclipse
11:49 egonw github is down
11:54 olass egonw: seems ok to me...
11:54 egonw then they are back up already
11:54 egonw that would be quick
11:55 egonw they spoke of a major down :)
11:55 egonw confirmed
11:55 olass well, the web gui seems to work
11:55 * egonw is impressed
11:55 olass github++
11:55 egonw "Site is down hard. We've tracked it to the database and are researching further."
11:55 egonw is what they tweeted 10mins ago
11:55 egonw ok, 11
11:55 egonw github++
11:59 egonw quick way to get your branch back in order with upstream master:
11:59 egonw git reset olas/master
11:59 egonw s/olas/foo/ where appropriate
11:59 egonw that's pretty neat...
11:59 olass yes
12:03 CIA-57 bioclipse.qsar: Ola Spjuth master * rf6c818b / plugins/net.bioclipse.qsar.ui/src/net/bioclipse/qs​ar/ui/wizards/NewQSARProjectFromChEMBLWizard.java : Added a progressmonitor to wizard - http://bit.ly/brcUAK
12:29 olass egonw: I am on the Bertrand question
12:30 egonw ack
12:30 olass will produce a screencast for him
12:30 egonw good
12:30 egonw I normally add something like "I'm sure Ola will reply too, but ..."
12:31 olass no, thanks for answering
12:31 egonw since he was folllowing the wiki tutorial, he should be doing most right
12:31 olass just wanted to update you
12:31 egonw ack
12:31 * egonw is tired
12:31 egonw and needs some easy stuff to do
12:31 egonw olass: please consider sending me some SMILES with failing atom types
12:31 olass did Martin send you some examples?
12:32 egonw yeah, and fixed the platinum ones
12:32 olass nice
12:32 egonw umm... he sent, I fixed :)
12:32 olass will do when I run into them
12:38 * jonalv is back from lunch
13:35 annzi egonw: where can I file issues with the rdf plugin?
13:36 egonw @pelezilla
13:36 zarah egonw: Pelezilla is at http://bugs.bioclipse.net
13:38 annzi thnx, signing up immediately
13:39 CIA-57 bioclipse.cheminformatics: jonalv master * r301687c / plugins/net.bioclipse.cdk.business/src/net/bioc​lipse/cdk/domain/CDKMoleculePropertySource.java :
13:39 CIA-57 bioclipse.cheminformatics: solved "inchi property is String isntead of InChi"
13:39 CIA-57 bioclipse.cheminformatics: Solved situation where there is a property named inchi which is a String instead of an Inchi object - http://bit.ly/bodJTg
13:39 CIA-57 bioclipse.cheminformatics: jonalv master * r4af7a3a / plugins/net.bioclipse.cdk.business/src/net/bioc​lipse/cdk/domain/CDKMoleculePropertySource.java : removed trailing white spaces - http://bit.ly/bDKKp0
13:40 CIA-57 bioclipse.structuredb: jonalv master * r54c0c23 / plugins/net.bioclipse.structuredb/src/net/b​ioclipse/structuredb/domain/DBMolecule.java : adapt dbmolecule to atomcontainer - http://bit.ly/bs5DMy
14:14 olass will be offline to make screencast
14:14 olass connecting soon again
14:15 olass until then...
14:15 jonalv olass: screencast++
15:11 olass joined #bioclipse
15:34 samuell joined #bioclipse
16:08 olass joined #bioclipse
16:10 CIA-57 bioclipse.structuredb: jonalv master * r3c1d9db / plugins/net.bioclipse.structuredb/src/sqlMap.xml :
16:10 CIA-57 bioclipse.structuredb: attempt at query optimization
16:10 CIA-57 bioclipse.structuredb: This shoulod theoretically be better (I think). I haven't made any measurements and the feeling is that it wasn't significantly better but the code is at least shorter so let's keep it. :) - http://bit.ly/cgQ6vG
16:40 sneumann joined #bioclipse
19:14 egonw joined #bioclipse
21:03 samuell joined #bioclipse
21:58 sneumann joined #bioclipse

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