Camelia, the Perl 6 bug

IRC log for #bioclipse, 2007-12-03

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

All times shown according to UTC.

Time Nick Message
07:25 rojas1 joined #bioclipse
08:15 olas joined #bioclipse
08:15 olas hi all
08:18 steinbeck joined #bioclipse
08:19 olas moin steinbeck
08:19 steinbeck moin
08:25 thomas_ku joined #bioclipse
08:57 jonalv joined #bioclipse
09:02 EskilA joined #bioclipse
09:21 masak joined #bioclipse
09:42 masak joined #bioclipse
09:49 jonalv egonw: I saw in the log that you where wondering about browsing molecules in database from Bioclipse
09:50 olas jonalv: searching and browsing of databases is really a central issue in Bioclipse2
09:50 olas we should really try to design good support for that
09:50 jonalv I guess that's pretty related to my structuredb project which is somewhat put on ice right now. because of bioclipse 2 development It will have a few requirements on tables though...
09:50 masak olas: please elaborate what 'good support' means
10:39 egonw a beaty: http://depth-first.com/articles/2007​/12/01/casual-saturdays-perspective
10:44 egonw also interesting: http://plindenbaum.blogspot.com/2007/12/pivot.html
10:44 egonw wrt QSAR
13:24 olas left #bioclipse
13:25 olas joined #bioclipse
13:25 edrin joined #bioclipse
13:25 edrin hi
13:27 edrin how are you ther?
13:36 olas fine thanks
13:36 olas egonw: around?
13:36 egonw yes
13:36 olas I have problems with the ReaderFactory
13:37 olas new ReaderFactory().createReader(bs) return null, even though bs is a working InputStream
13:37 egonw mmm...
13:37 egonw I would say that should work...
13:37 olas (I can read the stream and it look ok)
13:37 egonw seems that the format is not detected...
13:37 olas PDB?
13:37 egonw please verify that is detected the format...
13:38 egonw with getFormat(IS) or so...
13:38 egonw if that returns null too...
13:38 olas also returns null
13:38 olas it's a PDB header
13:38 olas HEADER    TRANSCRIPTION REGULATION                06-MAR-92   1D66      1D66   2 COMPND    GAL4 (RESIDUES 1 - 65) COMPLEX WITH 19MER DNA                 1D66   3 SOURCE    (SACCHAROMYCES $CEREVISIAE)
13:38 egonw OK, might be a problem that it did not nicely load the list of file formats...
13:38 olas works in bioclipse1
13:38 olas ok
13:38 olas how can I detect that?
13:38 egonw looking
13:39 olas where?
13:39 egonw getFormats()
13:39 olas don't have cdk source
13:39 olas ok
13:39 egonw olas: bookmark this: http://cheminfo.informatics.indiana​.edu/~rguha/code/java/nightly/api/
13:39 egonw it's the javadoc for HEAD of trunk/
13:40 olas ok
13:40 olas you are right
13:40 olas no formats loaded
13:40 olas what's the problem then?
13:40 olas I did: ReaderFactory fac=new ReaderFactory(); System.out.println("formats: " + fac.getFormats().size());
13:40 olas formats: 0
13:41 egonw OK
13:41 egonw has to do with the classloading...
13:41 egonw ah, no...
13:41 olas I get no eror messages
13:41 egonw we're using the source now...
13:41 olas are we?
13:41 egonw and the list of formats is autocreated...
13:41 egonw yes, via svn:externals
13:41 egonw OK, use this approach:
13:42 egonw factory.registerFormat(PDBFormat.getInstance())
13:43 egonw and the other formats we support for net.bioclipse.cdk
13:43 * olas is trying this
13:44 olas hmm that doesnt work
13:45 olas you can't use tha tsyntax
13:45 olas you need an IChemFormatMatcher as input
13:45 egonw yeah, cast the bastard
13:45 olas ok
13:46 olas ok, works now
13:46 olas thanks a lot
13:48 olas egonw: how should we expand a CDK molecule?
13:48 olas into atomContainers?
13:49 olas or chemModels?
13:49 olas Or both?
13:49 egonw http://cheminfo.informatics.indiana.edu/​~rguha/code/java/nightly/api/org/opensci​ence/cdk/graph/ConnectivityChecker.html
13:49 * olas will do IAtomCOntainers as reference
13:50 olas egonw: should I use that?
13:50 egonw if I understood correctly what you want to acchieve...
13:50 olas I have a chemfile
13:51 olas and want to expand it (like clicking on the arrow on a Java-class in Eclipse)
13:51 olas should reveal all molecules in the file
13:51 olas (and later on all spectra etc)
13:52 egonw in the outline?
13:52 olas no, in the navigator
13:52 egonw ok, about the same :)
13:52 olas (outline will be similar but only if editor is open)
13:52 olas yes
13:53 egonw doesn't that work sort of the same way as the ChemObject tree we have in BC1
13:53 egonw ?
13:53 egonw with getChildren() ?
13:53 olas hmm, kindof
13:53 olas but not too cluttery
13:53 olas I think expand to sub-molecules should be fine in navigator
13:53 olas outline can be mor eelaborate and expand to atoms/bonds etc
13:54 olas anyway, will do a reference and you can update it later if you like :-)
13:54 olas (I'm sure you will!)
13:54 olas egonw: do you have an MDL file with 3-10 molecules you can email me?
13:54 olas and also a CML file?
13:56 olas and another question:
13:56 olas How do I get the name of a molecule?
13:56 olas I mean, it if a file contains many molecules, could they be named in the file?
13:56 olas or should I just call them "molecule1", "molecule2) etc?
13:57 egonw olas: it's in net.bioclipse.cdk/
13:57 egonw I think...
13:57 egonw but I think you want ChemModelManipulator.getAllAtomContainers()
13:57 egonw not the one I gave earlier
13:58 egonw IMolecule.getProperty(CDKConstants.TITLE)
13:58 olas thx
14:10 olas egonw: how about em files with mols?
14:12 egonw oh, yeah... mom...
14:13 egonw https://cdk.svn.sf.net/svnroo​t/cdk/trunk/cdk/src/data/mdl
14:14 egonw pick one of the .sdf fils
14:14 egonw files
14:18 olas egonw: GeometryTools.has3DCoordinates says my PDB file does not have 3D coordinates
14:18 egonw ha :)
14:19 egonw umm... the things expects *all* atoms to have 3D...
14:19 egonw some might not?
14:19 egonw email me the PDB file, please
14:21 olas ok
14:22 olas egonw: sorry, my fault
14:22 olas it does work
14:23 egonw good :)
14:32 masak joined #bioclipse
14:43 olas egonw: around?
14:43 olas Varför får jag ClassCastException för fac.registerFormat((IChemFormatMatcher) XYZFormat.getInstance());
14:44 olas o, will not use xyz for the moment
14:45 olas egonw: around?
14:46 olas IF I read a PDB molecule with 5 frames=models in CDK, how are they read into CDK? as 5 ChemModels?
14:48 DrGTO joined #bioclipse
14:49 egonw olas: correct
14:49 olas ok
14:50 olas why is there no echo of SVN commits here?
14:50 egonw CIA client is present...
14:51 olas maybe a SF problem?
14:51 egonw no, don't think so...
14:51 egonw they *are* showing up on #cdl
14:51 egonw #cdk
14:52 olas egonw: can't get the name from the IMolecuke...
14:52 egonw might not be set
14:52 egonw depends on the file
14:55 olas an sdf file?
14:55 pohyg joined #bioclipse
14:55 egonw olas: should have
14:55 olas hmm
14:55 egonw but likely on ChemModel level?
14:55 olas will double-check
14:55 pohyg hi!
14:56 olas hi pohyg
14:56 pohyg hi!
14:56 olas who are you?
14:57 pohyg biologist?
14:57 olas have we met before?
14:57 olas here in IRC I mean?
14:57 pohyg a few days i was here asking a question on chemical searches
14:57 olas oh
14:57 olas then it's egonw you should talk to
14:58 olas and maybe even on the #cdk channel
14:58 pohyg I would like to know if bioclipse or other tools which we can search compound by formula/structure.. is it possible?
14:58 olas oh
14:58 pohyg yeah you are the second person to point me to egonw
14:58 olas it's an ongoing project
14:58 * egonw feels popular...
14:59 olas I'll let egonw answer
14:59 egonw also already named in the Royal Society of Chemistry circular Chemistry World ...
14:59 * jonalv is listening
14:59 * egonw has a good day
14:59 * masak has nothing to add
14:59 pohyg i hv problem compiling bioclipse
14:59 egonw if only he could get that Eclipse maven plugin installed...
14:59 pohyg but the binary works fine
15:01 egonw are you using eclipse?
15:01 jonalv wait a minute what happened to the search tool discussion?
15:01 pohyg no
15:02 pohyg alright will give that a try.... so it is tightly coupled with eclipse?
15:03 pohyg yeah search tool for chemicals... I am not sure where to go proper
15:03 pohyg compound names are relatively easy... but sometimes... there are several names to 1 compound.
15:11 egonw join #m2eclipse
15:12 DrGTO hello everyone
15:12 egonw hi DrGTO
15:12 DrGTO reading the last lines ... are there different channels related to bioclipse?
15:13 egonw no
15:13 DrGTO k
15:13 DrGTO so if I have some serious firstTimeUser problem with bioclipse I come here ? ;)
15:13 egonw yes
15:13 egonw sorry, no bioclipse-users yet
15:14 * DrGTO was trying the online help on bioclipse website ...
15:15 DrGTO well, I just finished reading the paper about CMLSpect and DL Bioclipse 1.1.3 for windows to try out the case study for NMRShiftDb
15:16 DrGTO looks like there are some plugins missing, still after doing the update process
15:16 DrGTO no jdx or cml file from the spectra samples can be opened
15:17 egonw good point...
15:17 egonw olas: the spectrum feature should be part of the main update site, not?
15:17 egonw DrGTO: did you install that feature?
15:17 DrGTO well, actually I didnt see it in the update features list
15:18 DrGTO Help->Add extensions ...
15:18 egonw DrGTO: please try 1.1.5 pre:
15:18 egonw http://bioclipse.blogspot.com/2007​/10/bioclipse-115-pre-release.html
15:18 egonw that should have that version
15:18 DrGTO ah :)
15:19 DrGTO DL ....
15:20 olas egonw: I donät think spectra are in 115
15:20 egonw oh, crap
15:20 DrGTO nope
15:20 DrGTO its not listed
15:20 olas egonw: I don't know how to use the spectra functionality
15:20 * egonw makes a note to set up that spectrum update website...
15:20 olas you have to fix the feature
15:20 egonw olas: no problem
15:20 egonw will short cut this with Christoph
15:21 olas or just fix it and I'll upload it to bioclipse update site
15:21 DrGTO hmm, is there a chance then to test the plugin features described in the paper?
15:21 olas egonw: we should really release 117
15:21 olas with spectrum support
15:21 egonw what happened to 1.1.6 ?
15:22 egonw yes, we should
15:22 olas :-)
15:22 egonw whatever the version :)
15:22 olas sure
15:22 olas if you have time to fix it this week?
15:22 olas (and test it)
15:22 olas I could imagine a 116 later this week then
15:23 olas egonw: what do you think?
15:25 egonw fix what?
15:45 olas egonw: fix the spectrum feature
15:45 egonw ah, sure
15:45 egonw what's wrong with it?
15:45 olas donät know
15:45 olas it might even work
15:45 egonw then it's not broken
15:45 olas but please test and verify that
15:45 egonw I am not aware of anthing wrong with it
15:45 egonw ack
15:45 egonw will do
15:45 olas then I will release it
15:46 olas later this week
15:46 olas I'd really like an "getting started with spectrum in Bioclipse"-article
15:47 olas step by step how to use it with examples
15:47 olas then I could test it too
15:47 olas but as it is now: no idea how it works
15:52 DrGTO hmm
15:52 DrGTO thats a point
15:52 DrGTO maybe I dont find it, but its hard to getting started with bioclipse. Not much manuals available :(
15:53 egonw ummm... there are cheat sheets I think
15:53 egonw for many things...
15:53 DrGTO I found plugin help pages on the wiki
15:56 DrGTO but it looks more like spending a weekend trying this and that :)
15:56 olas DrGTO: there is a getting started guide, have you seen it?
15:56 DrGTO yes
15:56 DrGTO I already read it
15:56 olas and your comments are?
15:56 DrGTO now I am hungry for more
15:56 olas ah :-)
15:57 olas I just asked egonw for an article about spectrum
15:57 DrGTO I heard a lot about some really nice workflow features
15:57 olas I don't know if it exists
15:57 olas oh, workflows are in VERY much development
15:57 DrGTO like the spectra stuff
15:57 olas hmm
15:57 DrGTO I think it was Stefan? who told me about the NMR workflow
15:57 olas oh, that must be stefan
15:57 egonw bioclipse/trunk/playground/plugin​s/net.bioclipse.tutorial.spectra
15:58 olas but egonw knows about that too
15:58 DrGTO ah, okay. So I have to go into the source code ..
15:58 egonw no
15:58 egonw will make sure that tutorial is included in the feature
15:58 olas egonw: is there no material about spectrum available?
15:58 olas ah, great
15:59 DrGTO thats what I found
15:59 DrGTO http://www.bioclipse.net/doc/1.0.1/index.html
15:59 olas hmm, are there no docs for the 113?
16:00 olas must fix that for next release
16:00 * olas has SO much to do right now
16:00 DrGTO sorry to bother you with stupid documentation ;)
16:00 egonw olas: make a mental note to periodically back up the website
16:00 egonw olas: and then be happy about removing old docs
16:01 egonw old docs are worse than no docs
16:01 egonw DrGTO: not at all
16:01 egonw DrGTO: it's important
16:01 DrGTO sure it is. But its always last on the list when you ask a Dev
16:02 DrGTO cause programming new features is much more fun
16:08 olas very true
16:08 olas and more importantly: You can't publish maintenance
16:08 olas and bug fixes
16:08 olas it doesn't bring your research forward
16:09 olas and I need to publish in order to get my PhD
16:09 olas but on the other hand
16:09 olas docu IS important in the longer run
16:09 olas so need some sort of tradeoff
16:09 olas I try to do both :-)
16:11 DrGTO nvm. I will try bioclipse tonight at home. More time, maybe a nice glas of wine abd things will sort out
16:11 egonw olas: two fixes commited to trunk/plugins
16:12 egonw one, actually
16:13 * DrGTO goes home
16:13 DrGTO thanks for the help so far. I am afraid I will come back ;)
16:13 egonw no problem
16:13 egonw who worked on renaming the IScriptProvider again?
16:14 egonw might he/she please fix net.bioclipse.biojava.ui ?
16:16 masak egonw: I think I might have... caused things there
16:16 masak I'll look into it
16:24 egonw olas: I'm happy with the feature
16:24 egonw will move things to appropriate locations in SVN
16:24 olas good
16:24 olas I will make a release later this week then
16:24 masak egonw: there, I hope that solved it
16:29 edrin joined #bioclipse
16:29 masak too bad name refactorings only concern the projects that one has checked out
16:34 egonw masak: that's why developers must have *all* plugins checked out in trunk/plugins and trunk/features
16:35 masak hm
16:35 masak guess so
16:35 masak at least in the presence of extension points
16:36 masak heh, GOTOs are evil because they relate completely different parts of code with each other, creating action-at-a-distance
16:36 masak instead we have exceptions and extension points :)
17:04 jonalv joined #bioclipse
18:02 edrin left #bioclipse
18:17 olas joined #bioclipse
19:10 olas left #bioclipse
20:16 egonw_ joined #bioclipse
23:33 kaskelot_ joined #bioclipse

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