Camelia, the Perl 6 bug

IRC log for #bioclipse, 2008-05-06

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

All times shown according to UTC.

Time Nick Message
05:59 rojasm joined #bioclipse
06:56 thomas_ku joined #bioclipse
07:10 rojasm joined #bioclipse
08:00 egonw joined #bioclipse
08:04 olas joined #bioclipse
08:08 jonalv joined #bioclipse
08:34 jonalv is it just me or is net.bioclipse.cdk not compiling today?
08:35 masak joined #bioclipse
08:35 egonw jonalv: did we break it again? :)
08:36 jonalv egonw, guess so... :)
08:47 CIA-45 bioclipse: jonalv * r5554 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): mangertest for creating and retrieving structures now passes
09:11 masak joined #bioclipse
09:13 egonw jonalv: please file a bug report
09:13 egonw will look at it asap
09:16 jonalv I don't have any problems except it doesn't compile fully :)
09:18 olas does not compile here either
09:19 olas for me it's the import org.kxml2.io.KXmlParser that's not included
09:19 egonw ah, that's easy
09:19 olas what's kxml2?
09:19 egonw in the jar/ folder there is a kxml2-2.3.0.jar
09:19 egonw add that to the build path, .classpath, etc
09:20 egonw (not export)
09:20 olas egonw: could you briefly explain what's best: XOM kXML, JAXB?
09:20 egonw JAXB sucks
09:20 masak I'm of a mind to move scripting back into net.bioclipse.core
09:20 egonw XOM and kXML are both good
09:20 masak anyone against?
09:20 olas no
09:20 egonw but different technology
09:20 egonw XOM is a DOM
09:20 egonw kXML is a pull parser
09:20 olas what's the diference?
09:20 olas hmm
09:20 * masak likes pull parsers
09:20 egonw scalability
09:20 olas and that means?
09:20 jonalv masak: I am not against but does it work?
09:20 masak jonalv: dunno
09:20 masak will try tho'
09:20 egonw a pullParser does not have to read the whole XML doc in one go
09:21 egonw like SAX
09:21 * masak really likes SAX
09:21 egonw but SAX is not a pull parser
09:21 egonw but an event driven parser
09:21 masak ah
09:21 olas what is good with XOM?
09:21 egonw it's a well design DOM library
09:21 egonw Jumbo uses it
09:22 egonw better than the w3c DOM specs
09:22 olas when use DOM and not kXML?
09:22 egonw small documents
09:22 egonw kXML requires you to set up a custom data structure
09:22 egonw for CDK that is no problem, that's the CDK data classes
09:22 olas but these are the two parsers you recommend?
09:22 egonw Jumbo uses XML-DOM as internal data model
09:23 egonw no, SAX is fine too :)
09:23 egonw SAX2 that is
09:23 olas when to use that?
09:23 olas large docs?
09:23 egonw yes, like kXML
09:23 olas When use SAX2 over kXML?
09:23 egonw SAX and kXML is on par
09:25 egonw pull parsers are fairly new
09:25 egonw are more suitable for iterators...
09:25 masak huh? pull parsers new?
09:25 egonw SAX is fine when you read the DOC in one go...
09:26 egonw masak: well, for XML, yes
09:26 masak they're not newer than 7-8 years
09:26 egonw ~3 years or so
09:26 egonw for XML?
09:26 olas kXML is from 1999
09:26 masak egonw: ah, sry, was thinking of event parsers again
09:26 * masak shuffles off learning about pull parsers
09:27 CIA-45 bioclipse: jonalv * r5555 /bioclipse2/trunk/plugins/net.bioclipse.jmol/src/n​et/bioclipse/jmol/adapter/cdk/CdkJmolAdapter.java: Switched from catching arrayindexoutofboundsexception into catching indexoutofboundsexception. Still not sure it makes sence...
09:27 egonw mmm... the also say this impl is better:
09:27 egonw http://www.extreme.indiana​.edu/xgws/xsoap/xpp/mxp1/
09:28 egonw so will likely switch at some point :)
09:32 olas egonw: where is the jar?
09:32 olas for kxml in cdk?
09:32 olas do I need to DL it?
09:32 egonw a svn update should do it
09:32 olas sry
09:32 egonw found it?
09:34 CIA-45 bioclipse: ospjuth * r5556 /bioclipse2/trunk/plugins/net.bioclipse.cdk/ (.classpath META-INF/MANIFEST.MF build.properties): Added kxml.jar to classpath.
09:34 olas yup
09:34 egonw all compile problem fixed now?
09:34 olas yup
09:34 CIA-45 bioclipse: jonalv * r5557 /bioclipse2/trunk/plugins/ (4 files in 3 dirs): Creating and retrieiving folders now work
09:42 olas egonw: around?
09:42 egonw yes
09:42 EskilA joined #bioclipse
09:42 olas for the outline for JCP I am constructing, what is the model object?
09:42 olas a ChemFile or IAtomCOntainer?
09:43 olas and what is the difference in scope?
09:43 olas it's just for JCP, hence only one mol, right?
09:44 olas I will do Atoms and Bonds first
09:44 egonw IChemFIle
09:44 masak olas: request permission to change a line of your code
09:44 egonw because that allow reactions too
09:45 olas hmm, ok
09:45 olas masak: what line?
09:45 egonw well, olas, this is exactly why I want one format, one editor
09:45 masak olas: line 56
09:45 masak of AbstractServiceContainer.java
09:45 egonw so the MDL molfile editor could indeed just use IAtomContainer...
09:45 masak in n.b.services
09:45 olas which says now?
09:45 egonw actually, JCP is using IAtomContainerSet, I think
09:45 masak olas: if (children!=null) return children; else return null;
09:46 masak it's very silly
09:46 olas hahaha
09:46 olas must have been tired
09:46 masak yes...
09:46 olas check what time I wrote that?
09:46 masak will check
09:46 olas egonw: ok
09:50 masak olas: it was 16:10 in the afternoon on a Monday
09:51 jonalv hate those mondays...
09:51 masak 'case of the Mondays?'
09:51 masak :D
09:54 CIA-45 bioclipse: jonalv * r5558 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): manager test for creating and retrieivng users now passes
09:55 masak olas: I changed it to a slightly more sensible if (children == null) return Collections.EMPTY_LIST; return children;
09:55 masak also, I did defensive copying on both getter and setter
09:55 masak just some small changes, that is
09:55 olas great
09:55 egonw lunch time
09:55 * olas @ lunch
09:59 CIA-45 bioclipse: carl_masak * r5559 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/n​et/bioclipse/ui/views/ScriptingConsoleView.java:
09:59 CIA-45 bioclipse: [ScriptingConsoleView.java]
09:59 CIA-45 bioclipse: * added a few checks for nulls
09:59 * masak >>> lunch
10:36 steinbeck joined #bioclipse
11:06 olas egonw: around?
11:06 egonw yes
11:06 olas from MDLEditor I can get
11:06 olas editor.getJcpModel().getChemModel();
11:06 egonw good
11:06 egonw that makes sense
11:06 olas duger ICHemModel som modellobject?
11:06 olas eller skall jag casta den till ngt?
11:07 olas (pratar om outline View nu)
11:07 olas hur får jag atoms och bonds från en IChemObject?
11:07 olas borde det inte vara en AC?
11:08 egonw a IChemModel contains ICrystal, IMoleculeSet and/or IReactionSet
11:08 egonw the IMoleculeSet contains one or more IMolecule
11:08 egonw each one is a AC
11:09 egonw I'd start they outline with ChemModel
11:09 egonw as JCPmodel may also contain IReaction's in a IReactionSet
11:14 olas could the MDLEditor support all that?
11:15 olas reactions?
11:15 egonw no, MDL is really for just molecules
11:15 egonw IMoleculeSet
11:19 olas ok
11:19 olas I'll start with that and we can extend it later for reactions
11:22 olas egonw: could a Molecule contain more than one AC?
11:22 egonw ok
11:22 egonw no
11:22 egonw Molecule extends AC
11:22 olas what is the difference between ICHemModel.getMolecule() and getAtomContainer?
11:23 olas hmm
11:23 olas ModelSet.getMol
11:23 olas and ModelSet.getAC
11:23 egonw semantics
11:23 olas in what sense?
11:23 egonw one return IAC, other IMolecule
11:24 olas whateve
11:24 EskilA joined #bioclipse
11:34 CIA-45 bioclipse: jonalv * r5560 /bioclipse2/trunk/plugins/net.bioclipse.st​ructuredb.tests/tests/net/bioclipse/struct​uredb/persistency/dao/FolderDaoTest.java: Test for putting structure in folder. It fails. :(
11:51 malick joined #bioclipse
11:54 thomas_ku joined #bioclipse
11:54 CIA-45 bioclipse: carl_masak * r5561 /bioclipse2/trunk/plugins/net.bioclipse.​services/src/net/bioclipse/services/view​s/model/AbstractServiceContainer.java:
11:54 CIA-45 bioclipse: [AbstractServiceContainer.java]
11:54 CIA-45 bioclipse: * prevented getChildren method from returning null
11:54 CIA-45 bioclipse: * added defensive copying
11:56 CIA-45 bioclipse: carl_masak * r5562 /bioclipse2/trunk/plugins/net.bioclipse.​services/src/net/bioclipse/services/view​s/model/AbstractServiceContainer.java:
11:56 CIA-45 bioclipse: [AbstractServiceContainer.java]
11:56 CIA-45 bioclipse: * removed useless check for membership
11:57 masak olas: there. I'm done tampering with your code now.
12:03 EskilA joined #bioclipse
12:25 olas *phew*
12:45 egonw does Uppsala have interest in reading PubChem XML.gz archives?
12:45 egonw CDK has now a reader for that :)
12:46 jonalv egonw, sounds great! (I have no idea of when I would want to use it but it sounds great!) :)
12:46 egonw check the XML dir of the PubChem FTP download side
12:46 jonalv notn ow
12:46 egonw you can download the full PubChem as binary ASN.1, XML and SDF
12:47 jonalv oki
12:47 egonw the SDF sucks, as it is too limited to represent the chemitry in PubChem
12:47 egonw (e.g. complexes)
12:47 egonw and I hate binary stuff
12:56 olas egonw: sounds excellent
12:56 olas can it be read on the fly from Pubchem, or does it need downloading of complete file?
12:57 olas Pubchem has Web services, but I don't think the new are available yet
12:57 olas guess I'll download the xml
12:57 olas you have a URL to that?
12:58 egonw where are the PC webservices documented?
13:04 olas non so far
13:04 olas I was contacted for evaluation a while ago
13:04 olas but not heard of it since
13:05 olas we'll see what turns out
13:05 olas Services infrastructure in BC is far from finished anyway
13:06 egonw I discovered this weekend how easy it is to write SOAP clients using the Java libs in javax.xml.soap
13:06 egonw (or so)
13:06 egonw SOAP1.1 only, but works quite well
13:11 olas aha
13:18 thomas_ku joined #bioclipse
13:22 CIA-45 bioclipse: jonalv * r5563 /bioclipse2/trunk/plugins/ (3 files in 3 dirs): Structuredb: All tests passes. Yup it's now possible to import a SDF file into the structuredb. I should make it possible to testrun it from the console real soon now...
13:27 olas cool
13:27 olas looking forward to integrating that with the structureEditor
13:28 jonalv olas : yup
13:33 masak olas: DnD is fun!
13:34 * masak is having a great time here
13:43 olas I know
13:44 masak ok, I have a working PoC now
13:46 olas PoC?
13:46 masak 'Proof of Concept'
13:46 masak which means, it does something
13:46 masak but maybe not what everyone wants
13:46 jonalv commit? :)
13:47 CIA-45 bioclipse: ospjuth * r5564 /bioclipse2/trunk/plugins/net​.bioclipse.cdk10.jchempaint/ (16 files in 5 dirs): Created a ContentOutline for MDLEditor for atoms and bonds. Selecting in Outline highlights bonds/atoms in JCP (similar to Bioclipse1).
13:47 masak jonalv: coming
13:47 olas will be a blog from me today/tomorrow about JCP with outline/highligting in BC2 with screenshot.
13:48 jonalv olas: w00t! :)
13:48 olas gotta run now
13:48 olas cu tomorrow
13:50 masak cu
13:52 CIA-45 bioclipse: carl_masak * r5565 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/n​et/bioclipse/ui/views/ScriptingConsoleView.java:
13:52 CIA-45 bioclipse: [ScriptingConsoleView.java]
13:52 CIA-45 bioclipse: * dragged and dropped files, directories and projects
13:52 CIA-45 bioclipse:  from the BioResource Navigator now shows up as
13:52 CIA-45 bioclipse:  strings in the ScriptingConsoleView
13:53 masak jonalv: try now
16:49 ilbot2 joined #bioclipse
16:49 Topic for #bioclipseis now An open source workbench for chemo- and bioinformatics - http://www.bioclipse.net/ | BiclipseWiki => http://wiki.bioclipse.net/ !! | logs => http://colabti.org/irclogge​r/irclogger_logs/bioclipse & http://moritz.faui2k3.org/irc​log/out.pl?channel=bioclipse
16:49 ilbot2 joined #bioclipse
16:49 Topic for #bioclipseis now An open source workbench for chemo- and bioinformatics - http://www.bioclipse.net/ | BiclipseWiki => http://wiki.bioclipse.net/ !! | logs => http://colabti.org/irclogge​r/irclogger_logs/bioclipse & http://moritz.faui2k3.org/irc​log/out.pl?channel=bioclipse
18:19 rojasm joined #bioclipse

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