Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-02-09

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

All times shown according to UTC.

Time Nick Message
05:10 masak joined #bioclipse
06:48 rojasm joined #bioclipse
08:04 Gpox joined #bioclipse
08:08 olass joined #bioclipse
08:08 olass moin
08:08 zarah saluton olass
08:13 egonw joined #bioclipse
08:16 olass hi egonw
08:16 egonw hi olass
08:16 zarah egonw: You have new messages. Write '/msg zarah @messages' to read them.
08:16 egonw at UU, or at home?
08:16 egonw @msg
08:16 zarah egonw: 10 h 24 m 17 s ago, edrin said could you make the RDF service a delayed service? responses sometimes need some seconds...
08:16 zarah egonw: 7 m 36 s ago, masak said re pz ticket for js exceptions not shown in console: yes, #236. it currently doesn't have a milestone.
08:17 * egonw is at UU
08:17 * masak is at UU
08:17 * olass is at home
08:17 * Gpox is at home
08:17 masak zarah?
08:17 * jonalv is studying
08:18 olass jonalv++ //for being a good student
08:18 egonw jonalv has a show stopper in his todo list
08:18 egonw release will not be possible unless that is fixed
08:18 olass which one?
08:18 egonw mom, checking
08:18 olass I'm sure he'll fix it today if it's important
08:19 masak egonw: could you please remove the Component named 'Bioclipse 2.0.0' from Pelezilla?
08:20 masak 2.0.0 is supposedly a Milestone, not a Component.
08:21 egonw masak: please file a bug
08:21 masak ok.
08:25 masak olass: kudos on the new welcome screen.
08:25 olass danke
08:25 masak but it says the same thing twice: "Install sample data" -- "Install sample data"
08:25 olass been quite some work
08:26 masak "Install extended sample data" -- "Install extended sample data"
08:26 masak we can't have that.
08:26 olass ok, please rephrase
08:26 olass (but that is what it is)
08:26 olass 1) small piece of data (no extra download)
08:26 masak what I mean is, we should write something else than the same thing twice.
08:26 olass 2) More data but requires downloading
08:26 masak there's no need to say it twice.
08:27 olass right
08:27 masak it just gives a confused impression.
08:27 olass much in intro is just stubs
08:27 masak better to say nothing the second time.
08:27 olass have not put much effort
08:27 olass agreed
08:27 egonw olass: @bug #422
08:27 olass @pz 422
08:27 zarah http://tinyurl.com/bh2sj7
08:28 olass yup, I see it
08:29 rojasm joined #bioclipse
08:29 egonw hi rojasm!
08:29 rojasm hi egonw
08:29 rojasm good morning
08:29 masak morning!
08:34 olass morning rojasm
08:37 * egonw has to drop of something, bbl
08:38 CIA-55 bioclipse: goglepox * r8667 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.jchempaint/src/net/bioclipse/cdk/jch​empaint/editor/SWTMouseEventRelay.java: SWTMouseRelay now only relays left button clicks (resolves bug #413)
08:38 zarah bug #413 | http://tinyurl.com/ats4a3
08:44 olass @tell edrin: please fix bug 372 before tomorrow to get webservices in RC1
08:44 zarah Consider it noted.
08:56 olass Gpox: what is the selection in the SDFEditor?
08:56 jonalv joined #bioclipse
08:56 olass are they ICDKMolecules?
08:58 Gpox yes
08:58 olass gut
09:05 maclean joined #bioclipse
09:05 maclean hello
09:05 zarah privet maclean
09:05 maclean privet
09:14 olass egonw: when is CDK update with new tag in SVN?
09:14 olass ETA?
09:14 olass today?
09:14 shk3 joined #bioclipse
09:16 olass hi shk3
09:17 egonw olass: yes, going to try to get that out today
09:17 olass great
09:18 shk3 hi olass
09:23 CIA-55 bioclipse: ospjuth * r8668 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: CDKManager loadMolecule() throws NPE for 1D66.pdb in testfiles. Wrote a Junit test case for this.
09:25 * egonw does not think that's a bright idea...
09:25 egonw olass: better use Jmol's PDB reader
09:25 olass egonw: it cannot read into an ICDKMolecule
09:25 egonw I'm not suprised
09:26 olass can CDK not read PDBs?
09:26 egonw this is why PDB is not supported in Bioclipse regarding the ICDK side
09:26 egonw well...
09:26 olass ah
09:26 maclean olass: kinda
09:26 olass did not now that
09:26 egonw right :)
09:26 egonw there actually is a patch...
09:26 egonw but the protein support in CDK needs fixing
09:26 egonw post 1.2
09:27 maclean sigh : my bioclipse wont start again
09:27 maclean :(
09:27 olass use the product
09:27 olass works every time
09:27 olass (hmm, SHOULD work)
09:27 olass much more stable than run config
09:28 maclean hmm
09:28 maclean well, it works when I switch to 'based on..plugins'
09:28 maclean so that's fine for me
09:30 egonw maclean: you did try the 'set target platform' right?
09:30 maclean yup
09:31 olass egonw: regarding pdb in bioclipse: is there an IJmolMolecule?
09:32 olass egonw: can jmols API be used in that way?
09:32 rojasm joined #bioclipse
09:32 olass and how much effort would it be to post Jmols PDB reader to cdk?
09:32 olass post -> port
09:32 egonw AFAIK there is no IJmolMOlecule
09:32 olass should there be one?
09:32 egonw a lot
09:33 egonw actually... what is the JmolEditor doing?
09:33 egonw that should use a IJmolMOlecule, I gues...
09:33 olass openClientFile(...)
09:33 olass with file as input
09:33 olass AFAIK
09:33 maclean also...
09:33 maclean jmol.load
09:34 * olass thinks it's a severe flaw of Bioclipse to not support PDBs
09:34 egonw well, current focus is cheminformatics
09:34 egonw not bioinformatics
09:34 olass true
09:34 egonw right?
09:34 olass PDB sits in between IMO
09:34 egonw yes, true too
09:34 maclean olass: also, jmolEditor.load(file)
09:34 egonw we'll get this back online for 2.1
09:34 olass maclean: yes
09:34 olass egonw: sounds good
09:35 olass file bug 429 for 2.1 then
09:37 maclean a hack would be to expose a jmolEditor.load(String) method and convert a CDKMolecule into a string in the JmolManager...
09:38 maclean (not very nice though, I suppose)
09:40 edrin joined #bioclipse
09:40 edrin hi
09:41 zarah privet edrin
09:41 zarah edrin: You have new messages. Write '/msg zarah @messages' to read them.
09:41 edrin @msg
09:41 zarah edrin: 56 m 45 s ago, olass said please fix bug 372 before tomorrow to get webservices in RC1
09:41 edrin @pz
09:41 zarah edrin: Pelezilla is at http://bugs.bioclipse.net
09:42 olass edrin: IMO, just putting everything in Virtual is not a good thing
09:42 olass edin: masak will merge in your jseditor in RC1 today
09:42 edrin ok, where to put it?
09:42 olass see unit test
09:42 edrin olass:
09:43 olass yes?
09:43 edrin olass: ?
09:43 olass see bug
09:43 edrin olass: i mean where to put downloaded pdb files?
09:43 olass where user tells you
09:43 olass updated to take a path as input param
09:43 olass we could add one
09:43 edrin ok
09:44 olass if no path given, store in /Virtual with name= pdbid+."pdb"
09:44 edrin that's not a problem... i am just wondering, if there is a default location
09:44 olass also, we donät pass around IFiles but Strings to Patha
09:44 edrin olass: ok
09:45 edrin Plugin Junit tests in net.bioclipse.webservices.tests fail <- what does this mean?
09:45 edrin masak: are u around?
09:45 olass if you run them, they fail
09:45 olass they should pass
09:45 masak edrin: hiya.
09:45 edrin olass: and what does fail mean? what are they doing?
09:45 olass however, it seems PDB parsing in CDK is broken and will not be fixed until 2.1
09:46 olass edrin: have you looked at the tests?
09:46 edrin hi masak
09:46 edrin olass: not yet
09:46 olass please do that first
09:46 edrin ok
09:46 olass it's very selfexplanatory
09:46 edrin how to run such a "test"?
09:46 olass read bug
09:46 olass there is a link to a wiki page
09:46 edrin ok
09:47 edrin masak: regarding the former .rhino now .jseditor plugin
09:47 edrin i renamed the plugin stuff to this name
09:47 edrin masak: i found some trouble with scripting
09:48 edrin it's a general issue
09:48 CIA-55 bioclipse: carl_masak * r8669 /bioclipse2/trunk/plugins/net.bioclipse​.scripting.ui/src/net/bioclipse/scripti​ng/ui/views/ScriptingConsoleView.java: [scripting] fixed tab completion shortcoming (#397)
09:48 masak edrin: re rename: great.
09:48 edrin the problem is really complicated
09:48 jonalv could someone explain cdk.save to me?
09:48 masak edrin: re trouble with scripting. what issue?
09:48 masak edrin: ok. I don't like the sound of this... :/
09:48 egonw jonalv: I wish I could...
09:48 edrin ok: i will explain
09:48 jonalv egonw: huh?
09:49 egonw jonalv: ok, ok, depends on your actual question... what's up?
09:50 edrin for example, if a plugin brings in new functionality, right now, the plan is to always do the actions/stuff through the manager
09:50 edrin this is working quite ok and stable as the managers have to be thread safe.
09:50 edrin the problem now is
09:51 edrin a plugin might also want to bring in "advanced" stuff
09:51 edrin for example it wants to offer the possibility to use interfaces of a lib
09:51 edrin javascript/rhino supports this
09:52 edrin and in fact this can be useful for advanced scripts
09:52 masak sure.
09:52 masak CDK is another example, no?
09:53 edrin for example xws supports functions like f = xmpp.getFunction(..); f.invokeAsync(input, listener) while listener is an object created from an interface
09:53 edrin cdk yes
09:53 edrin very same
09:54 edrin the problem now is, the context of mozilla rhino must have the lib (CDK, xws) on its classpath
09:54 masak ok.
09:54 masak that feels wrong, yes.
09:55 masak dependency failure.
09:55 edrin thus one can get interfaces like: ISimpleProcessListener = Packages.net.bioclipse.xws.client.​listeners.ISimpleProcessListener;
09:55 edrin yes
09:55 CIA-55 bioclipse: goglepox * r8670 /bioclipse2/trunk/plugins/net.bioclipse.cd​k.jchempaint/src/net/bioclipse/cdk/jchempa​int/widgets/JChemPaintEditorWidget.java: Removed dirty when only selection changed
09:55 edrin ISimpleProcessListener = Packages.net.bioclipse.xws.client.​listeners.ISimpleProcessListener; fails (not in this line but ISimpleProcessListener is an invalid object)
09:56 edrin now the workaround would be to export all packages of the lib in the privding plugin
09:56 edrin and import them in the plugin spawning the rhino context
09:56 masak I don't quite like this.
09:57 masak it means jumping through hoops for rhino's sake.
09:57 edrin however this will result in the requirement that all plugins that offer such things will be a dependecy!
09:57 masak yes, that's plain wrong.
09:57 edrin masak: jumping?
09:57 masak edrin: it's an expression.
09:57 edrin masak: sure, that's a working workaround right now
09:57 masak I've heard that before...
09:57 edrin but not good in general
09:58 masak why did you say you needed to import things into the rhino-using module again?
09:58 edrin in bioclipse i solved this problem with Dynamic-Import * thing in manifest, but it seems Spring "made" this option invalid...
09:58 edrin at least it does not work anymore
09:59 edrin masak: to use classes/objects/interfaces of libs i want to script with
09:59 masak right.
09:59 masak edrin: you should use managers for that.
09:59 edrin masak: for most things this works
09:59 edrin but not for interfaces
10:00 edrin at least i found not way yet
10:00 masak edrin: could you provide an use case for this, please?
10:01 edrin for example: listener = new ISimpleProcessListener() { function: ... onResult: ... }; f.invokeAsync(input, listener);
10:01 edrin for this you need ISimpleProcessListener = Packages.net.bioclipse.xws.client.​listeners.ISimpleProcessListener;
10:01 masak unh.
10:02 edrin anyway, it is a must
10:02 edrin otherwise we lose very much of the power of rhino
10:02 masak edrin: you have just necessitated an intra-office meeting.
10:02 edrin ?
10:02 masak @tell jonal we need to talk.
10:02 zarah Consider it noted.
10:02 masak @tell jonalv we need to talk.
10:02 zarah Consider it noted.
10:03 edrin masak: at weekend i was trying several workarounds
10:03 masak edrin: I mean, this changes things a lot. and we need to take it into account.
10:03 edrin ah ok
10:04 edrin the workaround i tried would be quite nice but they did not fully work, i am not 100% sure why they do not fully work
10:04 masak neither am I.
10:04 edrin (most likely because of duplication of imports)
10:04 masak jonalv is our Spring guy.
10:04 edrin i can explain what i tried:
10:04 masak yes, please do.
10:06 edrin fo example one idea was to modify the rhino context ClassLoader before the script is run.
10:06 olass maclean: I reopened bug #335. Do you think you have time to look at it?
10:07 edrin before the script is put in the context it collects the managers from the plugins, and then in my try it was using a (hidden) function in these managers to request an update on the plugins classpath
10:07 edrin like: xmpp.extendClassPath(ClassPath old_path);
10:08 CIA-55 bioclipse: ospjuth * r8671 /bioclipse2/trunk/plugins/net.bioclipse.jmol/s​rc/net/bioclipse/jmol/editors/JmolEditor.java: Fixed an NPE when a ResourceChangeListener was not removed properly (probably on Bioclipse abnormal termination).
10:08 edrin in my plugin i could then load for example xws4j.jar into the old_path...
10:09 edrin but this resulted in incomatibility of xws4j objects because they were spawned with different library instances...
10:10 edrin unfortunately i found no way to "append" the existing classpath of xws4j plugin to the rhino plugin classpath
10:11 edrin it seems a classpath must be loaded new each time...
10:11 edrin but there must be some way to do this... OSGI is doing it, too
10:11 edrin hm
10:13 edrin another way I was thinking about is modifying rhino in some way to support a midification of the Packages.net.bc...  trick. for example to add other classloaders to it's internal search mechanism...
10:13 edrin maybe that's the best way to go
10:13 edrin but i did not figure out how to do this yet
10:14 edrin unfortunately I do not know how "Packages." of rhino works...
10:15 edrin masak: more i can't tell right now
10:16 edrin hm
10:16 masak ok.
10:17 masak @tell jonalv http://irclog.perlgeek.de/bi​oclipse/2009-02-09#i_893980
10:17 zarah masak's link is also http://tinyurl.com/djh44r
10:17 zarah Consider it noted.
10:18 edrin masak: ah, maybe i found a way
10:18 masak oh, how?
10:18 edrin it's a hack but could work as we want it...
10:18 masak :)
10:19 edrin one would have to derive a class from ClassLoader and overwrite all functions... then we can make it to search in different *sub* classloaders
10:19 edrin hm
10:20 edrin maybe we do not even have to do this
10:20 edrin because I would expect that OSGI is doing exactly this
10:20 masak aye.
10:20 edrin maybe they have this "special" ClassLoader
10:20 edrin class
10:20 edrin hm
10:21 edrin there should be some osgi class that extends java.lang.ClassLoader
10:21 edrin and this should then have something like osgiclassloader.addClassLoader(loader);
10:22 masak mm
10:22 edrin this way one could collect the class loaders of the managers/plugins
10:25 maclean olass : about bug #335...
10:25 olass Gpox: what is the syntax to set a new Font in JCP?
10:25 olass maclean: yes
10:25 maclean I changed it to exporting within the workspace as that did not suffer from the bug
10:25 Gpox RendererModel.setFontName and .setFontManager( BOLD|NORMAL)
10:26 maclean but I agree that exporting outside would be better
10:26 olass Gpox: So the following should work: model.setFontName("courier");
10:26 olass maclean: good
10:26 olass Gpox: How do I set the size of the font?
10:26 maclean olass: yes
10:26 maclean olass: you can't
10:26 maclean sorry
10:27 Gpox if new Font("courier" ...) work in swt yes
10:27 * Gpox points at maclean
10:27 maclean <_< ... >_>
10:27 maclean there is a reason for this
10:28 olass if shown in non-symbol mode (no symbols but just atoms as squares) how do I make the atoms bigger?
10:28 olass looking at the SDFEditor, it is really hard to see the atoms
10:28 olass they are simply to small
10:28 maclean mom
10:29 Gpox maclean: base font size, to make them larger at a given scale
10:29 maclean gpox : agreed
10:29 maclean that was always the intention - to allow the user to set what size at 100% zoom
10:30 olass I want to be able to make atoms larger (if described by Font or just a square) so that when I color them they are easier to see
10:30 maclean olass: ack. the compact elements are not yet paramatised
10:30 olass can I do that in SDFEditor currently?
10:30 maclean the BasicAtomGenerator uses a value of 0.1 in model space
10:30 olass maclean: that is an RFE I'd very much like to see
10:31 maclean it should probably use atomRadius in the renderer model
10:31 olass maclean: should I file a bug?
10:31 maclean yes
10:31 olass oki
10:32 olass what do you call it when atoms are displayed as squares (and not symbols)?
10:32 olass minimized mode?
10:32 maclean compact mode
10:32 olass ok
10:32 olass and else?
10:32 olass expanded mode?
10:32 olass standard mode?
10:33 maclean err
10:33 maclean not sure
10:33 maclean !compactMode
10:33 olass :-)
10:33 masak SUV mode.
10:33 maclean :)
10:33 olass maclean: assign to you?
10:33 maclean sure
10:34 maclean Gpox : did you want to be able to set the compact mode shape - SQUARE|OVAL...?
10:35 olass maclean: bug #430 filed
10:35 maclean ok, cool
10:35 Gpox would you not change the generator for that?
10:36 maclean well that would work also, but it could be set in the renderer model
10:36 maclean like with selection shape
10:36 olass maclean: any more comments on bug #335?
10:38 maclean yes : the point I was making was that the SaveAsDialog (workspace export) worked on osx, but the other one (FileDialog : export anywhere) suffered from that odd bug
10:38 maclean So, unless I make a SWT dialog that exports outside the workspace (or if one exists) then there will be the same problem
10:41 olass hmm
10:41 olass SWT dialog should not be restricted to workspace by default
10:41 olass SWT is not an Eclipse Resources component
10:42 olass very strange
10:42 maclean Well, maybe there is some kind of option - I'll take a look
10:47 olass great
10:52 edrin masak: maybe it is not too difficult to extend ClassLoader
10:53 edrin and then one can pass this extended class loader to the manager's (hidden) special function and the manager then just appends its own classloader to the ClassLoaderCollector
10:54 edrin have to leave, cu later
10:54 masak edrin: ok
10:55 masak I've always wanted a good reason to extend ClassLoader. seems this might be it.
10:55 CIA-55 bioclipse: egonw * r8672 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Use the Activitor to get the manager; fixed to use a ICDKManager.save() method
10:56 CIA-55 bioclipse: egonw * r8673 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Use @BeforeClass to instantiate only one instance of the manager
10:57 CIA-55 bioclipse: egonw * r8674 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Keep track of the underlying problem.
10:57 CIA-55 bioclipse: egonw * r8675 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Forced loading of the org.springframework.bundle.osgi.extender, to fix the time out problem you get otherwise
11:00 olass maclean, Gpox: When I click the icon to add atoms and then hover over an atom, it is not visualized as selected. This makes it a but odd when adding new atoms.
11:00 olass is this a known bug?
11:00 olass else I'd like to file it
11:01 egonw olass: jonathan and I just sat down on those UnsupportedBla exceptions
11:01 egonw resulting in the above commits...
11:01 olass ok, good
11:01 olass problem solved?
11:01 maclean olass : not sure - shouldn't happen
11:01 egonw olass: no
11:01 olass maclean: what you mean?
11:01 egonw 14/49 unit tests still fail
11:02 maclean I mean that sounds like a bug
11:02 egonw still with Unsupported stuff
11:02 olass maclean: will file it then
11:02 olass egonw: what do you think about it?
11:03 egonw dunno...
11:03 egonw need lunch first now
11:03 olass ok
11:03 egonw will then further look at it with J
11:03 * olass too
11:03 olass great
11:04 egonw it seems there is a problem with IMockFile's
11:04 egonw on top of the Unsupported thingy
11:06 egonw olass: I don't think things will be ready by tomorrow...
11:06 olass ok
11:06 egonw olass: and reverting to a revision before the @Job removal would be rather nasty too
11:06 olass agreed
11:06 egonw but maybe we are lucky this afternoon
11:06 olass let's wait and see
11:07 olass else we reschedule accordingly
11:07 olass things need to be in order
11:07 olass it is very important
11:07 egonw I fully agree
11:15 CIA-55 bioclipse: ospjuth * r8676 /bioclipse2/trunk/plugins/net.bioclipse.jmol/ (icons/capture-screen.png plugin.xml): Updated icon for "Export as Image" in Jmol.
11:16 olass egonw: could we add some sample actions that should be available for all ICDKMolecules when right-clicked upon?
11:16 olass If you give me one proposal, I'll implement it
11:16 olass egonw: will file you a bug for this
11:16 egonw Show InChI ?
11:17 olass is that an expensive operation?
11:17 egonw for larger molecules, yes
11:17 olass I mean, should that not just always be visible in Properties?
11:17 olass ok
11:17 olass then it is not in properties
11:44 CIA-55 bioclipse: goglepox * r8677 /bioclipse2/trunk/plugins/net.bioclipse.cdk.ui.sdf​editor/src/net/bioclipse/cdk/ui/sdfeditor/editor/ (MoleculeTableContentProvider.java MoleculeTableViewer.java): Added 10 property columns to molecule table fixes bug #174
11:44 zarah bug #174 | http://tinyurl.com/79fbdp
11:59 maclean cool
11:59 maclean quite large sdf files open very nicely
12:03 maclean "Created Index in 14884ms" for an 140Mb file with 3.6 million lines
12:07 olass very nice
12:07 maclean joined #bioclipse
12:15 edrin olass: what was the thing with the pdb files? they will not work in bc right now?
12:15 olass cdk cannot parse them
12:15 olass it is not supported
12:15 olass hence we must use files :-(
12:15 edrin oh
12:15 egonw but you can open them with Jmol
12:15 olass yes
12:15 edrin ok ic
12:16 olass but must serialize to files (which is bad)
12:16 edrin it is bad? really?
12:16 olass yes
12:16 olass why save to file if user dowsn't want to?
12:16 olass just takes up space
12:17 maclean it will take up space at some point, as a temporary file
12:17 olass yes
12:18 edrin o
12:18 edrin ok
12:18 olass currently there is no /Temp
12:18 edrin and the virtual folder is not good for this?
12:18 edrin why remove the virtual folder?
12:18 edrin hm
12:18 olass Virtual is not good as it stores the persisted file in memory AND the parsed file in memory too
12:18 olass really dedundant
12:19 olass if a large file, bad
12:19 edrin ic
12:19 olass Virtual will be changed to a file-based Temp dir in bc 2.1
12:19 edrin getting large files with webservices is not possible anyway ;)
12:19 edrin ok
12:19 olass edrin: Bioclispe is larger than webservices
12:20 edrin sure
12:20 olass so we need to design accordinly
12:20 edrin for now dump the websservices results in virtual and later in temp "by default", with option to change it?
12:20 olass sounds like a good idea
12:21 edrin olass: a database plugin to put stuff into would be nice
12:21 olass structureDB will be in 2.1
12:21 edrin or something similar to this
12:21 edrin structureDB is not limited to structures?
12:21 olass sure
12:22 olass there will be sequenceDB too in due time
12:22 edrin ah cool
12:22 olass in due time = no one has even started that
12:22 olass :-)
12:23 edrin sure
12:23 olass but important
12:23 olass might use Biojava's BioSQL
12:27 CIA-55 bioclipse: ospjuth * r8678 /bioclipse2/trunk/plugins/net.bioclipse.jmol/s​rc/net/bioclipse/jmol/editors/JmolEditor.java: Attempt to reduce flickering in Jmol when Bioclipse window moved.
12:49 * maclean wonders how hard it would be to make an SWT-Jmol...
12:50 * egonw suggests maclean to look at the Jmol Feature, which is exactly aiming to do that, but not finished yet :)
12:50 egonw oh...
12:50 egonw sorry, mixing up SWT with RCP
12:50 maclean No an SWT _rendering_ Jmol...
12:50 olass maclean: interesting idea indeed!
12:51 egonw umm... you mean a SWT widget?
12:51 * maclean quickly points out that I am not going to attempt it
12:51 egonw or SWT-application
12:51 egonw maclean should be grant writing :)
12:51 maclean I mean that Jmol does custom graphics anyway
12:51 egonw maclean: the current Jmol in Bioclipse uses a SWT widget
12:51 maclean very low level, but in pure java
12:52 olass Google Sommer of Code project?
12:52 maclean it uses an SWT_AWT composite
12:52 egonw what?
12:52 egonw who did that?
12:52 egonw we had a perfectly nice SWT-Jmol in BC1
12:52 maclean really?
12:52 olass what?
12:52 olass did we?
12:52 egonw well, IIRC that's what I wrote
12:53 olass it was not at all perfect
12:53 olass had no good context menu
12:53 olass selections were shaky
12:53 olass but it was indeed very nice
12:53 olass A Google SoC project to write an SWT version of Jmol was just a proposal :-)
12:54 egonw well, I am sure I wrote a SWT widget without _AWT for Jmol...
12:55 egonw BC1 does not seem to use it in 1.2
12:55 egonw OK, another 2.1 task
12:55 egonw maclean: you are perfeclty right... it create an Image
12:55 egonw which we can easily copy to a SWT image and draw
12:55 egonw that
12:55 egonw that would also solve the flickering, I think
12:55 maclean indeed
12:56 egonw and I really recall doing that...
12:56 maclean I'm looking at the bc1 source
12:56 olass egonw: if I right-click an entry in an SDFile, what would you like to be able to do?
12:56 egonw because it was that code that caused a problem with how SWT handled transparency
12:56 olass apart from extract it to another file?
12:56 egonw olass: nothing much at this moment
12:56 olass why not?
12:56 egonw olass: select entries, then right-click Extract ...
12:56 olass you said generate inchi 1 hour ago
12:57 egonw was that for SDF?
12:57 egonw that was for ICDKMolecule
12:57 olass ICDKMolecule
12:57 olass an entry in SDFEditor is 1-N ICDKMolecules
12:57 egonw ah... ok
12:57 olass hence clicking an entry in MoleculesTable is an ICDKMolecule
12:57 egonw ic where you're going...
12:57 olass yes, want some sample actions for this
12:58 olass to demonstrate the power :-)
12:58 olass will add extract to file
12:58 olass but
12:58 olass do you want option to generate 3D for individual entries?
12:58 egonw no, rather not
12:58 olass does not seem appropriate to me...
12:58 olass ok
12:58 egonw because the SDFEditor doesnät to 3D
12:59 olass not yet
13:00 edrin olass: btw. the bioschemata and hivpred services are not running right now
13:00 olass what!?
13:00 olass :-(
13:01 olass I wonder why they crashed...
13:01 olass I have not restarted the ejabberd
13:01 edrin do you have a log? i guess they were disconnected because of netsplit or something?
13:02 olass what is a netsplit?
13:02 edrin routerproblem
13:02 edrin network connection problem
13:02 edrin http://en.wikipedia.org/wiki/Netsplit
13:02 zarah edrin's link is also http://tinyurl.com/9w9xy
13:03 olass log does not say anything
13:03 olass ends with a stanza sent :-(
13:03 edrin and the ejabberd log?
13:04 olass mom, need to look that up
13:04 edrin most likely there lot of stuff in between
13:04 edrin was experimenting with egons RDF service
13:05 olass no error in ejabberd.log either
13:06 olass it is clearly only my services that are down
13:06 olass cdk, qsar, rdf work
13:07 olass :-(
13:07 olass edrin: do you log on error in xws4j component?
13:07 olass I get no error output at all
13:11 olass egonw: is it INChI ?
13:11 egonw InChI
13:11 olass only lowercase h?
13:11 olass ok
13:11 olass sure?
13:11 egonw (In)ternational (Ch)emical (I)dentifier
13:11 maclean http://en.wikipedia.org/wiki/In​ternational_Chemical_Identifier
13:11 zarah maclean's link is also http://tinyurl.com/bzt78b
13:12 egonw (formerly: (I)nternational (N)IST (Ch)hemical (I)dentifier
13:12 egonw )
13:12 maclean hmmm. it's InChI (TM)
13:12 egonw (formerly: (I)nternational (Ch)hemical (I)dentifier -> IChI)
13:12 egonw ummm... no, like this:
13:12 egonw (In)ternational (Ch)emical (I)dentifier
13:13 egonw formerly: (I)UPAC/N(IST) (Ch)emical (I)dentifier
13:13 egonw formerly: (I)UPAC (Ch)emical (I)dentifier
13:13 egonw you feel the policitics here :)
13:13 maclean indeed
13:19 olass egonw: what is required of a molecule in order to generate inchi?
13:19 olass 2D?
13:19 olass 3d?
13:19 olass no coords required?
13:19 egonw neither
13:19 egonw just connectivity
13:19 olass ok
13:23 edrin olass: errors should be written to console
13:23 olass I get none
13:23 olass so dunno what went wrong :-(
13:23 edrin olass: i think your components were disconnect on firday or something
13:24 edrin why, i dont know
13:24 edrin egonw 's RDF service i killed on friday or saturday when calling it 200x in one second...
13:25 egonw I think the HTTP it did in the backend killed it...
13:25 edrin (this bug we have to fix at some time)
13:25 edrin yes
13:25 edrin that's what i think, too...
13:28 olass ehabberd log file doesn't tell
13:28 olass I cannot help
13:28 olass I can restart them
13:29 edrin ok
13:29 edrin are they running on a different machine? maybe this is the reason, technical connection problems
13:29 olass yes they are
13:29 olass but so is egons
13:31 egonw ports are OK?
13:32 egonw they were already down when I rebooted mine...
13:32 egonw what happens if you boot them now?
13:33 olass ok, booted them
13:33 olass looks fine in dosco
13:34 olass disco
13:39 egonw shk3: ping
13:40 shk3 hi egonw
13:40 CIA-55 bioclipse: ospjuth * r8679 /bioclipse2/trunk/plugins/n​et.bioclipse.ui/plugin.xml: Created a contenttype for MDLMolfile that is a supertype to MDLMolfile2D and MDLMolFile3D.
13:40 egonw shk3: do you work with oracle?
13:41 egonw olass++
13:42 shk3 no
13:43 egonw ok, going to ask anyway...
13:43 egonw do you know if oracle can export SQL?
13:43 egonw ASCII SQL maybe even?
13:43 masak @google can oracle export SQL
13:43 zarah masak: http://www.orafaq.com/wiki/Import_Export_FAQ
13:43 masak ta-da!
13:44 egonw http://www.oracle.com/technology/​oramag/oracle/06-may/o36sql.html
13:44 zarah egonw's link is also http://tinyurl.com/busa5x
13:44 egonw that one seems better
13:44 egonw it's a bit vague
13:44 masak ok :)
13:44 egonw "yes, but"
13:44 egonw that seems the answer...
13:45 maclean egonw: we could ask Mark when he's around
13:45 egonw yeah, a shame I missed him...
13:46 egonw I hope he will be around at the CDK workshop...
13:46 maclean when is that, by the way
13:46 maclean ?
13:46 maclean (I know that I should know, but I don't)
13:47 egonw april-ish
13:47 egonw yes, we need to set up a good workshop homepage
13:47 egonw and start advertsize things...
13:47 maclean true
13:57 shk3 sorry, can't say anthing about oracle exports
13:57 egonw ok, thanx
14:28 olass egonw: I'd like your blessing to rename net.bc.inchi.business to net.bc.inchi as it now contains a command and a menu, and I donät think it's large enough to motivate a separate UI plugin
14:29 egonw fine, except that it breaks using .business for where the manager is...
14:29 egonw just that I can say later 'told you so' :)
14:29 olass that should be for package names solely
14:29 olass IMO
14:29 olass managers are defined as OSGI components
14:30 olass hence independent of plugin ID
14:33 olass egonw: I'd like to request that you don't log anything in format detection
14:33 olass now, just moving down in Bioclipse Navigator logs things
14:33 olass please turn that off
14:34 olass it is VERY annoying
14:34 egonw olass: please file a bug report
14:35 olass ok
14:39 olass shk3: please revisit bug #273
14:39 shk3 will do so
14:40 CIA-55 bioclipse: ospjuth * r8680 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business/src/net/bioclipse/cdk/d​omain/CDKMoleculeSelectionHelper.java: Constructed a CDKMoleculeSelectionHelper that parses an IStructuredSelection into a List<ICDKMolecules>.
14:41 olass Gpox: Could bug #208 be moved to 2.0?
14:42 olass We will most likely postpone RC1 until thursday or friday
14:44 CIA-55 bioclipse: ospjuth * r8681 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.ui/src/net/bioclipse/cdk/ui/domain/ (. CDKMoleculeTester.java): Implemented a PropertyTester that can be used to target actions for ICDKMolecules having distinct properties. Currently has2D and has3D are implemented but could easily be extended.
14:45 Gpox sure, i can have a look just tell me how to prioritize
14:48 olass what are you doing now?
14:49 olass cmlmultieditor?
14:49 olass http://pele.farmbio.uu.se/cgi-bin/bugzilla/buglis​t.cgi?bug_file_loc=&amp;bug_file_loc_type=allword​ssubstr&amp;bug_id=&amp;bug_status=NEW&amp;bug_st​atus=ASSIGNED&amp;bug_status=REOPENED&amp;bugidty​pe=include&amp;chfieldfrom=&amp;chfieldto=Now&amp​;chfieldvalue=&amp;deadlinefrom=&amp;deadlineto=&​amp;email1=arvid&amp;email2=&amp;emailassigned_to​1=1&amp;emailassigned_to2=1&amp;emailcc2=1&amp;em​ailreporter2=1&amp;emailtype1=substring&amp;email​type2=substring&amp;field-1-0-0=target_milestone&​amp;field-1-1-0=bug_status&amp;field-1-2
14:49 zarah olass's link is also http://tinyurl.com/cyukps
14:49 masak that's one impressive URL.
14:50 olass pelezilla
14:50 egonw luckily we have tinyurl
14:50 masak too late :)
14:50 olass masak: I have an rfe
14:50 masak olass: let's hear it
14:51 olass @pz user=ola target=2.0 lists all olas bugs for target=2.0
14:51 zarah http://tinyurl.com/bql2pv
14:51 olass easy to list one persons bugs for a certain target
14:51 Gpox olass: no, Version 2.1 target milestone 2.0 ?
14:51 masak olass: it's a pretty tall order. care to make that into a bug ticket?
14:52 olass sure
14:52 masak dz
14:52 olass Gpox: what you mean?
14:52 Gpox is that for 2.0 or 2.1 i think we said 2.1
14:53 olass well, could we reschedule multiselections for 2.0? That was my question...
14:53 olass or did I read you wrong?
14:54 olass Gpox: target Milestone is what is important. Version we don't really use yet...
14:54 Gpox you asked if i worked on cmlmultieditor
14:54 olass Gpox: might it be so that you have mixed up version and target milestone properties in pelezilla?
14:55 olass version is what system you report on
14:55 olass target milestone is when it should be ready
14:55 shk3 olass: revisited #273
14:55 shk3 see comment
14:55 olass shk3: ack
14:56 olass Gpox: I have problems listing your bugs for target milestone 2.0
14:56 olass could you paste me a link?
15:00 Gpox it seams ok, there are some that is set for version 2.1 instead of target milestone 2.1
15:02 Gpox and it is you who have them mixed up, you changed the version
15:09 CIA-55 bioclipse: ospjuth * r8682 /bioclipse2/trunk/plugins/ne​t.bioclipse.inchi.business/ (7 files in 4 dirs):
15:09 CIA-55 bioclipse: * Renamed net.bioclipse.inchi.business to net.bioclipse.inchi
15:09 CIA-55 bioclipse: * Added command, handler, menu to generate inchi from 1-N Files and ICDKMolecules in mixed selections.
15:09 Gpox http://pele.farmbio.uu.se/cgi-bin/bugzilla/buglist​.cgi?query_format=advanced&amp;short_desc_type=all​wordssubstr&amp;short_desc=&amp;product=Bioclipse+​2.0&amp;version=2.0&amp;version=2.1&amp;version=un​specified&amp;target_milestone=2.0+release&amp;lon​g_desc_type=substring&amp;long_desc=&amp;bug_file_​loc_type=allwordssubstr&amp;bug_file_loc=&amp;dead​linefrom=&amp;deadlineto=&amp;bug_status=NEW&amp;b​ug_status=ASSIGNED&amp;bug_status=REOPENED&amp;ema​ilassigned_to1=1&amp;emailtype1=substring&amp;emai
15:09 zarah Gpox's link is also http://tinyurl.com/cwpeo9
15:10 olass Gpox: all your bugs are important for 2.0 release
15:10 olass how much time do you expect for all of them?
15:11 olass lets talk more tomorrow and discuss priorities
15:12 * egonw is ready to release CDK 1.1.5
15:12 egonw aka CDK1.2 RC1
15:14 maclean good good
15:25 * olass has updated http://wiki.bioclipse.net/index.php​?title=How_to_add_menus_and_actions
15:26 olass we're getting there step by step
15:28 egonw so, who has been googling for "IAtomContainer PDB file example" ??
15:29 CIA-55 bioclipse: ospjuth * r8683 /bioclipse2/trunk/plugins/net​.bioclipse.cdk.ui/plugin.xml: Declared the propertytester in plugin.xml.
15:30 maclean egonw: I don't know, why?
15:30 egonw well, someone did today :)
15:30 CIA-55 bioclipse: ospjuth * r8684 /bioclipse2/trunk/plugins/net.bioc​lipse.cdk.ui/META-INF/MANIFEST.MF: Added deps to core expression for propertytester.
15:30 egonw and since we discussed PDB/CDK here today
15:31 maclean how did you find this out?
15:31 maclean (it wasn't me)
15:31 egonw :)
15:31 egonw piwiki
15:32 egonw one of the Hosted Apps at SF
15:32 maclean ohh.
15:37 CIA-55 bioclipse: ospjuth * r8685 /bioclipse2/trunk/features/net.bioclips​e.chemoinformatics_feature/feature.xml: Updated feature since inchi pluginID changed.
15:41 olass Gpox: reopened and commented on bug #174
15:45 maclean it's a bit dispiriting for the fixer when bugs are reopened :)
15:47 edrin egonw: could you make the RDF Service async?
15:47 egonw yes, I guess so...
15:47 egonw but not before Fri
15:48 edrin we need a long lasting job
15:48 edrin what needs long?
15:48 edrin a blast
15:48 edrin but we do not have a blast database
15:48 maclean generating 3d coordinates can take time
15:48 edrin ok
15:48 CIA-55 bioclipse: egonw * r8686 /bioclipse2/trunk/cdk-externals/trunk/ (46 files in 44 dirs): Uploaded CDK 1.1.5
15:49 maclean I was trying for superspiro.mol, and that was taking ages (or it was hanging)
15:49 egonw UFF forcefield geometry optmization
15:49 egonw superspiro is not so large, is it?
15:49 edrin egonw: the generate3Dcoordinates takes long? is it async set?
15:49 maclean no, but it was taking ages
15:49 egonw edrin: no, that's not async either
15:49 egonw will do that one too
15:49 edrin egonw: but this one can take ages, not?
15:50 edrin no?
15:50 egonw 3D generation should not take ages
15:50 egonw OK, got to catch the bus now
15:50 edrin maclean said it does
15:50 edrin hrm
15:50 egonw 3D generation is a basic look up...
15:50 egonw yes, about that... I expect a bug...
15:50 egonw ok, bye gyes
15:50 maclean oh, well, maybe a bug then
15:50 egonw ok, bye guys
15:50 maclean bye
15:50 egonw see you online tonite
15:50 edrin bye
15:50 edrin hehe
15:51 maclean no rest for the wicked. or cdk developers. or wicked cdk developers.
15:56 CIA-55 bioclipse: goglepox * r8687 /bioclipse2/trunk/plugins/net.bioclipse.cdk.u​i.sdfeditor/src/net/bioclipse/cdk/ui/sdfedito​r/editor/MoleculesEditorLabelProvider.java:
15:56 CIA-55 bioclipse: Added Generated Molecules table if coordinates was generated
15:56 CIA-55 bioclipse: bug #434
15:56 zarah bug #434 | http://tinyurl.com/cvuche
16:06 jonalv oh hai
16:06 zarah oh hai jonalv
16:06 zarah jonalv: You have new messages. Write '/msg zarah @messages' to read them.
16:07 jonalv zarah: how many?
16:07 zarah jonalv: that was not so easy for a little bot to understand
16:17 maclean what is this, a masked ball?
16:19 masak sorry :/
16:19 masak olass: when's this week's status meeting?
16:19 olass I guess we need one tomorrow
16:20 masak I guess so too.
16:20 olass egonw: ok with you?
16:20 jonalv edrin: ping
16:20 maclean he's gone
16:20 olass right
16:20 maclean egon, that is
16:23 jonalv @pz
16:23 zarah jonalv: Pelezilla is at http://bugs.bioclipse.net
16:28 jonalv uhm, olass Gpox egonw: CDKManager doesn't even compile for me... Van someone confirm_
16:28 jonalv ?
16:28 olass compiles here
16:28 jonalv olass: laterst?
16:28 olass ah, no
16:28 olass same problem here
16:29 olass probably a missing commit from egonw
16:29 * masak pulls
16:29 jonalv The method guessFormat(InputStream) is undefined for the type ReaderFactory
16:29 olass please email egonw
16:29 olass so he can fix that asap
16:29 masak aye, b0rken here too.
16:31 olass I sent egon an sms
16:33 maclean sms! maybe he can commit that way too - text patches directly to the repository :)
16:33 masak he'd better. :)
16:33 olass he answered: "Ok, first thing I'll do when at home"
16:34 jonalv @pz 437
16:34 zarah http://tinyurl.com/bu9hgy
16:38 * olass will work late tonight
16:44 * jonalv is trying to get rid of a few bugs now...
16:44 jonalv (mostly by tossing them off to someone else I am afraid...)
16:45 jonalv =D
16:48 egonw joined #bioclipse
16:48 jonalv anyone wants a bug from me?
16:48 jonalv @pz 41
16:48 zarah http://tinyurl.com/67mutl
16:48 jonalv it's a very intresting problem
16:48 jonalv masak: do you want bug #41?
16:49 jonalv anyone?
16:49 masak jonalv: not really, no.
16:49 jonalv olass?
16:49 jonalv egonw?
16:49 jonalv Gpox?
16:49 egonw no, thanx
16:49 jonalv edrin?
16:49 jonalv :)
16:49 egonw got so many on my own list
16:49 * jonalv does not really want it
16:49 masak I can suggest a solution, and then sit and watch when someone else implements it, though. :)
16:50 jonalv masak: what's you suggested solution?
16:50 egonw what's 41 about?
16:50 masak jonalv: having five or so numbered log files, and rotating them.
16:50 egonw ah, yak
16:51 egonw no, don't want that one
16:51 masak ...or not.
16:51 jonalv I am pretty sure the logging framework has some solution for this
16:51 masak oh, that's good then.
16:51 jonalv I just don't know how to configure it
16:51 jonalv I have logging and logging hates me right back
16:51 jonalv |hate
16:51 jonalv we hates it
16:51 * olass will be away for some time
16:52 masak olass: o/
16:52 Gpox jonalv: i can take it
16:52 olass brb
16:52 jonalv woho
16:52 maclean log by dates?
16:52 jonalv zarah: praise Gpox
16:52 zarah jonalv: excuse me?
16:52 jonalv Gpox++
16:53 Gpox masak: pleas pleas put your suggestion in a comment
16:53 masak Gpox: awright.
16:53 masak but do recall that egonw rates it as 'yak'.
16:54 masak (which, I guess, means big and hairy and smelly)
16:54 egonw masak: he, it seems I had the spelling right :)
16:54 masak indubitably. :)
16:55 Gpox what size should the log files be?
16:55 maclean joined #bioclipse
16:55 masak http://bitworking.org/news/39​0/text-editor-saving-routines
16:55 masak Eclipse--
16:55 zarah masak's link is also http://tinyurl.com/7exbjt
16:56 Gpox any suggestions?
16:57 masak Gpox: to what?
16:57 Gpox log file size
16:57 Gpox max size i should say
16:57 masak ah.
16:57 masak that's a hard question.
16:57 masak 1Gb?
16:58 egonw 55% of the hard disk space?
16:58 Gpox my file is 27MB and if we should have five of them ...
16:58 CIA-55 bioclipse: egonw * r8688 /bioclipse2/trunk/plugins/net.bioclipse.cd​k.business/src/net/bioclipse/cdk/business/ (CDKManager.java CDKManagerHelper.java): Fixed the compile errors... sorry, guys
16:58 egonw ok, make that partition space
16:58 jonalv egonw: @pz 437
16:59 egonw @pz 437
16:59 zarah http://tinyurl.com/bu9hgy
16:59 egonw yes, fixed
17:00 Gpox uhmn i think that the size was the reason for the bug in the first place
17:00 masak "Kanye West wants the world to know that he's not doing bisexual porn, contrary to the claims made by those who have allegedly hacked his Gmail, Twitter, and MySpace accounts." -- http://arstechnica.com/web/news/2009/01/week-in-we​b-doomed-dvds-muxtape-returns-wikipedia-drama.ars
17:01 zarah masak's link is also http://tinyurl.com/cc6mks
17:01 egonw ummm... ok, but then splitting it up in 5 files does not make sense...
17:01 egonw or does it?
17:01 masak egonw: no, it doesn't.
17:01 masak we need a fixed size.
17:02 Gpox masak> jonalv: having five or so numbered log files, and rotating them. ??????
17:02 edrin jonalv: yes?
17:02 jonalv I don't think
17:02 jonalv so
17:02 jonalv hi edrin
17:02 jonalv edrin: I saw your comment about Spring making it impossible to do
17:02 masak Gpox: maybe rotate by day instead of a fixed size?
17:02 jonalv uhm
17:02 jonalv what do you call it?
17:03 edrin the Dynamic-Import thing in manifest but this is not good anyway
17:03 jonalv edrin: anyway I am pretty sure it's not Spring's fault. But what happends if you try doing it?
17:03 edrin jonalv: it simply does not work :)
17:03 edrin jonalv: but it is not the way to go anyway
17:04 jonalv edrin: as in you get an error?
17:04 edrin we rather need a certain extendable classloader
17:04 edrin jonalv: the error is, i cant see the exported packages of other plugins in my plugin
17:04 jonalv edrin: Uhm I am not 100% sure I understand when this is needed
17:05 edrin jonalv: it is needed if you want to use certain features of a lib provided with a plugin
17:05 jonalv you want to instantiate java classes directly from JS?
17:05 edrin yes
17:05 edrin sure
17:05 edrin very important
17:05 jonalv is that the only case where you run into trouble or?
17:05 jonalv or are there more situations?
17:05 edrin no, it's basically the reason
17:06 edrin but it's an important reason
17:06 jonalv it's a bit messy
17:06 edrin what do you mean?
17:06 jonalv do you do some form of import in the beginning of such a script?
17:07 jonalv you would have to import _everything_
17:07 jonalv it doesn't exactly scale
17:07 edrin that would be one option
17:07 jonalv but I guess the manager system suffers from the same problem...
17:07 edrin no, not everything but a lot
17:07 edrin ?
17:07 edrin no, managers are fine
17:08 jonalv all managers must be instantiated (meaning that all plugins with managers need to be started) before the JS-environment can be used
17:08 edrin in fact i do not want to import everything, therefore the Dynamic-Import I do not like so much
17:08 edrin on the other hand the advantage of dynamic-import would mean that the stuff is better controled...
17:08 jonalv edrin: so what do you want to import and how would you like to define what to import in each case?
17:08 edrin hm
17:09 edrin jonalv: for example i need to import the xws4j .jar
17:09 edrin and it must be the same classloader (containing it) as it is used by the xws4j plugin
17:09 edrin not a parallel instance
17:10 jonalv are you doing binding to the thread?
17:10 edrin so, the Dynamic-Import would scale quite nice, but unfortuanately it does not work in bc2, no clue why
17:10 edrin binding to the thread?
17:10 edrin i dont understand
17:11 jonalv never mind then
17:11 jonalv uhm could it have changed in Eclipse 3.4.1?
17:11 jonalv I know they talked about removing it...
17:12 edrin really?
17:12 egonw OK, going for dinner
17:12 edrin where?
17:12 egonw bye for now
17:12 edrin jonalv: do you have an url?
17:12 jonalv or btw: Why do you have to use the same class loader?
17:12 jonalv edrin: nope
17:12 jonalv I just thought I heard something
17:12 jonalv might be imagination
17:13 edrin jonalv: yes, because i want to pass stuff generated with it to the manager
17:13 edrin jonalv: the "Packages." of rhino is a quite powerful and important feature
17:14 jonalv edrin: I still don't understand why it's important that it is done by the same class loader
17:16 jonalv Gpox++ //for accepting the bug
17:18 * jonalv leaves the ship
17:18 edrin jonalv: rhino has problems
17:18 edrin jonalv: i will explain
17:18 jonalv uhm could you do it in a mail?
17:18 jonalv I really have to get going...
17:18 edrin one sec :)
17:18 Gpox i will bring it up for discussion tomorrow, i is easy to implement just need to know how it should work
17:19 jonalv Gpox: what?
17:19 edrin for example if you have xmpp.function(blah input) and you get an instance of input now from another classloader rhino thinks it is not compatible with xmpp.function(blah input)
17:20 edrin jonalv: do you understand?
17:21 jonalv edrin: if you use interfaces?
17:21 jonalv that probabably wouldn't effect...
17:21 jonalv I see
17:21 edrin jonalv: it does not work with interfaces, actually that's what i need it for
17:21 jonalv it sort of makes sence
17:21 jonalv but it's a real problem I thin
17:21 jonalv k
17:22 jonalv I don't know what to do about it...
17:22 edrin jonalv: i need it because i must define a listener according to an interface
17:22 jonalv may I leave now?
17:23 masak no you may NOT!
17:23 edrin jonalv: solution 1 Dynamic import packages *, solution 2 modification of the classloader used by rhino's context, for this we must create a powerful classloader collector that extends ClassLoader
17:23 jonalv can't you just switch classsloader oin the thread? :)
17:23 masak I don't like either solution, to be honest. :/
17:24 masak do we have to?
17:24 jonalv masak: what do you suggest then?
17:24 edrin i think that probably 1 would be better because it only imports stuff that *is* exported by plugins
17:24 masak jonalv: I dunno. just... not doing it.
17:24 edrin while 2 will bring in strange stuff that *sould* be hidden
17:24 masak edrin: aye.
17:24 jonalv but #1 isn't working...
17:25 masak edrin: have you figured out why #1 isn't working yet?
17:25 edrin masak: if we do not do it we loose the power of javascript
17:25 masak edrin: you have a point.
17:25 masak but I'm still only partly convinced of that point.
17:25 jonalv personally I don't understand why the manager can't create your listener...
17:26 jonalv oh
17:26 jonalv aha\
17:26 jonalv never mind...
17:26 jonalv :)
17:26 masak jonalv++
17:26 edrin jonalv: doesnt work with rhino
17:27 edrin jonalv: does work from java sure but not from rhino i think
17:27 masak jonalv will be right back after these commercials.
17:29 edrin hm...
17:31 jonalv oh well masak and me heads out to swim now. You all have fun now! now.
17:31 olass sure
17:31 edrin bye
17:32 olass jonalv, masak: cu tomorrow
17:32 olass at bmc
17:34 masak olass: not if I see you first! :P
17:34 * masak goes swimming
17:35 olass jonalv: take care of masak
17:35 olass too late :-)
17:35 olass edrin: I'll be around later if you want to discuss xws
17:36 edrin olass: ok, but will be here only very late
17:36 olass how late?
17:36 edrin around 22:00
17:36 edrin and after
17:36 olass that's fine with me
17:37 edrin have to meet my parents
17:37 olass say hi to them from me
17:37 olass :-)
17:37 olass well
17:37 edrin heh ok
17:37 edrin cu
17:37 olass bye
17:37 edrin left #bioclipse
18:02 maclean left #bioclipse
18:06 Gpox joined #bioclipse
18:10 olass grr
18:10 * olass hates bank-id
20:06 rojasm joined #bioclipse
20:10 CIA-55 bioclipse: ospjuth * r8689 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/ne​t/bioclipse/ui/ApplicationActionBarAdvisor.java: Merged update actions into one: Software Updates. Solves bug #340.
20:10 zarah bug #340 | http://tinyurl.com/acaxxk
20:15 CIA-55 bioclipse: ospjuth * r8690 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/ne​t/bioclipse/ui/ApplicationActionBarAdvisor.java: Added action for Dynamic Help. Solves bug #347.
20:17 egonw joined #bioclipse
20:17 ilogger2 joined #bioclipse
20:17 zarah joined #bioclipse
20:20 CIA-55 bioclipse: ospjuth * r8691 /bioclipse2/trunk/plugins/net.bi​oclipse.scripting.ui/plugin.xml: Removed Virtual Navigator since non-functional.
20:23 olass egonw: around?
20:23 olass egonw: My JCP is not working anymore
20:28 rojasm joined #bioclipse
21:08 CIA-55 bioclipse: ospjuth * r8692 /bioclipse2/trunk/plugins/net.biocli​pse.ui/src/net/bioclipse/ui/actions/ (AddExtensionsAction.java SoftwareUpdatesAction.java): Merge for Software Updates Action.
21:36 CIA-55 bioclipse: ospjuth * r8693 /bioclipse2/trunk/plugins/net.bioclipse.ui/ (3 files in 2 dirs): GUI tweakings: Filtered out unwanted Eclipse views and shortcuts, added Open Perspective submenu.
21:39 CIA-55 bioclipse: ospjuth * r8694 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/ne​t/bioclipse/ui/ApplicationActionBarAdvisor.java: Fixed bug that placed perspective menu in SHow View menu.
21:45 CIA-55 bioclipse: ospjuth * r8695 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Add Chemoinformatics Perspective as PerspectiveShortcut.
21:56 CIA-55 bioclipse: ospjuth * r8696 /bioclipse2/trunk/plugins/n​et.bioclipse.ui/plugin.xml: Filter out Eclipse's own Project Explorer View.
22:25 edrin joined #bioclipse
22:26 edrin hi
22:26 edrin hi olass
22:26 olass hi edrin
22:27 olass (me will go to bed soon
22:27 * olass will go to bed soon
22:27 edrin olass: currently looking at the webservice plugin
22:27 olass ok
22:27 olass how does it look?
22:27 edrin why should it not return IFile ?
22:28 edrin and where to save data? why not in virtual (and later in temp)=
22:28 edrin ?
22:28 olass we have decided to not work with Files in Bioclipse, hence all manager methods do not declare methods taking IFile as input as @published
22:29 olass can ywe not do what you ended up saying?
22:30 olass you may use Virtual if you like
22:30 olass it will just be removed
22:30 edrin sure
22:30 olass and I dislike everyone just dumping stuff there, expectibg users will look there
22:30 edrin we can change virtual to the temp folder later
22:30 edrin ah ok
22:30 olass users are not smart enough to search temp
22:30 olass they need more assistance than that
22:30 edrin so have a web service folder?
22:31 olass why not do as in BC1?
22:31 olass you lcicked in ProgressView on the job and got result opened in dialog
22:31 olass you remember?
22:31 edrin sure, that is possible, too
22:31 edrin the code is even available
22:32 edrin but what if you download 20 files?
22:32 olass I want two levels: One scripting (for power users) and one REALLY SIMPLE (for normal users)
22:32 olass edrin: what if?
22:32 edrin if you download 20 files, 20 times a dialog will popup
22:32 edrin hm
22:33 edrin in bc1 it was dumped in a virtual folder
22:33 olass yes, this was not so intuitive IMO
22:33 olass more of a quick and dirty solution
22:33 olass but fine in many cases I guess
22:33 edrin yes
22:34 edrin hm...
22:34 olass I agree, ResultDialog should not be mandatory
22:34 olass but be default :-)
22:34 olass powerusers should be able to bypass it
22:34 edrin this dialog is available... you see it if you download a file with the wizard
22:35 olass hmm, ok
22:35 olass haven't tested that
22:35 olass edrin: at this time you can do as you like, I won't argue
22:35 olass just take my comments into account, OK?
22:35 edrin sure
22:36 olass and please update and run the Plugin Junit test
22:36 edrin wondering how to popup a dialog where user selects the folder to save to
22:36 edrin must research in eclipse API
22:36 edrin complicate
22:37 edrin olass: ok, i will put stuff in virtual now and change this later, will now look in the junit thing
22:37 olass ok
22:42 CIA-55 bioclipse: ospjuth * r8697 /bioclipse2/trunk/plugins/​net.bioclipse.statistics/ (2 files in 2 dirs): Removed traces of BC1 code and updated manifest for 3.4.
22:42 CIA-55 bioclipse: ospjuth * r8698 /bioclipse2/trunk/ (3 files in 3 dirs): Updated for experimental update site.
22:43 CIA-55 bioclipse: ospjuth * r8699 /bioclipse2/trunk/experimental-updatesite/ (. artifacts.xml content.xml site.xml): Updated with QSAR feature as content.
22:46 olass that's it, I'm going to bed
22:46 olass edrin: will print and read XMPP article tomorrow afternoon. OK?
22:47 olass really tired now
22:47 olass g'night
23:24 olass joined #bioclipse
23:24 olass egonw: around?

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