Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-02-23

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

All times shown according to UTC.

Time Nick Message
07:59 Gpox joined #bioclipse
08:15 egonw joined #bioclipse
08:18 egonw moin
08:18 zarah ni hao egonw
08:35 CIA-55 bioclipse: egonw * r8966 /bioclipse2/trunk/plugins/net.biocl​ipse.scripting/schema/net.bioclipse​.core.scriptingContribution.exsd: Fixed number of contributions allowed, which should be one or more. The other stuff is done by Eclipse, and 3.4.1/.2 seems to fix a number of XML issues :) (fixes #552)
08:35 zarah bug #552 | http://tinyurl.com/dhz7hk
08:37 rojasm joined #bioclipse
08:40 egonw px 552
08:40 egonw @pz 552
08:40 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=552
08:49 CIA-55 bioclipse: edrin_t * r8967 /xws/trunk/xws4j-blastproxy-service/: Initial import.
08:53 masak joined #bioclipse
08:54 CIA-55 bioclipse: edrin_t * r8968 /xws/trunk/xws4j-blastproxy-service/ (65 files in 29 dirs):
09:10 shk3 joined #bioclipse
09:11 jonalv joined #bioclipse
09:11 jonalv oh hai
09:11 zarah saluton jonalv
09:11 egonw hi jonalv
09:12 jonalv egonw: your latest blog post need a screen dump :)
09:13 egonw why?
09:13 jonalv uhm
09:13 jonalv good question
09:13 jonalv it was a feeling I had
09:13 jonalv must think a little to figure out exactly why
09:13 jonalv oh now I see
09:14 jonalv Google reader dösn't pick up the gists
09:14 egonw ah...'
09:14 jonalv (google reader)--
09:14 egonw http://chem-bla-ics.blogspot.com/2009/02/so​lubility-data-in-bioclipse-2-handling.html
09:14 zarah egonw's link is also http://tinyurl.com/afnebj
09:14 egonw afnebj... I knew it... :(
09:15 egonw well... feel free to make a script that does this:
09:15 jonalv egonw: so that's a Bioclipse manager with state?
09:15 egonw use rdf to download the ONS data, use pubchem to get 2D diagrams, aggregate them as SDF and open that
09:16 egonw oi
09:16 egonw jonalv: yes, it might be...
09:16 jonalv egonw: that sounds cool
09:16 egonw there *is* a global model
09:16 jonalv egonw: just wanted to make you aware about that...
09:16 egonw which is not synchronized...
09:16 egonw good catch
09:16 jonalv *of that
09:17 egonw otherwise, I think it is rather cool :)
09:19 egonw jonalv: it's now sort of to you, for making anything ?s ?p ?mol, ?mol ?p2 ?o, ?mol a chem:Molecule persistent in structure db
09:20 jonalv egonw: Structuredb persists CDKMolecules as cml
09:20 egonw yeah, that's fine
09:21 egonw ?mol chem:hasConnectionTable ?cml
09:21 jonalv that ofcourse means that everything that doesn\t survive cml translation is lost
09:21 egonw no
09:22 egonw the above is on the annotation level
09:22 egonw you do not store the annotation as CML, or do you?
09:22 jonalv egonw: yet I do not imort the annotations atall
09:23 jonalv so they would be in cml
09:23 jonalv there are somet htings left to figure out here...
09:25 jonalv my computer is not working
09:25 jonalv needs a restart
09:25 jonalv bbl
09:25 egonw ack
09:42 jonalv joined #bioclipse
09:42 jonalv ...finally
09:43 jonalv restarting Mac are a bit like what I can imagin working with donkeys would be like
09:45 edrin joined #bioclipse
09:46 edrin hi
09:46 zarah hello edrin, you fantastic person you
09:47 masak hi edrin
09:47 jonalv hi edrin
09:47 edrin hi masak
09:47 edrin hi jonalv
10:03 edrin jonalv: masak is the jsexecution plugin clear to you?
10:04 masak clear in what way?
10:04 jonalv edrin: define clear
10:04 edrin (structure, files, class usage)
10:04 edrin what it does
10:04 * masak looks
10:04 edrin what the classes are supposed to do
10:04 rojasm joined #bioclipse
10:05 * jonalv tries to find it
10:05 jonalv edrin: you have put a js.jar in it
10:05 jonalv why you do this?
10:06 jonalv but youy are depending on the js bundle...
10:06 edrin jonalv: i added it already before you wrote your mail as i initially thought we would use our own .jar then. was thought to be prepared for you
10:06 jonalv oki
10:07 jonalv edrin: and this is working now?
10:07 jonalv we can add any amount of deps to that plugin and you won't mind? :)
10:07 edrin in fact i don't mind if we use the spring version, on the other hand in this case, as this is really a specific rhino plugin, we can add the js.jar ourself
10:08 jonalv edrin: there are other plugins using the js classes in Bioclipse as well
10:08 edrin jonalv: oh? what? plugins?
10:08 edrin what plugins?
10:08 jonalv scripting.ui at least.
10:08 jonalv I think one more
10:09 jonalv masak?
10:09 masak don't know offhand.
10:09 masak do a find.
10:09 edrin jonalv: in my opinion the dependency should be removed from this plugin, and the js specific code should go into the rhino plugin
10:10 jonalv edrin: I hear you but take that up with masak
10:10 edrin masak: what do you think?
10:10 masak edrin: with 'rhino plugin', you probably mean something else.
10:10 masak but, yes.
10:10 edrin this way one can plugin new scripting engines to the environment
10:10 masak I've followed the discussion, and I believe that separating out things is a good consensus action.
10:10 * jonalv is not sure...
10:10 edrin masak: i mean the jsexecution plugin
10:11 masak edrin: I thought so. :)
10:11 edrin must go
10:11 edrin visitor
10:11 edrin cue
10:11 edrin later
10:11 edrin left #bioclipse
10:11 masak did the visitor have a cue? :)
10:12 * masak has no clue
10:14 jonalv actually edrins editor is not working at all for me right now...
10:14 jonalv anyone else?
10:14 * masak will look at that too
10:14 jonalv egonw, masak?
10:14 egonw cannot try right now...
10:15 egonw trying to figure out how to output something to the console
10:15 egonw without gettting a invalid thread access error
10:15 egonw keep forgetting the proper way to do that...
10:15 egonw whether Display. bla... or some CONSOLE.bla
10:17 jonalv js.print
10:17 egonw yes, doing that now...
10:17 egonw and that gets me the exception
10:18 jonalv masak: sounds like someone forgot to do things in the ugi thread...
10:18 masak ugi!
10:19 masak and would I be that someone, you mean?
10:19 egonw ah... it's not the js code doing the exception
10:19 masak I should think not! :)
10:20 jonalv -me ducks
10:20 CIA-55 bioclipse: jonalv * r8969 /bioclipse2/trunk/features/net.​bioclipse_feature/feature.xml: added net.bioclipse.jsexecution to the bioclispe feature
10:21 jonalv now edrin++s editor should work
10:34 jonalv zarah: you should be able to ping me when edrin comes back online...
10:34 zarah jonalv: Hokay. You provide the commits, I provide the "should"-ing
10:37 CIA-55 bioclipse: egonw * r8970 /bioclipse2/trunk/plugins/net.bioclipse.pubchem/ (4 files in 3 dirs): Ported the Search on Pubchem wizard (fixes #366)
10:37 zarah bug #366 | http://tinyurl.com/az3ljl
11:15 CIA-55 bioclipse: egonw * r8971 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/net​/bioclipse/ui/contenttypes/CmlFileDescriber.java: Empty CML files are now opened by JCP
11:16 jonalv food
11:18 CIA-55 bioclipse: egonw * r8972 /bioclipse2/trunk/plugins/net.bioclipse.ui/src/net​/bioclipse/ui/contenttypes/CmlFileDescriber.java: Removed debug info again
11:19 masak egonw: was just going to notify you about the System.out.println in r8971 :)
11:20 egonw yeah, sorry about that
11:20 egonw @pz 498
11:20 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=498
11:36 CIA-55 bioclipse: egonw * r8973 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: Fixed writing of SDF files
11:36 maclean joined #bioclipse
11:43 CIA-55 bioclipse: egonw * r8974 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Removed to failing tests, which use a method whihch is not supposed to be used; since the method is tested anyway, it's save
11:47 CIA-55 bioclipse: egonw * r8975 /bioclipse2/trunk/plugins/net.bioc​lipse.cdk.business/src/net/bioclip​se/cdk/business/ICDKManager.java: Added missing method summary
11:48 egonw OK, CDKManagerTest 38/38
11:48 egonw green
12:02 CIA-55 bioclipse: gilleain * r8976 /bioclipse2/trunk/plugins/net.bioclipse​.ui/src/net/bioclipse/ui/contenttypes/ (CmlFileCoordinatesDescriber.java CmlFileDescriber.java): Copyright notices.
12:03 CIA-55 bioclipse: egonw * r8977 /bioclipse2/trunk/plugins/ (2 files in 2 dirs):
12:03 CIA-55 bioclipse: Revert "Removed to failing tests, which use a method whihch is not supposed to be used; since the method is tested anyway, it's save"
12:03 CIA-55 bioclipse: This reverts commit 471c3fd273b48efd023c5a880a37195f88b52c63.
12:03 CIA-55 bioclipse: Better option: just use the variant with a IProgressMonitor.
12:10 CIA-55 bioclipse: egonw * r8978 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business.test/src/net/bioclipse/​cdk/business/test/CDKManagerTest.java: Fixed test: getSMILES() is used, so had to make tin the input SMILES symmetrical or canonical. Since the last is a pain, I've chosen the first
12:25 CIA-55 bioclipse: jonalv * r8979 /bioclipse2/trunk/plugins/net.biocl​ipse.rhino/src/net/bioclipse/jsexec​ution/tools/MonitorContainer.java: Wrote container class for monitors
12:26 CIA-55 bioclipse: jonalv * r8980 /bioclipse2/trunk/plugins/ (4 files in 4 dirs): Use the new MonitorContainer for keeping a monitor updated when running JS scripts through the editor
12:26 CIA-55 bioclipse: jonalv * r8981 /bioclipse2/trunk/plugins/net.bioclip​se.scripting/src/net/bioclipse/script​ing/business/JSJobCreatorAdvice.java: Made the SubProgressMonitors use 0 ticks so that the overall progressmonitor can stop at 50% after 1 tick for the entire run.
12:28 egonw I hate it when something just decided to stop working after a svn update :(
12:29 jonalv egonw: what is not working?
12:29 egonw CDKManagerPluginTest...
12:29 egonw it was running an hour ago...
12:29 jonalv oh
12:29 jonalv strange
12:30 egonw and now it fails with a NPE cdk.business.Activator.getDefault()
12:30 egonw I guess I am not being warned about a missing plugin again...
12:30 egonw actually... is jsexecutor I guess...
12:30 jonalv I added taht one to the feature...
12:31 egonw yes, but Eclipse does not seem complain if you do not actually have that plugin in your workspace
12:32 jonalv egonw: no it doesn't. I learned that the hard way last week...
12:32 egonw and this is a big usability problem...
12:32 egonw even more than a broken SVN, I hate Eclipse silently eating critical bugs
12:34 egonw fuck
12:34 egonw that did not help
12:40 CIA-55 bioclipse: egonw * r8982 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Use the methods with a IProgressMonitor to fix a few UnsupportedMethod exceptions
12:43 egonw jonalv: ping
12:44 egonw or anyone... suggestions?
12:44 jonalv egonw: you pang?
12:44 masak egonw: did you just ping 'anyone'? :P
12:45 egonw can someone explain me why BC2 boots fine, but the PluginTest does not...
12:45 egonw failing on a cdk.business.Activator.getDefault() returning nul...
12:45 * masak can't
12:45 jonalv I have no idea
12:45 egonw osgi> ?? what again
12:45 jonalv ss
12:45 egonw crap
12:46 egonw in debug mode it does boot...
12:46 egonw this osgi stuff sucks at timing
12:46 masak a Heisenbug!
12:47 egonw oh, btw...
12:48 egonw the NPE is still there...
12:48 egonw just at a different time
12:48 egonw jonalv: so, what changed in the js stuff?
12:48 egonw because I can only imagine that to be the root of evil
12:48 edrin joined #bioclipse
12:48 egonw the axis of wickedness
12:48 jonalv egonw: alot...
12:48 edrin hi
12:49 zarah oh hai edrin
12:49 jonalv edrin: hi
12:49 jonalv edrin: I got it working now :)
12:49 edrin jonalv: good
12:49 egonw but not at my machine
12:49 edrin the jsexecution?
12:49 jonalv egonw: Bioclipse is not running at your machine?
12:50 egonw it is...
12:50 egonw but please try the CDKManagerPluginTest too
12:50 egonw that one is not
12:50 edrin egonw: if you want to run the blast service you must add all the jars from the project's jar folder to the classpath of the xws4j-component
12:50 egonw while it did prior to your js commits
12:50 egonw edrin: I have no idea what that means
12:51 masak edrin: what's your email address in Bugzilla? I'd like to assign a ticket to you.
12:51 edrin aliban at gmx.net
12:51 masak thanks.
12:51 edrin egonw: are you talking about the jseditor plugin?
12:52 egonw no clue
12:52 * egonw will locally revert the js commits, and see if the PluginTest runs again then
12:52 edrin masak: the blast service i uploaded to svn xws folder: named xws4j-blastproxy-service
12:52 masak edrin: what about it?
12:53 edrin egonw: jonalv changed the platform, reset target platform to get the js.jar from spring
12:53 edrin sorry
12:53 egonw edrin: so, CDKManagerPluginTest works for you?
12:53 edrin egonw: the blast service i uploaded to svn xws folder: named xws4j-blastproxy-service. there are jars in the /jars folder, you need them to run the blast service...
12:54 edrin egonw: I cant tell you, i am not on my computer
12:54 edrin masak: the question was directed to egonw
12:54 egonw edrin: then why do you think resetting target platform is going to help?
12:54 edrin egonw: because jonalv changed the spring base, the js.jar might be missing in your platform
12:55 edrin but this was friday I think
12:55 egonw edrin: well, you may have missed the start of the discussion... but BC2 has no problem booting...
12:55 egonw it's the test that does not
12:55 edrin and the test is associated with js ?
12:55 edrin ok
12:55 edrin no clue then :) I give up
12:57 egonw jonalv: your patches do not seem to be the source of the problem...
12:57 * jonalv will try running that test...
12:57 egonw with your patches reverted, I still can't start the tests :(
12:58 jonalv egonw: okey
12:58 egonw jonalv: but it still is a osgi problem
12:59 jonalv which test fails for you?
12:59 egonw all
12:59 egonw none run
12:59 jonalv they run for me...
12:59 egonw net.bioclipse.cdk.business.Activator.getDefault() returns null
12:59 jonalv 15 error 1 failures
13:00 edrin oh, i just realized it is named net.bioclipse.rhino in the svn?
13:00 egonw what is?
13:00 edrin egonw: not related to your problem
13:03 maclean joined #bioclipse
13:04 egonw maclean: do you have a ETA on the JCP changes wrt static zoom?
13:04 maclean I understood about half of that sentence
13:04 maclean it zooms right now
13:04 egonw scale
13:04 maclean jcp.setFitToScreen(false)
13:04 maclean jcp.setZoom(jcp.getZoom() * 0.5)
13:05 maclean jcp.updateView()
13:05 edrin jonalv: indeed this monitor was not much todo :)
13:05 jonalv edrin: sry?
13:05 maclean what it doesn't do is scroll
13:06 edrin jonalv: i mean it was only one line to add to get the monitor associated with your spring
13:06 jonalv edrin: ah no all the pieces where there and once I realised that I should use a weak HAsh map with Threads as keys it was pretty straight forward
13:06 egonw maclean: OK, that script works...
13:06 maclean good
13:06 egonw Gpox: the inital zoom is too large
13:06 jonalv edrin: I have added a complete class to that plugin..
13:06 masak jonalv: so a threadsafe hashmap wasn't necessary?
13:06 egonw and I don't see any margin anymore...
13:06 edrin yes, have seen it
13:07 egonw but that's because fitToScreen is false, I guess?
13:07 jonalv masak: it would have been usefull but I synched all calls to it instead...
13:07 maclean initial zoom is too LARGE? TOO LARGE?
13:07 egonw yes, *TOO* *LARGE*
13:07 maclean initial zoom is fit-to-screen
13:07 maclean as that is what was requested
13:08 masak :)
13:08 maclean margin is my fault.
13:08 edrin must go again
13:08 edrin cu
13:08 edrin left #bioclipse
13:09 egonw jonalv: start is not being called...
13:10 egonw cdk.business.Activator.start()
13:11 egonw at least, not before cdk.business.Activator.getDefault() is called...
13:15 jonalv egonw: strange but it seems to work for me so I am afraid I can't help you...
13:17 egonw jonalv: to whom should I assign the bug report?
13:17 CIA-55 bioclipse: egonw * r8983 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/Activator.java: Properly fail when initialization did not work
13:28 shk3 I am having problems compiling bc
13:28 shk3 the JsEnvironment in net.bc.scripting
13:29 shk3 wants stuff from core which is not accessible
13:29 shk3 I exported everything in core, but I still get
13:29 shk3 Access restriction: The type Context is not accessible due to restriction on required library net.bioclipse.core/libs/js.jar
13:29 jonalv shk3: that's strange. What stuff?
13:29 shk3 import org.mozilla.javascript.Context;
13:29 shk3 import org.mozilla.javascript.EcmaError;
13:29 shk3 import org.mozilla.javascript.EvaluatorException;
13:29 shk3 import org.mozilla.javascript.Scriptable;
13:29 shk3 import org.mozilla.javascript.ScriptableObject;
13:29 shk3 import org.mozilla.javascript.WrappedException;
13:30 jonalv but there shouldn't even be a js.jar in core
13:30 jonalv edrin, masak: _Why_ is there a js.jar in core?
13:32 shk3 guys, could you try to sort this out?
13:33 shk3 right now, I can't do anthing in bioclipse.
13:33 shk3 will continue jcp for the time being.
13:34 masak jonalv: it probably shouldn't be there.
13:35 egonw shk3: why are they undoredo classes start with Swing* ?
13:36 shk3 because the implement the UndoableEdit interface from swing
13:36 shk3 please see the wiki page I created
13:36 shk3 it explains inter alia this
13:36 egonw cdk wiki?
13:37 shk3 yes
13:37 shk3 mom
13:37 egonw found it
13:37 shk3 http://apps.sourceforge.net/mediaw​iki/cdk/index.php?title=Undo/Redo
13:37 zarah shk3's link is also http://tinyurl.com/csg5q5
13:43 egonw jonalv: the plugin test works again...
13:43 egonw no clue what is really different now... :(
13:49 CIA-55 bioclipse: gilleain * r8984 /bioclipse2/trunk/cdk-externals/trunk/or​g.openscience.cdk.control/src/org/opensc​ience/cdk/controller/ControllerHub.java: Fixes bug #467 - jcp.zap() was removing all atoms from all atom containers but leaving the atom containers.
13:49 zarah bug #467 | http://tinyurl.com/c6x6os
13:49 egonw maclean: but it still leaves one, I hope?
13:49 maclean no
13:49 egonw why not?
13:49 egonw would make better memory use if it did...
13:50 maclean why
13:50 egonw because instantiating an IAtomContainer is expansive
13:50 maclean eh?
13:51 maclean pfft. compared to the 10,000 other things biocilpse is doing?
13:51 jonalv masak: http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=569
13:51 zarah jonalv's link is also http://tinyurl.com/d3m82c
13:51 maclean well, whatever - the behaviour of zap is now consistent with the hub's constructor
13:52 masak jonalv: I agree with egonw on that one.
13:54 * jonalv gives up
13:54 masak overwriting files without warning when saving is a really bad default.
13:55 CIA-55 bioclipse: gilleain * r8985 /bioclipse2/trunk/cdk-externals/trunk/​org.openscience.cdk.render/src/org/ope​nscience/cdk/renderer/Renderer.java: A reset method, that is equivalent re-creating the same object.
13:58 jonalv egonw: ping
13:58 egonw pong
13:58 egonw was just going to say:
13:58 jonalv egonw: regardding 560
13:58 jonalv Jonathan, please file a bug report with the CML file that fails for you.
13:58 egonw if you think the API is inconsistent... I agree
13:58 egonw clear fault of no design
13:59 jonalv egonw: how would you check that?
13:59 egonw jonalv: add a unit test...
13:59 egonw with the file that fails to read for you...
13:59 jonalv egonw: are you gonna read through all files and figure out whter it's cml and whetther it's a molceuls?
13:59 egonw using the manager method which makes it fail for you
13:59 egonw jonalv: umm... yes... that's what the content-type describers are doing, not?
14:00 jonalv egonw: it is?
14:00 CIA-55 bioclipse: gilleain * r8986 /bioclipse2/trunk/plugins/ (3 files in 3 dirs): Paint the whole chem model in the editor, and move the scroll-handling and non-fittoscreen down from the PaintWidget.
14:00 * jonalv has no idea
14:00 jonalv so a file only called "bla" should be opened in jchempaint if it is correct cml for ethane?
14:00 Gpox I belive they check the extention first, to speed up things
14:01 egonw jonalv: I think you might find @pz 516 interesting
14:01 shk3 yes cml type is defined by extension
14:01 Gpox or can do it.
14:01 egonw shk3: no
14:01 shk3 no?
14:01 jonalv which is my point\
14:01 shk3 I looked it up a few days ago
14:01 jonalv a cml file is defined by the file ending
14:01 egonw not the CML content types which are supported by the editors
14:01 shk3 really
14:01 shk3 let me check
14:01 jonalv if saveCML does not add .cml then it will not always produce correct cml...
14:02 egonw jonalv: huh?
14:02 jonalv which I think it should...
14:02 egonw jonalv: what makes you think that?
14:02 jonalv egonw: from the users perspective
14:02 egonw jonalv: huh?
14:02 egonw jonalv: what makes you think that?
14:03 jonalv egonw: FRom the users perspective a file which Bioclipse recognises as a cml file and opens in jchempaint is a correct cml file containing a molecule
14:03 jonalv egonw: if Bioclipse does not open it as a Molecule in JchemPaint then the file is broken
14:03 egonw shk3: the top level CML content-type might still have cml as extension, but files with that type are not openend AFAIK
14:03 jonalv that's how it looks from the suers perspecitve...
14:03 egonw jonalv: yes, sounds accurate to me
14:03 egonw but that has nothing to do with file extensions
14:04 egonw file extension != file format
14:04 masak indeed.
14:04 jonalv egonw: it seems as if it is implemented like that right now anyway
14:04 shk3 sorry guys,
14:04 egonw jonalv: no, it is not
14:04 shk3 cml is defined by extension
14:04 egonw jonalv: open CmlFileDescriber
14:04 egonw shk3: what content-type ID?
14:04 shk3 <content-type
14:04 shk3 base-type="org.eclipse.core.runtime.xml"
14:04 shk3 id="net.bioclipse.contenttypes.cml"
14:04 shk3 name="Chemical Markup Language (CML)"
14:04 shk3 file-extensions="cml,xml"
14:04 shk3 priority="normal">
14:04 jonalv egonw: both Gpox, and shk3 says that it is like that...
14:05 egonw shk3: yes, that's the top level
14:05 egonw but JCP does not open such files
14:05 shk3 and this is the cml definition
14:05 egonw JCP opens a subtype
14:05 shk3 yes, but still it is the cml definition
14:05 shk3 and the subtype inherits
14:05 shk3 from cml content type
14:05 shk3 which means
14:05 shk3 according to my understanding
14:05 egonw then that's a bug
14:05 shk3 that if no extension
14:06 shk3 the subtype is not consdering it
14:06 egonw I'm fine with extensions speeding things up...
14:06 egonw but now it actually causes a bug...
14:06 egonw shk3: was not aware of this inheritance...
14:06 egonw apologies for that
14:06 jonalv okey I will file this new bug and close the old one
14:07 jonalv Good thing we cleared this out :)
14:07 egonw jonalv: masak prefered to keep the old one open
14:07 egonw until the new one is fixed
14:07 masak well, you may do as you wish, of course.
14:07 egonw jonalv: please add a dep
14:07 egonw ah, ok
14:08 egonw shk3: ping
14:08 masak my comment wasn't so much about dependencies as about the nature of file formats.
14:08 egonw shk3: what to do with with CML validation?
14:08 shk3 ?
14:08 shk3 why do we need to do something?
14:08 egonw shk3: keep tied to this top level CML type?
14:09 shk3 can't we leave as it is?
14:09 shk3 yes
14:09 shk3 absolutly
14:09 egonw OK
14:09 shk3 1. because
14:09 shk3 not all valid cml has a subtype
14:09 shk3 (only single molecule etc. have, but there are other possibilities)
14:09 egonw yes
14:10 shk3 2. because restricting this action only to valid cml files does not make sense
14:10 egonw but if we make the top level assume CML namespace
14:10 shk3 I tried to get this throught for ages
14:10 egonw it likely means that the doc has to be well-formed
14:10 shk3 , but it seems hard to understand.
14:10 shk3 so I repeat:
14:10 shk3 this action makes only sense if it bound to something which also may be invalid cml files
14:11 shk3 having a "check valid cml" only on valid cml is nonsense.
14:11 jonalv LOL
14:11 shk3 so jonalv at least got the point
14:11 shk3 it seems
14:11 egonw shk3: there is also something as well-formed, and not well-formed
14:11 jonalv perhaps we don't need that action at all then?
14:11 egonw shk3: that's what I was asking about
14:11 egonw not about validness
14:12 egonw shk3: but I guess that point still got lost
14:12 shk3 must valid cml not be well-formed?
14:12 egonw yes
14:13 shk3 so?
14:13 egonw what about CML files that are not well-formed?
14:13 egonw or not recognizable as CML?
14:13 egonw e.g. DTD-based files?
14:14 shk3 not valid, I would say
14:14 egonw without actually running the validation?
14:14 masak "well-formed" only pertains to XML nesting and such.
14:14 egonw look, this is the other way around...
14:14 masak "valid" is about validation against a schema.
14:15 egonw masak: no
14:15 masak egonw: http://en.wikipedia.org/wiki/XML#.22Well-formed.22_and_.22Valid.22_XML_documents
14:15 zarah masak's link is also http://tinyurl.com/evvbp
14:15 egonw masak: can be against a DTD, Schematron...
14:15 masak egonw: I know.
14:15 masak I meant 'schema' in a general sense.
14:15 egonw anyway...
14:15 egonw shk3: at least I now understand the scope you want to validate to work in...
14:16 egonw CML files which are already somewhat valid
14:16 shk3 well, I really don't want to discuss this every day for hours, but no, I can't see why not run this on any sort of file.
14:16 shk3 if it is not well formed xml
14:17 shk3 ->not valid cml
14:17 shk3 if yes, check against schema or whatever
14:17 shk3 check attribute namespaces
14:17 shk3 and so on
14:17 masak shk3: so you're simply saying that things which are not well-formed cannot be valid?
14:18 masak shk3: if so, then I agree fully. you're right, per definition.
14:18 shk3 yes
14:18 shk3 I think that's easy
14:18 egonw interestingly, validation is often run to see why something is not valid...
14:18 jonalv shk3: Okey so you want cml validation on all files? There are about a zillion other file formats used in life science. You realsie that if each of them should have an action on all files then there will be quite a big context menu...
14:19 shk3 well, that's why having it on all *.xml and *.xml files
14:19 shk3 seems fine.
14:19 jonalv shk3: yes, but then someone wants Bioclispe to not care about file endings.
14:20 jonalv egonw, shk3: Either we use file endings or we don't. We can't mix and match...
14:20 egonw jonalv: agreed
14:20 shk3 well, then we need to come up with another definition of the
14:20 shk3 cml type
14:20 egonw shk3: net.bioclipse.cml.handlers.validatecml should use the XML content-type
14:20 shk3 we can still have the validate on it
14:21 egonw shk3: will send a patch
14:21 shk3 what is the XML content-type defined by?
14:21 egonw Eclipse
14:21 egonw shk3: it's what the top CML content-type extends
14:21 shk3 well, yes
14:21 shk3 but what is a xml file for Eclipse?
14:21 egonw mom, checking the source code
14:22 egonw org.eclipse.core.internal.c​ontent.XMLContentDescriber
14:23 grandm44 joined #bioclipse
14:31 jonalv egonw: you are wrong, there is no method saveCML taking a boolean force
14:31 egonw save*Molecule*()
14:32 jonalv egonw: okey, but I want to save cml...
14:32 egonw save*Molecule*()
14:32 jonalv egonw: or don't I?
14:32 egonw jonalv: yes, I guess you do
14:33 jonalv egonw: What is the difference and how do I do it?
14:33 egonw one of the 50 saveMolecule() methods allows you to pick a file format
14:33 egonw but it default to CML anyway
14:33 egonw diff between what?
14:33 jonalv saveCML and saveMolecule
14:35 egonw saveCML does not allow saving as other format...
14:35 grandm44 hello
14:35 zarah saluton grandm44
14:35 grandm44 hi bot
14:35 shk3 hi grandm44
14:35 egonw [14:58:47] <egonw> if you think the API is inconsistent... I agree
14:35 grandm44 where is net.bioclipse.jsexecution ? I can't see it on the svn
14:35 jonalv egonw: aha, perhaps you could take it upon you to do something about it?
14:36 CIA-55 bioclipse: gilleain * r8987 /bioclipse2/trunk/cdk-externals/tr​unk/org.openscience.cdk.control/sr​c/org/openscience/cdk/controller/ (AddRingModule.java ControllerHub.java): Unattached phenyls now placed properly. Fixes bug #509.
14:36 zarah bug #509 | http://tinyurl.com/akjzwp
14:36 egonw maclean++
14:36 egonw you beat me too it...
14:36 egonw that was an important one
14:36 maclean hah
14:36 egonw jonalv: yes...
14:36 jonalv egonw: awesome
14:37 shk3 grandm44: i have the same problems
14:37 shk3 updated and stuff is not found
14:37 egonw jonalv: it just did not have priority... and people were changing it all the time anyway...
14:37 jonalv egonw: I see
14:38 jonalv shk3, grandm44: I think it mgiht still be called net.bioclipse.rhino....
14:38 jonalv edrin should do something about that...
14:38 grandm44 jonalv: thx, i try now
14:38 masak jonalv: aye.
14:38 masak jonalv: I've sent him a ticket.
14:40 shk3 for me, also com.springsource.org.mozilla.javascript is not found
14:40 shk3 and lots of these access problems
14:40 jonalv shk3: reset target platform
14:40 shk3 trying
14:40 jonalv it should be there...
14:41 egonw shk3: and make sure you updated the net.bioclipse.platform folder, and add that to you eclipse workspace
14:41 shk3 ah, that helps
14:41 shk3 thanks jonalv
14:41 grandm44 jonalv, shk3: that's it: take net.bioclipse.rhino for net.bioclipse.jsexecution
14:42 shk3 ok
14:42 shk3 compiles for me now
14:42 jonalv @pz 522
14:42 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=522
14:52 edrin joined #bioclipse
14:53 jonalv @pz 567
14:53 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=567
14:54 jonalv edrin: I think that bug: #567 is creating problems for people here...
14:55 CIA-55 bioclipse: shk3 * r8988 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.ui.sdfeditor/src/net/bioclipse/cdk​/ui/wizards/NewSDFileWizardPage.java: default file name is unnamed.sdf
15:01 edrin jonalv: strange, I dont know why this happened
15:01 edrin in fact i thought i renamed the project properly
15:04 jonalv mhm
15:05 jonalv I don't know. I only know that people have problems with it...
15:05 egonw shk3: as undo/redo is widget set independent... do you think it could be possible to have undo() and redo() as methods in the hub?
15:05 shk3 hm
15:06 shk3 you mean method calls for doing the undo and redo?
15:07 egonw yes
15:07 shk3 but that is using widget stuff
15:08 shk3 e. g. in jcp swing classes are used for posting the events
15:08 edrin jonalv: there is no net.bioclipse.rhino anymore... although the stuff is checked in in the wrong directory...
15:09 shk3 and keeping the stack etc.
15:09 shk3 you mean we should do that ourselves?
15:09 edrin net.bioclipse.rhino.rhinoeditor !?
15:09 jonalv edrin: that sounds wrong...
15:10 egonw shk3: dunno... that could be reason why that would not be possible...
15:11 edrin net.bioclipse.rhino.rhinoeditor <- what is this? i think this does not exist
15:11 shk3 I think it could be done, but would require some work
15:11 shk3 plus in elcipse, if you post to the eclipse undo/redo stack,
15:11 edrin indeed some stuff must be cleaned :/
15:11 shk3 it probably get "hooked up" to the menu automatically
15:12 shk3 if it's in the hub, we would need to make the menu/toolbar actions to call the hub
15:12 egonw ok, another good reason not to...
15:12 egonw thanx for the info
15:13 edrin egonw: do you think there is a way to find out, why the cdk schema does not compile with XMLBeans? we should really try to solve this...
15:14 egonw can XMLBeans be run from the command line?
15:14 edrin egonw: yes
15:15 edrin the tool you would need to use it "scomp"
15:15 egonw scomp?
15:15 egonw ok..
15:15 edrin http://xmlbeans.apache.org/
15:16 edrin Compile your schema. Use scomp to compile the schema, generating and jarring Java types. For example, to create a employeeschema.jar from an employeesschema.xsd file:
15:16 edrin scomp -out employeeschema.jar employeeschema.xsd
15:16 edrin gtg
15:16 edrin left #bioclipse
15:21 olass joined #bioclipse
15:31 CIA-55 bioclipse: jonalv * r8989 /bioclipse2/trunk/features/net.​bioclipse.usermanager_feature/ (.project build.properties feature.xml): Deleted net.bioclipse.usermanager. Fixes bug: #267
15:32 CIA-55 bioclipse: jonalv * r8990 /bioclipse2/trunk/plugins/net.bioclip​se.scripting/src/net/bioclipse/script​ing/business/JSJobCreatorAdvice.java: Added some status information when running JS-scripts
15:43 maclean eclipse crashed on me again. something to do with target.platform :(
15:48 maclean aaaarghhggghghh
15:50 jonalv maclean: Eclipse crashed? I thought Eclipse was sort of stable...
15:50 maclean ppft.
15:50 jonalv :)
15:50 maclean nope, every time I double-click the target.platform file it crashes.
15:50 jonalv wow
15:51 jonalv that's sort of problematic...
15:51 maclean Java frames: (J=compiled Java code, j=interpreted, Vv=VM code)
15:51 maclean j  org.eclipse.swt.internal.carbon.OS.HI​ScrollViewSetScrollBarAutoHide(IZ)I+0
15:51 maclean dammit dammit dammit
15:54 egonw jonalv: ping
15:54 egonw shk3: ping, if you want to talk a bit more about content-types
15:55 shk3 well, not too keen on it
15:55 egonw file extensions are important for things in the GUI...
15:55 shk3 but if there is something worth discussing
15:55 maclean oh well, back to running by plugins not features. good while it lasted (all of 3 days)
15:55 egonw so, while things have improved, Eclipse still needs .xml, .cml :(
15:56 egonw there is a Eclipse bug report on this, which I attached to 560 and 570
15:56 egonw but the bottom line is... /Virtual/bla will not be recognizable as CML file
15:56 egonw even if the Described is perfectly fine with that :(
15:56 egonw Describer
15:56 egonw *unless*
15:56 CIA-55 bioclipse: gilleain * r8991 /bioclipse2/trunk/cdk-externals/trunk/​org.openscience.cdk.render/src/org/ope​nscience/cdk/renderer/Renderer.java: Margin is back.
15:57 egonw the user adds bla as extension for the CML atom type
15:58 egonw I don't want to restrict on the manager level how people name there files...
15:59 egonw but the dummy user should be encouraged to use a registered file extension...
15:59 egonw which is a bit nasty...
15:59 egonw because we do not know what those are...
15:59 egonw any plugin can add extensions, and people can register additional manually...
16:00 egonw practically this means, that there is *no* way to decide what file extension should be required, and when to throw a warning
16:00 shk3 did I understand correctly that editor associations always work on extensions?
16:00 egonw Eclipse--
16:00 egonw shk3: no, not generally...
16:00 egonw but
16:00 egonw yes, on a GUI level
16:01 egonw however
16:01 egonw not just based on the extensions you give or don't
16:01 egonw the file extensions are meant to decide only which icon to show in the navigator
16:01 egonw and which editors to show in the Open With menu...
16:02 egonw that's the scope...
16:02 shk3 ah
16:02 shk3 and which editor is actually used for opening?
16:03 egonw depends how the users opens the file
16:03 shk3 anyway
16:04 shk3 what does that mean for us?
16:05 egonw sorry... was away for a sec
16:05 egonw it depends also on:
16:05 egonw 1. what wat the last editor used...
16:05 egonw 2. what plugins the user has installed
16:06 egonw 3. what extensions it has registered
16:06 egonw what it means to us...
16:06 egonw wait for that eclipse bug to be fixed?
16:07 egonw I think the GUI should give a warning when no recognized ext is being used...
16:08 egonw but not sure how to decide if the file name is OK...
16:08 egonw see, 1,2 and 3
16:08 shk3 at which occasion?
16:08 jonalv egonw:
16:08 shk3 file creation?
16:08 jonalv @pz 574
16:08 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=574
16:08 egonw shk3: in the GUI at least...
16:08 egonw not in the manager method
16:08 shk3 ok
16:08 shk3 yes
16:08 egonw but no clue how to plug a warning in a general SaveAs dialog either...
16:09 olass egonw: we have to live with file extensions
16:09 olass but
16:09 olass we can add content types above that level IMO
16:09 olass I mean, many many many apps/OS work with file extensions
16:10 * olass wants to build a new beta
16:10 olass any blockers pending?
16:10 olass 574 is one I guess..
16:11 olass talked to PMR about file extensions and recognizing file formats
16:11 egonw jonalv: I cannot reproduce 574...
16:11 egonw New -> CML works here
16:11 olass his comments were: most files cannot be read by a describer
16:12 egonw I do not understand that statement...
16:12 jonalv egonw: I have both unnamed.cml and unname1.cml can that be related?
16:12 egonw define: read and describer
16:12 stain .xml is good enough for everyone!
16:12 olass egonw: any chance of requesting "one wizard to rule them all" (New Molecule) with a combobox to select format?
16:12 olass stain: even binary blobs?
16:12 egonw jonalv: yes
16:13 jonalv egonw: please try with that setup and see what happends to you...
16:13 stain olass: it was meant as irony.. Taverna 1 saves 4-5 different file formats as ".xml"
16:14 olass stain: sort of the same here, but we add .cml which is xml-based (just to increase complexity) ;)
16:14 egonw jonalv: it seems that Gpox fix was not 100% accurate
16:14 egonw stain: yes, that would be perfect... everyone using .xml
16:14 egonw stain: than our problem was solved
16:14 stain could just as well go with .file
16:14 stain it's a file, right?
16:14 olass egonw: read=easily parsed, describer=the code that inspects a file and decides if a contenttype is valid
16:14 egonw stain: yes, I like just no file extension :)
16:15 jonalv egonw: okey I see
16:15 egonw olass: I do not know the context, but most operating systems have really no problem with that set up...
16:15 egonw jonalv: will fix it
16:15 jonalv egonw: yes please :)
16:15 egonw olass: not sure what PMR had in mind with that statement
16:15 olass just that deciding file types is hard
16:15 olass really hard
16:15 jonalv egonw, olass Waht are you talking of?
16:16 olass jonalv: potatoes
16:16 olass :-)
16:16 olass jonalv: nvm
16:16 jonalv :(
16:16 olass where is jsexecution plugin?
16:16 jonalv olass: nvm
16:16 egonw olass: .rhino
16:16 jonalv :P
16:16 egonw but filed
16:16 olass ?
16:16 egonw but = bug
16:17 olass ah
16:17 jonalv olass: it's in the net.bioclispe.rhino folder IO thjionk... :(
16:17 olass oki
16:17 olass thx
16:17 olass about next beta
16:17 egonw olass: forget that beta for today
16:17 olass ok
16:17 olass tomorrow?
16:17 egonw go home, think about ethics, etc
16:17 olass lol
16:18 olass it's actually unethical for me having to take that class...
16:18 olass so boring
16:18 olass and focused on clinical studies
16:23 CIA-55 bioclipse: egonw * r8992 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Fail with a meaning full message
16:23 CIA-55 bioclipse: egonw * r8993 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Split up test
16:24 CIA-55 bioclipse: egonw * r8994 /bioclipse2/trunk/plugins/net.bioclipse.​chemoinformatics/src/net/bioclipse/chemo​informatics/wizards/WizardHelper.java: Fixed creating resources beyond unnamed1.ext (fixes #574)
16:24 zarah bug #574 | http://tinyurl.com/av6fzc
16:38 maclean joined #bioclipse
16:43 jonalv maclean: I am prettty sure I don't like this new behavior od the JCHempaint editor much better....
16:43 jonalv maclean: Now I get scroll bars...
16:43 jonalv and it rescales all the time... :)
16:43 jonalv maclean: When do you think we will see real zooming in jchempaint?
16:44 egonw maclean: I like the current better than the last
16:44 maclean o\
16:44 egonw maclean: definately in the right direction
16:45 maclean jonalv : jcp.setFitToScreen(false)
16:46 maclean or, wait like 10  minutes until I port the zoom-by-scroll wheel over from jcp-swing...
16:46 maclean and, yes the scrolling is not there yet
16:47 egonw dinner, bbl
16:48 jonalv maclean: ah so scrolling and zooming :)
16:49 maclean well, it now has the capability to zoom-to-fit on first opening, then using scrollbars when the diagram overlaps the screen, I just haven't set it up that way yet
16:50 jonalv okey
16:50 jonalv but real zoom are comming?
16:50 maclean eh? what? eh?
16:50 maclean eh?
16:50 jonalv zoom, like in jmol
16:51 jonalv or the sequnce editor
16:51 maclean jcp.setFitToScreen(false); jcp.setZoom(jcp.getZoom() * 0.5); jcp.updateView()
16:51 jonalv or google earth
16:51 maclean or Quake.
16:51 jonalv I think you get my point
16:51 jonalv it is comming, right?
16:51 maclean gah!
16:51 edrin joined #bioclipse
16:51 jonalv gah?
16:52 masak g'ah!
16:52 maclean if you'll stop asking me about it, I will finish porting the changes from swing-jcp...
16:52 maclean :}
16:52 jonalv uhm :)
16:52 jonalv but it is comming? right? ;)
16:52 maclean YESSSSS
16:53 masak HSSSSSSS!
16:53 jonalv weeeee
16:53 jonalv maclean++ //Real zoom is comming to jcp
16:53 masak http://bash.org/?400459
16:53 maclean ohhhh
16:53 maclean python, right?
16:53 masak aye :)
16:53 jonalv masak: when do you think we are gonna be commming to the pool? ;)
16:54 maclean (I've been to that site too much :)
16:54 masak jonalv: uhm.
16:54 masak jonalv: your inability to spell 'coming' seems to get progressively worse.
16:55 jonalv masak: do you mean I might be commmming down with something?
16:55 masak probbbbably
16:55 masak jonalv: next you will be asking how babby is formed.
17:15 jonalv w00t
17:23 edrin left #bioclipse
17:42 maclean awww yeah.
17:49 CIA-55 bioclipse: gilleain * r8995 /bioclipse2/trunk/ (39 files in 6 dirs):
17:49 CIA-55 bioclipse: - Porting undo/redo capabilities from swing-jcp.
17:49 CIA-55 bioclipse: - Moving some control classes that were in the render package due to incorrect annotation.
17:49 CIA-55 bioclipse: - Implemented zooming using the mouse wheel.
17:52 egonw maclean++
17:52 maclean did you try it?
17:52 egonw no, not yet
17:52 egonw but since we hope to release beta3 tomorrow...
17:52 maclean indeed
17:52 egonw I was kinda hoping you had :)
17:52 maclean I did try before committing, yes :)
17:55 egonw booting...
17:56 egonw zooming works...
17:56 maclean good
17:56 egonw I do no see Undo in the Edit menu get active
17:56 maclean Oh, ha. No.
17:56 maclean I haven't actually /implemented/ undo_redo yet...
17:57 egonw ok, no worries
17:57 maclean I just ported over all of shk3++ 's work :)
17:57 egonw was actually suprised to see it show up in the commit message :=
17:57 egonw was actually suprised to see it show up in the commit message :)
17:57 maclean indeed.
17:57 maclean It's next on the list, though...
17:57 shk3 maclean++
17:57 egonw undo and redo are greyed out
17:58 egonw but that's fine for now...
17:58 egonw I can ask Arvid
18:03 edrin joined #bioclipse
18:06 edrin hi again
18:07 edrin egonw, did you try the blast service?
18:07 egonw no, ran into several bioclipse bugs...
18:07 egonw some new, some older...
18:07 edrin ok
18:11 CIA-55 bioclipse: egonw * r8996 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/Activator.java: Reverted patch. In non plugin tests this throwing makes CDKMolecule crash :(
18:16 CIA-55 bioclipse: edrin_t * r8997 /bioclipse2/trunk/plugins/ (net.bioclipse.jsexecution/ net.bioclipse.rhino/): renamed to jsexecution
19:13 CIA-55 bioclipse: egonw * r8998 /bioclipse2/trunk/plugins/ (4 files in 2 dirs): Added support to instantiate CDK IChemFormats used for saving files
19:32 CIA-55 bioclipse: egonw * r8999 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Identified failing unit test as bug in the CDK Mol2Writer
19:52 CIA-55 bioclipse: egonw * r9000 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Any exception thrown in a unit test is bad; only JUnit is downstream, so no reason to be explicit with all exception types
19:53 CIA-55 bioclipse: egonw * r9001 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Split test upto into individual testing: they should be unit tests
19:54 CIA-55 bioclipse: egonw * r9002 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Ignore tests: PDB is not yet supported... 2.1 is the target
20:33 CIA-55 bioclipse: edrin_t * r9003 /bioclipse2/trunk/plugins/ (10 files in 9 dirs): removed all "rhino" names
20:48 CIA-55 bioclipse: egonw * r9004 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Removed try/catch, as the test expects the exception to be thrown
21:28 CIA-55 bioclipse: egonw * r9005 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Split up into two unit tests
21:28 CIA-55 bioclipse: egonw * r9006 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Fixed loading of a test file: it is not available from the workspace, but from the classpath
21:29 CIA-55 bioclipse: egonw * r9007 /bioclipse2/trunk/plugins/net.bioclipse.​cdk.business.test/src/net/bioclipse/cdk/​business/test/CDKManagerPluginTest.java: Set resource, so that previous format can be determined; correct to actual given SMILES
21:29 CIA-55 bioclipse: egonw * r9008 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: Use the old file to determine the format, not the new one; complain if we try 'overwrite' and there is no old format; if not in overwrite mode, and still no format known, use CML
22:04 ChanServ joined #bioclipse
22:32 ChanServ joined #bioclipse

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