Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-02-25

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

All times shown according to UTC.

Time Nick Message
07:35 Gpox joined #bioclipse
07:39 masak joined #bioclipse
08:09 egonw joined #bioclipse
08:09 rojasm joined #bioclipse
08:11 egonw moin
08:11 zarah oh hai egonw
08:13 masak oh hai egonw
08:31 CIA-55 bioclipse: egonw * r9047 /bioclipse2/trunk/plugins/net.bioclip​se.data.sampledata/data/SDF/test.sdf: Removed trivial test file
08:32 egonw Gpox: fixed the missing props you requested
08:36 maclean joined #bioclipse
08:36 shk3 joined #bioclipse
08:42 olass joined #bioclipse
08:42 CIA-55 bioclipse: ospjuth * r9048 /bioclipse2/trunk/plugins/net.bioclip​se.core/src/net/bioclipse/core/domain​/props/BioObjectPropertySource.java: Now uses interfaces. fixes bug #591.
08:42 zarah bug #591 | http://tinyurl.com/csxg4x
08:42 olass hi
08:42 zarah hello olass, you fantastic person you
08:42 olass @botsnack
08:42 zarah ;)
08:45 egonw still 73 bugs marked M2.0
08:46 olass yes
08:47 olass hopefully we can get them down somewhat today
08:47 olass and release beta3 in the afternoon
08:47 olass need to go to class now :-(
08:47 olass be back at 15
08:47 olass @pz
08:47 zarah olass: Pelezilla is at http://bugs.bioclipse.net
08:49 olass @tell edrin: please check your bugs for bioclipse on pelezilla
08:49 olass zarah?
08:49 olass zarah: ping
08:49 olass @ping
08:49 olass masak: help! zarah is sleeping...
08:50 masak :/
08:50 masak zarah: reboot
08:50 * masak goes in to investigate
08:51 zarah Consider it noted.
08:51 masak she's just really slow.
08:51 masak don't know why.
08:51 olass google?
08:51 zarah joined #bioclipse
08:51 egonw no
08:52 egonw PZ is fine
08:52 masak egonw: what is it you're running on pele? like four java threads.
08:52 masak it seems to be eating some resources.
08:52 egonw good question
08:52 masak esp 14030.
08:52 egonw I think the XMPP services
08:52 masak :(
08:54 egonw no
08:54 egonw Nightly
08:55 masak they're running java?
08:58 egonw sure
08:58 egonw yes, no Perl[5|6]
09:02 egonw masak: if you have a java compiler written in Perl...
09:02 jonalv joined #bioclipse
09:04 jonalv oh ahi
09:04 jonalv hi
09:04 zarah ni hao jonalv
09:05 masak egonw: no, I don't. :)
09:06 * jonalv wonders if he do
09:08 edrin joined #bioclipse
09:13 maclean joined #bioclipse
09:13 edrin good morning
09:13 zarah edrin: You have new messages. Write '/msg zarah @messages' to read them.
09:13 edrin @msg
09:13 zarah edrin: 24 m 27 s ago, olass said please check your bugs for bioclipse on pelezilla
09:13 olass edrin: just emailed you
09:13 olass must run off to a course
09:13 olass back at 15
09:13 olass bye
09:14 maclean " if you have a java compiler written in Perl.." - do not speak of such things!
09:15 egonw #!/usr/bin/perl -e
09:15 egonw `java $*`;
09:16 egonw umm
09:16 egonw `javac $*`;
09:16 egonw just has a few deps
09:18 masak maclean: well, Perl 6.
09:19 maclean ohhh. perl6. well maybe, then.
09:19 maclean I do know of a project that translated perl code to java.
09:20 masak sounds painful.
09:20 masak on many levels, actually.
09:20 maclean yes, I suspect that it was.
09:20 jonalv edrin: is scriptingexecution still called rhino?
09:21 * jonalv can't get his Bioclispe to run... :(
09:21 jonalv not his Bioclipse either...
09:21 egonw jonalv: no, that's fixed
09:21 masak jonalv: I closed the bug about that, so it's fixed.
09:21 jonalv okey that's probably why I can't get things to work... :?
09:22 jonalv and the rhino one is still around? :/
09:23 maclean http://search.cpan.org/~philcrow/Java-Javap-0.04/
09:23 zarah maclean's link is also http://tinyurl.com/2bkmfk
09:28 edrin jonalv: there is no "rhino" stuff left
09:28 edrin i mean no net.bioclipse.rhino
09:28 edrin zarah: pz
09:28 zarah edrin: Pelezilla is at http://bugs.bioclipse.net
09:30 edrin jonalv: i think the thing should run fine
09:30 edrin jonalv: did you change the folder to checkout to net.bioclipse.jsexecution ?
09:30 edrin jonalv: do you consider 562 fixed now?
09:31 jonalv @pz 562
09:31 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=562
09:32 jonalv edrin: ah yea
09:32 edrin so i close it?
09:32 jonalv I have done that now :)
09:32 egonw edrin: it is custom that only the reported can close bugs
09:33 edrin ok
09:33 egonw edrin: if you fixed something, you never close the bug, but only mark it as fixed
09:33 egonw that allows the reporter to verify the fix
09:34 edrin ok
09:35 edrin jonalv: bug 561 and 572, 572 i can fix, regarding 561 i don't know yet how to solve, at least the most important stuff is printed. as i do not have much time now
09:36 edrin i would postpone both of them
09:36 edrin do you think this is ok?
09:36 jonalv @pz 561
09:36 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=561
09:36 jonalv @pz 572
09:36 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=572
09:37 jonalv edrin: are you at least printing a stack trace?
09:37 edrin no
09:37 edrin to the log`?
09:38 jonalv yea
09:38 jonalv I want to be able to debug and see wehat went wrong...
09:38 edrin ok
09:38 edrin i can try this
09:39 edrin but not today
09:46 CIA-55 bioclipse: shk3 * r9049 /bioclipse2/trunk/plugins/net.bioclipse.cdk.u​i.sdfeditor/src/net/bioclipse/cdk/ui/wizards/ (NewSDFileWizardPage.java NewSdfWizard.java): using the WizardNewFileCreationPage instead of NewSDFileWizardPage
09:51 CIA-55 bioclipse: jonalv * r9050 /bioclipse2/trunk/plugins/net.bioclip​se.jsexecution/src/net/bioclipse/jsex​ecution/tools/MonitorContainer.java:
09:51 CIA-55 bioclipse: Now returning NullProgressMonitor and printing warning to log if no
09:51 CIA-55 bioclipse: progressmonitor can be found connected to the current thread. (This should not
09:51 CIA-55 bioclipse: happen but sadly enough it does. If you find it fun to debug multi threading
09:51 CIA-55 bioclipse: feel free to give it a try...)
09:57 masak is a NullProgressMonitor a monitor which fails with a NullPointerException, or a monitor reporting the progress on null? :)
09:57 CIA-55 bioclipse: egonw * r9051 /bioclipse2/trunk/cdk-externals/tr​unk/org.openscience.cdk.ioformats/ (3 files in 2 dirs): Updated CDK to r14254: fixes missing file ext for MDL V2000 format, format name, and incorrect writer assigned to the original (non-v2000) MDL molfile format
09:59 CIA-55 bioclipse: egonw * r9052 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): Expose a method to guess the file format based on the file extension (blah, yuck)
09:59 maclean masak: it's what you get when your language doesn't have optional method arguments :)
10:00 CIA-55 bioclipse: egonw * r9053 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.jchempaint/src/net/bioclipse/cdk/j​chempaint/editor/JChemPaintEditor.java: In lack of a drop down box with formats, guess a format from the file extension (blah, yuck, but fixes #592)
10:00 zarah bug #592 | http://tinyurl.com/a9ub9n
10:01 masak maclean: indeed.
10:02 maclean Simply create a NullOptionArgumentClassFactoryConfigurator...
10:03 masak ...Manager.
10:03 egonw ...Thingy?
10:03 masak it's like this: http://steve-yegge.blogspot.com/2006/​03/execution-in-kingdom-of-nouns.html
10:03 jonalv uhm perhaps it should be abstract? :)
10:03 zarah masak's link is also http://tinyurl.com/kxubn
10:04 masak jonalv: good idea!
10:04 jonalv public class AbstractNullOPtionArgumentClassFactoryConfigurator implements INullOptionArgumentClassFactoryConfigurator
10:05 jonalv :)
10:05 masak sometimes I don't know whether to laugh or cry.
10:05 jonalv sometimes I don't know whther I am serious or not...
10:06 maclean :`)
10:06 masak jonalv: as long as you keep within 80 columns...
10:23 CIA-55 bioclipse: jonalv * r9054 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: [cdk.loadMolecules] only create subProject for every 100 molecule. MUCH FASTER
10:23 CIA-55 bioclipse: jonalv * r9055 /bioclipse2/trunk/plugins/net.bioclipse.script​ing/src/net/bioclipse/scripting/JsThread.java: Made the JSConsole execution add it's monitor to the MonitorContainer as well. Now we can get feedback from the JSConsole as well.
10:25 egonw jonalv: what GUI is actually using loadMolecules() ?
10:25 egonw only structuredb?
11:06 egonw maclean, Gpox: ping
11:06 Gpox egonw: pong
11:07 egonw it seems that drawing a bond between two exsiting atoms is broken
11:07 egonw that is... in draw mode, drag from one atom to another...
11:07 egonw only 'sprout' seems to be working now
11:08 Gpox yes, that is how it works
11:09 egonw any particular reason why the drag-draw is no longer available?
11:09 egonw that is a quite commonly used drawing practise...
11:10 Gpox I changed module
11:12 egonw why?
11:13 Gpox i changed it to the one the JCP application was using
11:13 egonw OK
11:13 egonw that's a good thing anyway...
11:14 egonw I guess drag-draw does not work in JCP either, then...
11:28 egonw Gpox: going commit a fix for the JCP editor now...
11:28 egonw please check it
11:29 egonw ok, being commited now
11:29 egonw 9056
11:30 CIA-55 bioclipse: egonw * r9056 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.jchempaint/src/net/bioclipse/cdk/j​chempaint/editor/JChemPaintEditor.java: Fixed saving of files created from an empty canvas: synchronize the CDKMolecule model with the ControllerHub model
11:36 maclean joined #bioclipse
11:36 maclean I'm back
11:44 maclean egonw : did you read the bug report on click-sprout drag-draw?
11:44 egonw maclean: no, have not seen that
11:44 egonw so, someone already filed a bug for that?
11:44 maclean mom just finding the number
11:44 maclean guess who :)
11:45 egonw maclean: btw, I was just bitten by the fact that the hub removed all atomcontainers upon .zap()
11:45 maclean oh?
11:45 egonw please check commit 9056
11:45 maclean ok
11:45 egonw I think the fix is OK, but please have a look yourself
11:46 maclean errr
11:46 maclean I think that fix is unecessary with the code I have locally
11:46 maclean sorry
11:46 maclean maybe not, not sure
11:49 maclean the molecule should be stored in the hub, and only the hub, I think
11:51 maclean ie : the 'model' (in the eclipse sense) of the JCPEditor is a ChemModel, not an ICDKMolecule, surely
11:51 egonw no
11:51 egonw actually not
11:51 egonw JCPEditor has a ICDKMolecule model
11:51 egonw line 69
11:51 egonw of JChemPaintEditor
11:52 maclean yeah, I do see it
11:52 egonw Bioclipse doesn't do anything other than molecules at this moment
11:52 maclean well, ok
11:52 maclean also @bug590
11:52 jonalv egonw: We also have IMolecule...
11:52 maclean @bug 590
11:52 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=590
11:52 jonalv egonw: but that only has a smiles string...
11:52 maclean 590 = the bond-drag stuff
11:52 jonalv sry SMILES string
11:54 maclean why do the models have to be these special eclipse-only things?
11:54 maclean sorry, bioclipse only. ICDKMolecules and so on.
11:54 * maclean feels there's something he's missing
11:56 egonw shk3: speclipse sort of works...
11:57 egonw lots of exceptions though...
11:57 egonw and axis errors could be handled better, but informing the user why something failed...
11:58 egonw two usability problems: SMR resource with more than one spectrum... how to switch between them...
11:58 egonw two usability problems: removing a peak from the SMR editor does not seem to work
11:58 egonw ok, that was 1 and 2
11:59 shk3 switching between spectra is in the outline
11:59 egonw also, I noted that for my submission, the compound I queried against NMRShiftDB which indicated no exact matches yet, though the GUI actually shows there was one from a mainz data set
11:59 shk3 should use children, but didn't get children working
11:59 egonw shk3: ah... good one, did not try that... sounds like a good approach
12:00 shk3 deleting peaks in smr is not possible
12:00 shk3 you cannot edit the spectrum in smr editor
12:00 egonw no, I noted...
12:00 egonw which is a bit nasty... I know the BC1 did allow that...
12:01 shk3 via the children
12:01 egonw yes
12:01 egonw and you did not get that to work...
12:01 egonw understood
12:02 shk3 exactly
12:05 maclean jonalv : I have an idea why my fix for scroll-wheel zooming in JCP is not working for scroll-wheel zooming in Jmol....
12:06 jonalv maclean: oh :)
12:06 jonalv hehe
12:06 maclean sorry about that
12:06 maclean I guess it /might/ have worked, somehow. Perhaps I should read bug reports more closely....
12:07 egonw OK, maclean, my patch is not good
12:07 jonalv maclean: but Now I can zoom in jchempaint at least
12:07 jonalv cool :)
12:07 jonalv maclean: Now I want a hand to move the spot I am zooming on... :)
12:07 maclean egonw: oh?
12:07 maclean jonalv : possible
12:08 egonw maclean, Gpox:  when I save now, it synchs... but the hub seems empty now...
12:08 CIA-55 bioclipse: egonw * r9057 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.jchempaint/src/net/bioclipse/cdk/j​chempaint/editor/JChemPaintEditor.java:
12:08 CIA-55 bioclipse: Revert "Fixed saving of files created from an empty canvas: synchronize the CDKMolecule model with the ControllerHub model"
12:08 CIA-55 bioclipse: This reverts commit cc7d7343e8a5e40e375de497105f5e1745f15a36.
12:08 CIA-55 bioclipse: It breaks other saves.
12:08 jonalv maclean++ //I can haz zoom
12:08 egonw maclean: I'll assign it to you...
12:08 maclean egonw: okay, cool
12:08 maclean jonalv : it doesn't work very well on windows, though :(
12:09 jonalv oh :(
12:09 maclean it scrolls and zooms simultaneously
12:09 jonalv oh *lol*
12:09 maclean and doesn't update when zooming, so you have to mouseover to force it
12:10 jonalv something to woek on...
12:10 jonalv |work
12:10 maclean yes, I think there might be ways round it
12:10 jonalv yea probably
12:14 shk3 egonw: did you use the "download nmr spectra for this molecule" for quering nmrshiftdb?
12:14 egonw yes
12:14 egonw "Identity"
12:14 shk3 will check#
12:15 egonw it has likely to do with matching aromatic bonds...
12:15 egonw btw... now that I submitted that one...
12:15 egonw it does no longer complain about not finding something...
12:16 egonw however...
12:16 egonw it does nothing either...
12:16 egonw that is, nothing like showing up the actually found spectrum...
12:16 egonw ah...
12:16 egonw they end up in /Virtual
12:16 egonw maybe you can have them open automatically...
12:17 egonw ui.open("/Virtual/NMRShiftDB Results ##/nmrshiftdbXXXXXX.smr
12:17 egonw ")
12:18 shk3 yes, that was the way to hold results
12:18 egonw yes, that's fine with me
12:18 egonw nothing happened...
12:18 shk3 perhaps doing a new virtual folder named search results would be more obvious
12:18 egonw or just automatically open it...
12:18 shk3 yes, that's probably the best
12:18 shk3 will do that
12:20 egonw oh, and please filter out .project in the AddSpectrum page of the NewSpecMol wizard
12:22 egonw it's always a bit getting used to the tweaks, but the second submission went flawlessly
12:24 egonw shk3: btw, I did not have the Outline open...
12:24 egonw may I suggest a Speclipse perspective?
12:25 shk3 well, if that's usefull
12:25 shk3 isn't the outline part of the chemoinformatics perspective?
12:25 shk3 I didn't see an advantage in making one
12:25 egonw no sure about it being in the cheminfo perspective...
12:25 egonw but it's rather crucial to the assigning
12:26 egonw I'd see it must always be open when doing assignments...
12:26 egonw unlike with cheminfo
12:27 shk3 will do one
12:29 egonw maclean: please see @bug 597
12:29 maclean ok
12:30 egonw it's a rather critical bug to beta3
12:30 egonw so I hope we can solve it this afternoon
12:31 maclean ok
12:32 maclean oh how I wish my bioclipse started up normally like it did last week
12:36 egonw so do I
12:36 egonw what's wrong?
12:36 egonw since when does it not boot anymore?
12:36 egonw did you check you have jsexecution
12:36 egonw did you reset target platform and sorts?
12:41 egonw shk3: the good thing of the perspective would be that you can also make New Spectrum, New Assigned Spectrum, show up in the right click in the navigator
12:42 egonw shk3: so that people do not have to go via New -> Other
13:10 maclean egonw: you do see the away status, right? :)
13:10 egonw no
13:10 maclean ohhhh
13:10 maclean sorry
13:11 maclean didn't realise
13:11 egonw this IRC client is not so good
13:11 shk3 can you make that dependant on perspective?
13:11 egonw yes
13:11 egonw see the cheminfo perspective
13:12 egonw plugin.xml
13:13 CIA-55 bioclipse: gilleain * r9058 /bioclipse2/trunk/plugins/net.bioclipse.jmol/s​rc/net/bioclipse/jmol/editors/JmolEditor.java: Scroll-wheel zooming. Sure wish I didn't have to go through the scripting interface for this...
13:15 maclean dammit, just realised something
13:17 jonalv aaaahhhhh
13:17 jonalv why people scan articles_
13:18 jonalv I reading scanned article as pdf but made with latex
13:18 jonalv impossible to see what it says
13:18 jonalv why people do this?
13:18 jonalv I can has future soon?
13:21 jonalv zarah: ping
13:21 zarah jonalv: pong, dear
13:22 maclean Online articles that are scans of just the first page are especially irritating.
13:22 jonalv oh off course it can be worse... :/
13:24 maclean Articles in dead languages? In code?
13:24 egonw maybe in Perl6 :)
13:24 jonalv yea...
13:24 maclean Articles translated into the medium of interpretive dance.
13:24 egonw DNA?
13:25 egonw "Purchase PDB (3185 K)"
13:25 egonw kB, right??
13:25 egonw Kelvin?
13:26 egonw "Purchase PDF (3185 K)"
13:26 egonw buy this PDF and burn in hell?
13:26 egonw at least not, 3.185 M, right?
13:26 egonw bit expensive I'd say
13:26 jonalv yea
13:27 egonw I mean, WTF??
13:27 egonw a scientific journal not giving units??
13:27 jonalv yea...
13:27 egonw jonalv: indeed, it can be worse than plain PDF
13:27 jonalv well but itäs not awesome if the pdf is a low resolution bitmap...
13:28 egonw intentionally, I'm sure...
13:28 jonalv you mena they donät wanna be read?
13:28 egonw otherwise you could copy/paste a some data, a citation, a sentence you want to quite, ... from it
13:28 egonw quite -> quote
13:29 egonw sure
13:29 jonalv uhm and that would be bad?
13:29 egonw sure: copyright infringment is illegal
13:30 egonw as downloading in general
13:30 jonalv quoting is not copyright infingement
13:32 egonw jonalv: they'll see you in court
13:32 egonw (unless someone goes to the media with the thing, and they'll back down)
13:32 maclean egonw: I can't reproduce bug 597 (the showstopper), but probably due to changes I made while half-way through implementing undo/redo
13:32 maclean do you want me to commit this?
13:32 jonalv egonw: are yo utalking about some special case?
13:33 egonw maclean: when they are finished?
13:33 egonw maclean: ETA?
13:33 maclean whut?
13:33 maclean I haven't finished undo/redo
13:33 egonw jonalv: don't you remember this poor sool who copied then data points from a Wiley paper?
13:33 egonw maclean: you don't watch US series?
13:33 maclean but if I commit the partial implementation
13:34 jonalv egonw: nope, no idea...
13:34 maclean egonw: I know what ETA means
13:34 egonw ah, that dunno
13:34 egonw jonalv: never mind
13:34 egonw jonalv: nevermind
13:35 * jonalv neverminds
13:35 * maclean will commit, then ask forgiveness as permission doesn't seem to be forthcoming...
13:36 olass joined #bioclipse
13:36 olass moin
13:36 zarah oh hai olass
13:36 olass beta3?
13:36 olass anyone: any commites almost ready?
13:36 olass commits
13:37 maclean getting there
13:37 egonw from my point of view, there is one blockers...
13:37 egonw but maclean is about to commit a fix for that
13:37 olass ok
13:37 olass will wait then
13:37 olass Gpox? jonalv? masak?
13:38 olass masak: how is exception to console getting along?
13:38 olass that is important
13:38 * jonalv is trying to work on his litterature project fro that course.
13:38 jonalv he is not very succesfull so far...
13:38 olass ok
13:38 egonw masak does not seem around
13:38 jonalv olass: masak is not even around mate
13:38 egonw @seen masak
13:38 zarah masak was last seen 3 h 18 m 18 s ago in #november-wiki saying 'there, pinged.'
13:38 CIA-55 bioclipse: gilleain * r9059 /bioclipse2/trunk/plugins/ne​t.bioclipse.cdk.jchempaint/ (10 files in 5 dirs): Partial implementation of undo/redo that fixes critical bug #597.
13:38 zarah bug #597 | http://tinyurl.com/b8rd5s
13:39 egonw maclean: updating and testing...
13:39 maclean ok
13:40 maclean seems to save ok
13:40 maclean save as also
13:40 maclean but I may not be testing what you are testing :)
13:42 egonw first results...
13:42 egonw compile errors..
13:42 egonw exploring...
13:43 egonw access restriction for AddAtomsAndBondsEdit
13:43 egonw maclean: did you upload everything?
13:43 maclean think so
13:43 maclean hmmm
13:44 egonw I have a SWTAddAtomsAndBondsEdit, but not a  AddAtomsAndBondsEdit
13:44 egonw it seems
13:44 maclean no! exporting from the o.o.cdk.control package
13:44 CIA-55 bioclipse: gilleain * r9060 /bioclipse2/trunk/cdk-externals/trunk/org.op​enscience.cdk.control/META-INF/MANIFEST.MF: Exporting package.
13:44 egonw ack, thanx
13:45 egonw ok, that helped
13:45 maclean good
13:47 egonw no, still broken
13:47 maclean oh?
13:47 maclean I can save changes, and reopen them
13:47 maclean using /Virtual
13:48 egonw will try that
13:48 maclean will try not that
13:48 maclean (as in, I will try not using virtual)
13:49 egonw :)
13:49 egonw tried /Virtual
13:49 egonw but that does not work for me either
13:49 grandm44 joined #bioclipse
13:50 maclean weird. it works everywhere for me
13:51 maclean I do New->Cml Molecule
13:51 maclean add a phenyl, save
13:51 maclean and then open in xml-lite, and it has extra atoms
13:51 egonw mmm :(
13:51 egonw I surely hope it is not platform dependent...
13:51 * maclean updates to see if that breaks things
13:52 maclean I hope that it isn't, too.
13:52 egonw others, please try too...
13:52 maclean although if it works on the mac, I guess that's what's important - there aren't any other OSes, are there?
13:56 egonw mac, what's that?
13:58 maclean sigh. eclipse still crashing for me on trying to open target.platform in the target editor D:
13:58 egonw oh, brilliant
13:59 maclean this is why I can't run bioclipse as normal from source
14:00 maclean oh, no wait, now I can
14:00 maclean aha! I hadn't updated - that's why it worked for me
14:00 maclean sorry about that, will now try and fix it properly
14:01 egonw ok, great
14:01 * jonalv thinks he whished CDK used RuntimeExceptions instead of Exceptions
14:01 egonw olass: you are trying already to export?
14:01 egonw jonalv: feel free to propose on the mailing list...
14:01 egonw CDK 2.0 will be so cool
14:02 jonalv egonw: you actuallly agree with me for once?
14:02 * jonalv isnä
14:02 egonw I don't always disagree with you! :)
14:02 * maclean is away (meeting) :(
14:02 * jonalv isn't even sure he agrees with himnself in this matter
14:02 egonw maclean: how long?
14:02 maclean er 45 mins?
14:02 egonw olass: I guess that will delay the beta until tomorrow?
14:03 egonw maclean: ack
14:17 olass nmo worries
14:18 CIA-55 bioclipse: egonw * r9061 /bioclipse2/trunk/plugins/ (3 files in 3 dirs): Made the human-readable string human-readable
14:18 olass just talked to Gpox, the new MolTable with be online tomorrow ot friday
14:18 olass and beta3 will include it
14:19 * olass is off for a meeting
14:19 * jonalv joins that meeting
15:02 * maclean is back
15:13 egonw maclean: the beta3 will not be released today
15:13 maclean I saw
15:13 egonw so, the pressure is off
15:13 maclean the pressure is always on!
15:13 egonw no, that's just the job description
15:13 maclean have you seen the crunch? we can all be tourists to the crunch! :)
15:14 maclean (mighty boosh quote, sorry)
15:14 egonw crunch?
15:14 egonw I have seen photos of the plain crash
15:14 maclean egonw : I don't quite get the hydrogen positioning rendering bug - the pic looks fine
15:15 egonw the left two carbons are not good...
15:15 egonw hydrogens misaligned
15:15 egonw let me check I have the correct image attached...
15:16 maclean no, I guess I see
15:16 egonw @pz hydrogens
15:16 zarah [598] Orientation of implicit hydroge... http://tinyurl.com/b396uy
15:16 zarah Entire list at http://tinyurl.com/dcebmk
15:16 maclean you wrote that they should be H3C-
15:16 egonw indeed
15:16 egonw because the one bond points to the right...
15:16 egonw the hydrogens go left
15:17 egonw actually.. if you move the carbon a bit more left...
15:17 egonw the right bound of the full label becomes left of the bond
15:17 maclean this seems like the generator's fault
15:17 egonw and then it is correctly aligned again
15:17 egonw which generator is doing this?
15:17 egonw then I can have a look at the code
15:18 maclean well, it's the BasicAtomGenerator
15:18 maclean I think
15:18 maclean but with text stuff, you have to do some of it at render-time, as you don't have the width of the text
15:19 egonw right, and that's when it goes wrong, I think...
15:19 maclean the generator uses the GeometryTools.positionForLabel method
15:19 egonw but not entirely sure...
15:19 egonw GeometryTools.getBestAlignmentForLabel
15:19 maclean that's the fellow
15:20 egonw OK, I think I see the problem
15:20 maclean oh?
15:21 masak joined #bioclipse
15:21 egonw let me try
15:21 maclean ok
15:22 egonw it's a bug in the CDK :)
15:22 maclean ahhhh
15:23 egonw commit pending...
15:27 CIA-55 bioclipse: egonw * r9062 /bioclipse2/trunk/cdk-externals/trunk/or​g.openscience.cdk.standard/src/org/opens​cience/cdk/geometry/GeometryTools.java: Fixed deciding on alignment, which should be double based, as the Point2d is too (fixes #598)
15:27 zarah bug #598 | http://tinyurl.com/b396uy
15:28 maclean egonw++
15:28 maclean that was fastly fixed
15:36 egonw the int based code assumed pixels were still used as coordinate space :(
15:37 egonw which made it break down on a angstrom scale
15:41 maclean ahh, makes sense
15:42 * egonw wonders if anyone has every used the RDF functioanlity in the CDK...
15:42 egonw shk3: you have, not?
15:42 egonw for the NMR prediction paper...
15:42 egonw how did you get the proton descriptor to work at all?
15:42 egonw shk3: I could not find a filter to decide which protons belong to which of the 5 groups
15:43 egonw and the RDFCalculator I wrote is buggy
15:43 egonw (which I am fixing now)
15:44 shk3 hm
15:44 shk3 only some RDF values were effectivly used
15:44 shk3 and in general, some error in some descriptor won't hurt too much
15:45 egonw as long as the general thing is OK
15:45 egonw the default resolutions of the proton descriptors is too low
15:46 egonw and the default peak too wide, I think
15:57 CIA-55 bioclipse: gilleain * r9063 /bioclipse2/trunk/plugins/net.bioclipse.cdk.​jchempaint/src/net/bioclipse/cdk/jchempaint/ (4 files in 3 dirs): First undoable operation in jcp-bioclipse! (the add-atoms-and-bonds edit).
15:58 egonw gilleain++
15:58 egonw gilleain: now I am going to have to buy you a beer
15:58 egonw make sure not to have prearrangements with friends during the CDK workshop!
16:05 masak so it's both scriptable and undoable?
16:08 olass joined #bioclipse
16:09 olass hi
16:09 zarah privet olass
16:10 egonw I wonder if a scriptable command can be undone with the current design
16:10 egonw hi olass
16:10 olass hi egonw
16:11 masak egonw: I find that question highly relevant.
16:11 olass me too
16:14 egonw joined #bioclipse
16:15 egonw oops
16:27 * masak is unhappy we're catching Throwable in JsThread
16:27 masak not Exception. Throwable!
16:28 egonw how so?
16:28 egonw there is plenty of RuntimeExceptions etc thrown around
16:29 egonw which actually are Errors, not?
16:29 masak I'm not sure those should be caught...
16:29 masak they are often quite serious.
16:36 jonalv masak: you think it's better to just have the JS-thread dieing?
16:36 masak jonalv: I think it depends.
16:37 jonalv okey do you have an example of an exception that you really reallly should not catch?
16:37 masak out of memory.
16:37 jonalv masak: that one doesn't really matter
16:37 masak yes, it does.
16:37 jonalv masak: when you are out of memory anything can happend no matter whether you catch it or not
16:39 masak jonalv: can you name one Throwable that we know that we want to catch?
16:40 jonalv masak: honestly I don't think I can name any Throwable that is not also an exception...
16:40 jonalv masak: can you?
16:40 jonalv IOError?
16:41 masak jonalv: AssertionError
16:41 masak jonalv: LinkageError
16:41 masak jonalv: VirtualMachineError
16:41 jonalv masak: I think we want to catch assertionerrors...
16:41 masak jonalv: ThreadDeath
16:41 masak jonalv: I don't.
16:41 jonalv masak: okey, why not?
16:41 masak because they're meant to kill the program.
16:42 jonalv masak: uhm, they are?
16:42 masak yes.
16:42 masak that's why you put assertions in non-production code.
16:42 jonalv masak: you want Bioclipse to die on assertion exception in a manager?
16:42 jonalv that sounds very dangerous...
16:43 maclean joined #bioclipse
16:44 masak jonalv: in non-production code, yes.
16:47 CIA-55 bioclipse: shk3 * r9064 /bioclipse2/trunk/plugins/net.​bioclipse.specmol/plugin.xml: better wording
16:48 egonw any R wizard around?
16:49 egonw how can I get greek letters or other Latex in R plots?
16:50 CIA-55 bioclipse: shk3 * r9065 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (2 files in 2 dirs): bug #595 (double file extension) solved
16:50 zarah bug #595 | http://tinyurl.com/aaku5h
16:51 egonw going for dinner... bbl
17:02 CIA-55 bioclipse: shk3 * r9066 /bioclipse2/trunk/plugins/ (3 files in 3 dirs): Tried to make the wording clearer, no unnecessary "new" etc.
17:46 CIA-55 bioclipse: shk3 * r9067 /bioclipse2/trunk/plugins/net.bioclipse.specmol/ (4 files in 2 dirs): using the better new file page, better wording
18:20 rojasm joined #bioclipse
18:39 edrin left #bioclipse
18:44 CIA-55 bioclipse: shk3 * r9068 /bioclipse2/trunk/plugins/net.biocl​ipse.specmol/src/net/bioclipse/spec​mol/wizards/NewSpecMolWizard.java: added convention

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