Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-01-22

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

All times shown according to UTC.

Time Nick Message
00:54 ChanServ joined #bioclipse
00:54 stain joined #bioclipse
00:54 CIA-31 joined #bioclipse
00:54 ilogger2 joined #bioclipse
03:57 CIA-30 joined #bioclipse
06:54 olass joined #bioclipse
07:35 olass @tell egonw: I have several bugs assigned to write unit tests for CDKManager. Is this OK or would you like to write them?
07:35 zarah Consider it noted.
07:43 Gpox joined #bioclipse
08:07 egonw joined #bioclipse
08:14 egonw moin
08:14 zarah hello egonw, you fantastic person you
08:14 zarah egonw: You have new messages. Write '/msg zarah @messages' to read them.
08:14 egonw @msg
08:14 zarah egonw: 39 m 53 s ago, olass said I have several bugs assigned to write unit tests for CDKManager. Is this OK or would you like to write them?
08:15 olass up, I did
08:15 olass some are simple, like has2D
08:15 egonw I can write unit tests for those methods of which I understand what they have to do...
08:15 olass they are all straightforward
08:15 egonw but in general, the one who wrote the method should write the unit test, I think...
08:15 olass has2D, has3D
08:15 olass I agree
08:16 egonw I can have a look at it...
08:16 olass great
08:16 olass will assign to y9ou
08:16 egonw olass: btw, we really need to start building and testing automatically...
08:16 olass if you want, assign back
08:16 egonw after the beta release
08:16 egonw we can't keep postponing this
08:16 olass egonw: maybe after 2.0
08:16 olass Gpox has it working he says
08:16 olass at least partially
08:16 olass so yes, we should
08:17 egonw no, we *mustÄ
08:17 egonw no, we *must*
08:17 olass :-)
08:17 olass ok ok
08:17 egonw seriously...
08:17 olass next week, you and Gpox try to bring his stuff to pele?
08:17 egonw if we know we can trust to get kicked in the ass when we mess up...
08:17 egonw by this testing services
08:17 egonw we can both work faster...
08:17 olass yes
08:17 egonw because we do not have to run all plugin tests for each minor change
08:17 olass but it can't postpone 2.0
08:18 olass yes
08:18 olass agreed
08:18 egonw not doing it will postpone 2.0
08:18 egonw because the project maintaince becomes to laborous
08:18 egonw really
08:18 egonw I can't keep up with what everyone is doing in the cheminfo namespace
08:18 olass I understand
08:19 olass let's look at it next week then
08:19 olass this week is all about beta release
08:19 egonw I'll be in the UK next week
08:19 olass oh
08:19 egonw so the week after that
08:19 olass are you at BMC today?
08:19 egonw will have a talk in Cambridge
08:19 egonw yes
08:19 olass talk to Gpox about this
08:20 olass and set something up on pele
08:20 olass Gpox has it, at least, partially working
08:20 olass on his machine
08:20 olass Gpox: what is the status of the auto building?
08:21 Gpox I got a zip that i could run on my mac iirc
08:22 olass egonw: since you are gone next week, maybe you can interact with Gpox today at bmc?
08:22 olass brb
08:23 egonw Gpox: can we meet in about 10 minutes?
08:23 Gpox yes
08:23 egonw so that you can show me how to do it...
08:23 egonw I'll look up the wiki page now and get started
08:23 egonw it only builds things... right?
08:23 egonw or can we run the unit tests too?
08:24 Gpox only build
08:24 egonw ok, that's a good start
08:33 olass joined #bioclipse
08:33 olass oi
08:33 olass @ping
08:33 zarah olass: pong, honey
08:34 olass @botsnack
08:34 zarah ;)
08:44 egonw Gpox: can you please upload your config file to SVN?
08:44 CIA-30 bioclipse: egonw * r8421 /headless/ (. trunk/): Made room to share config files
08:44 egonw see commit 8421
08:45 Gpox sure, just gona check so it works
08:45 egonw and the next
08:45 egonw 8422
08:45 CIA-30 bioclipse: egonw * r8422 /headless/trunk/pele/: Added folder for the config for Pele
09:07 CIA-30 bioclipse: egonw * r8423 /bioclipse2/trunk/plugins/ (3 files in 3 dirs): Moved test methods to the proper testing class: atom type perception is in CDKDebugManager, not CDKManager
09:11 ChanServ joined #bioclipse
09:11 stain joined #bioclipse
09:12 masak joined #bioclipse
09:17 edrin joined #bioclipse
09:17 egonw edrin: check my blog
09:17 shk3 joined #bioclipse
09:21 edrin hi, I'm doing
09:22 shk3 great to hear you are doing
09:22 masak I'm doing too.
09:23 shk3 you too?
09:23 shk3 even better
09:25 masak shk3: are you doing?
09:25 shk3 a bit
09:26 shk3 do you know about help context ids in Eclipse help system?
09:30 CIA-30 bioclipse: egonw * r8424 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Added missing test methods for has2d() and has3d() (fixes #86)
09:30 zarah bug #086 | http://tinyurl.com/62zqah
09:32 CIA-30 bioclipse: egonw * r8425 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Added missing test method for fromCml (closes #97)
09:32 olass shk3: I am working on the welcome page
09:32 olass will set it up better now
09:32 shk3 "contexts.xml eclipse help" in google gives bioclipse wiki as first hit
09:32 shk3 wow
09:32 olass cool
09:33 olass but not so good after all
09:33 shk3 perhaps you get the dynmic conent in welcome.xml working
09:33 shk3 would be great
09:33 olass yes
09:33 olass will set that up
09:33 olass so each feature can add to welcome page
09:33 olass that is the objective
09:35 egonw ah, that's cool!
09:37 shk3 it seems my problem was I did not properly identy the anchor element in the welcome page
09:37 shk3 in the plugin which should contribute
09:39 shk3 btw, my idea was to have a "what can i do with Bioclipse section"
09:39 shk3 and plugins contribute things like "Analyze spectra" pointing to cheat sheets and/or help texts
09:42 edrin egonw: cool entry :)
09:42 egonw thanx
09:42 edrin :)
09:42 egonw will be giving a presentation at Unilever Center Cambridge next week
09:42 egonw will show results there too
09:42 edrin very cool :)
09:45 CIA-30 bioclipse: goglepox * r8426 /headless/ (build.properties build.xml): My configs files don't forget to edit build.properties to match your environment
09:46 CIA-30 bioclipse: shk3 * r8427 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (3 files in 2 dirs): the cheat sheet now has a link to help texts
09:51 HCS_ joined #bioclipse
09:51 egonw hi chris
09:51 HCS_ Hi Egon
09:51 olass hi chris
09:52 HCS_ Cheers
09:52 * HCS_ is playing with xchat on Ubuntu
09:52 egonw HCS_: quassel is nice
09:52 HCS_ ack
09:53 edrin ah
09:53 HCS_ its hard find the right one. Too much choice
09:53 edrin steinbeck
09:53 HCS_ yes
09:58 H-C-S joined #bioclipse
09:59 HCS_ left #bioclipse
10:03 H-C-S Can someone hear me?
10:04 egonw yes
10:06 olass H-C-S: try @ping
10:06 H-C-S cool
10:06 olass @ping
10:06 zarah olass: pong
10:06 H-C-S hehe
10:06 olass zarah can hear me
10:06 olass or just say hi
10:06 olass hi
10:06 zarah ni hao olass
10:06 olass hi
10:06 zarah hi olass
10:06 H-C-S wow, chinese bot
10:07 olass masak++
10:07 masak :P
10:07 masak that's all the chinese she knows.
10:07 masak zarah: ni hui shuo zhongwen ma?
10:07 zarah masak: excuse me?
10:08 H-C-S left #bioclipse
10:10 egonw guys: random wicked idea:
10:10 egonw should we annotate IManager methods with @Gist("48101") or whatever Int is appropriate?
10:14 CIA-30 bioclipse: egonw * r8428 /headless/ (5 files in 2 dirs): Move Arvid's config file in a subdir inside trunk
10:20 CIA-30 bioclipse: egonw * r8429 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Added missing unit test for getInfo
10:21 olass egonw: let us discuss that post bioclispe 2.0
10:21 CIA-30 bioclipse: egonw * r8430 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business.test/src/net/bioclipse/​cdk/business/test/CDKManagerTest.java: Use assertNull instead of assertTrue(==null)
10:21 olass too busy now
10:22 olass we must prioritize
10:22 olass file a bug for 2.1
10:25 CIA-30 bioclipse: ospjuth * r8431 /bioclipse2/trunk/plugins/net.bio​clipse.gist/META-INF/MANIFEST.MF: Removed dependency to cdk.business from gist plugin.
10:25 shk3 btw, wiki pages http://wiki.bioclipse.net/index.php​?title=How_to_add_help_to_a_plugin
10:25 zarah shk3's link is also http://tinyurl.com/chsmnz
10:25 shk3 and http://wiki.bioclipse.net/index​.php?title=Writing_Context_Help
10:25 zarah shk3's link is also http://tinyurl.com/aokoph
10:25 shk3 are redundant
10:25 olass shk3: feel free to fix
10:25 shk3 partly, at least
10:41 steinbeck joined #bioclipse
10:45 egonw ready for next week?
10:48 CIA-30 bioclipse: shk3 * r8432 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (4 files in 4 dirs): working on the cheat sheet
10:49 Gpox eecliqnano -wflskf
10:50 steinbeck egonw: yes :-)
10:50 steinbeck very exciting
10:51 steinbeck Is Gpox having a problem?
10:52 CIA-30 bioclipse: egonw * r8433 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business.test/src/net/bioclipse/​cdk/business/test/CDKManagerTest.java: Fixed unit tests: the returned object has the 2D coords
10:52 masak edrin: re commit messages, what does 'xx' mean?
10:53 edrin i think move a to b
10:54 edrin instant commit message
10:54 masak I see.
11:16 CIA-30 bioclipse: shk3 * r8434 /bioclipse2/trunk/plugins/​net.bioclipse.nmrshiftdb/ (2 files in 2 dirs): re-added an action for new molecule
11:19 CIA-30 bioclipse: egonw * r8435 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Added missing unit test for loadMolecules (fixes #101)
11:19 zarah bug #101 | http://tinyurl.com/5s3c2p
11:22 CIA-30 bioclipse: shk3 * r8436 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (2 files in 2 dirs): ready so far, saveAs still missing
11:27 CIA-30 bioclipse: shk3 * r8437 /bioclipse2/trunk/plugins/net.biocl​ipse.spectrum/src/net/bioclipse/spe​ctrum/editor/SpectrumEditor.java: formating
11:37 olass Gpox: I know you are very busy, but could you comment on bug #223 depending on #272?
11:38 olass I would so much like to have save and save as in JCP...
11:39 olass the CDKManager should be ready for this, right egonw?
11:39 olass I mean, there are the saveMolecule actions etc
11:39 olass Gpox: is it the dirty state that is problematic?
11:40 egonw there are 10M saveMolecule methods
11:40 egonw none of them tested
11:41 egonw people have been adding them without unit tests :(
11:41 olass yes, I fear I am to blame for some
11:41 egonw olass: yes, dirty state is a dep for save/saveAs
11:41 egonw we're working on it
11:41 CIA-30 bioclipse: egonw * r8438 /bioclipse2/trunk/plugins/net​.bioclipse.chemoinformatics/ (2 files in 2 dirs): Don't encourage the MDL molfile
11:41 olass ok, good
11:42 olass egonw: think can get that in beta?
11:42 CIA-30 bioclipse: egonw * r8439 /bioclipse2/trunk/plugins/net.bioclipse.chemoinfor​matics/src/net/bioclipse/chemoinformatics/wizards/ (NewCMLFileWizard.java NewMDLMolfileWizard.java): Make the new file contain ethane, to work around a known bug in the JChemPaintEditor: does not handle zero and one-atom files too well
11:42 egonw we have 28 hours left
11:42 olass :-)
11:42 egonw aka... perhaps
11:42 egonw we're trying
11:42 shk3 did anybody implement saveAs for an editor?
11:42 * olass is setting up a new welcome page with dynamic content
11:42 olass shk3: look in bc1
11:42 shk3 I am trying for spectrumeditor, but doesn't work
11:43 egonw Gpox: commit 8439 has the NewCMLWizard create ethane to encourage opening in JCP
11:43 olass JCP had that in bc1
11:43 shk3 well, it looks like you need isSaveAsAllowed and doSaveAs in editor
11:43 shk3 but that doesn't have an effect for me
11:54 stefan__ joined #bioclipse
11:55 stefan__ hello together
11:55 stefan__ >I just installed bioClipse and do some tryouts. Everything work fine so far - excepting the js - console. If i try to run extractsmiles.js i get the error that the function getSmiles can not be found. Has somebody an idea about that.
11:56 egonw stefan__: which bioclipse version?
11:56 egonw from SVN?
11:56 egonw getSMILES
11:56 stefan__ 2.0
11:56 egonw that was turned upper case
11:56 egonw where did you find that extractsmiles.js?
11:56 egonw will fix it...
11:56 stefan__ Version: 2.0.0.A20080808
11:57 egonw that .js came with the release?
11:57 Gpox yeah but withe the new
11:57 egonw stefan__: get getSMILES bug is already fixed in SVN, so will be in the tomorrow beta release
11:58 stefan__ ah... cool. thanks.
11:58 egonw Gpox: ?
11:58 Gpox wrong window
11:58 egonw stefan__: for now... just convert your js to say 'getSMILES' and then test it again
11:59 egonw stefan__: more example scripts are linked here:
11:59 egonw http://delicious.com/tag/bioclipse+gist
11:59 zarah egonw's link is also http://tinyurl.com/5gqjet
11:59 egonw but the newer ones will need tomorrows release
12:01 stefan__ thanks. i just renamed to upper - case. it works now :-)
12:01 egonw good :)
12:01 egonw the console support tab completion
12:01 egonw try: cdk. and then hit TAB
12:01 egonw try: "cdk." and then hit TAB
12:02 egonw for each method there is inline help
12:02 egonw man cdk.fromSMILES
12:02 egonw I think...
12:02 egonw masak: right?
12:02 masak I'm about to go to lunch...
12:02 masak egonw: but yes.
12:02 masak right.
12:02 stefan__ ahhh.... i tried Ctrl + Space for code competion
12:02 egonw :)
12:02 egonw stefan__: yes, that makes sense :)
12:03 egonw masak: would that be possible?
12:03 masak egonw: file ticket plz
12:03 egonw masak: will do
12:05 egonw @pz 276
12:05 zarah http://tinyurl.com/cmdod4
12:39 CIA-30 bioclipse: egonw * r8440 /bioclipse2/trunk/features/​net.bioclipse.test_feature/ (. .project build.properties feature.xml): Added a draft Test feature; at least needed for auto-testing
12:59 jonalv joined #bioclipse
13:16 jonalv oh hai
13:16 zarah hi jonalv
13:17 jonalv zarah: talk to me
13:17 zarah jonalv: excuse me?
13:17 jonalv haha
13:17 jonalv zarah: tell me
13:17 zarah jonalv: You have no new messages.
13:27 jonalv olass: I am confused. IDatabase and IDatabaseChild are classes...
13:27 masak *lol*
13:28 olass jonalv: what's confusing with that?
13:28 masak the only thing worse than Hungarian notation is Hungarian notation that lies about its identifier.
13:28 masak olass: you're actually asking that?
13:28 masak olass: what does the "I" stand for?
13:29 olass IntegratedClass
13:29 masak hahaha
13:29 olass :-)
13:29 olass ok ok, rename it yo
13:29 olass it was an EXAMPLE
13:29 olass proof of concept
13:29 jonalv "det just ett skönt exempel..."
13:29 masak olass: examples are supposed to be correct. :P
13:29 olass and I might have written interfaces at some point
13:30 olass and generated the classes from plugin.xml
13:30 jonalv anyway it got me a bit confused...
13:30 masak rightly so.
13:30 masak olass: after this, we might revoke your license to use the letter I.
13:30 olass jonalv and masak: you are so easily confused
13:31 olass try to think outside the box
13:31 jonalv :)
13:31 masak pfft.
13:31 * olass usually thinks below the box
13:32 * masak usually thinks inside of the cranium
13:32 * olass has a very big box
13:33 egonw yeah, houses in stockholm are big carbon boxes
13:33 egonw cardboard coxes
13:33 egonw forget it
13:33 egonw blah
13:33 * masak wants a carbon box
13:33 masak just full of C12 goodness.
13:33 jonalv masak: move to Stockholm then
13:34 masak jonalv: I don't want it _that_ much. :P
13:46 CIA-30 bioclipse: shk3 * r8441 /bioclipse2/trunk/plugins/​net.bioclipse.nmrshiftdb/ (html/search.html plugin.xml): minor changes
13:47 CIA-30 bioclipse: shk3 * r8442 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (3 files in 3 dirs): finished the spectrum cheat sheet
13:52 * jonalv is trying to build for the new database view but it's hard to do while still maintaining the old functionlaity... :(
13:57 shk3 olass: any success with the welcome screen?
13:57 CIA-30 bioclipse: goglepox * r8443 /bioclipse2/trunk/plugins/net.biocl​ipse.cdk.business/src/net/bioclipse​/cdk/domain/CDKAdapterFactory.java: Removed use of BioclipseStore
13:58 olass shk2: working hard on it
13:58 CIA-30 bioclipse: goglepox * r8444 /bioclipse2/trunk/plugins/net.bio​clipse.cdk.jchempaint/plugin.xml: Removed add atom from toolbar
13:58 olass really tricky
13:58 shk3 good luck
13:58 CIA-30 bioclipse: goglepox * r8445 /bioclipse2/trunk/plugins/net.bio​clipse.cdk.jchempaint/plugin.xml: changed id for mdl and cml editors and updated definiton expression
13:59 CIA-30 bioclipse: goglepox * r8446 /bioclipse2/trunk/plugins/net.bioclipse.cdk.​jchempaint/src/net/bioclipse/cdk/jchempaint/ (2 files in 2 dirs): Added dirty check and Save / Save As
13:59 egonw Gpox:++
13:59 egonw Gpox++
14:00 jonalv karma :
14:00 zarah : has a karma of 0
14:00 Gpox egonw: is it saveCML /saveMDLMolefile i should use?
14:00 jonalv :++
14:00 jonalv karma :
14:00 zarah : has a karma of 0
14:00 egonw Gpox: no idea
14:00 egonw that one, or any of the other saveMolecule method
14:00 egonw ...
14:00 egonw there are 5 million of them...
14:01 jonalv karma Gpox:
14:01 egonw no idea why we have so many
14:01 zarah Gpox: has a karma of 0
14:01 egonw Gpox: please use saveMolecule(IMolecule, String, boolean)
14:01 egonw the last boolean indicates overwrite or not
14:06 egonw I guess the saveAs dialog will ask about overwrite or not
14:07 CIA-30 bioclipse: goglepox * r8447 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.jchempaint/src/net/bioclipse/cdk/j​chempaint/editor/JChemPaintEditor.java:
14:07 CIA-30 bioclipse: Changed to use saveMolecule(IMolecule, String, boolean)
14:07 CIA-30 bioclipse: (resolves bug #223)
14:07 zarah bug #223 | http://tinyurl.com/d43xx4
14:10 CIA-30 bioclipse: egonw * r8448 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Added five missing saveMolecule() methods
14:16 jonalv (many_save_molecule_methods)++
14:17 jonalv (#save_molecules_methods)++
14:19 jonalv http://www.crosswalk.com/fun​/computers-internet/1350622/
14:19 zarah jonalv's link is also http://tinyurl.com/c7be48
14:19 jonalv niiiice :)
14:23 egonw ok, guys
14:23 egonw that was 5 missing unit test methods for 5 existing saveMOlecule methods
14:24 egonw who is interested for a coffee brake upstairs?
14:25 jonalv mmmm coffe brreak!
14:25 jonalv I am game!
14:26 jonalv now?
14:26 egonw yes
14:26 jonalv oki
14:26 egonw jonalv: will you take along whoever is in your room too?
14:26 jonalv they don't wanna, masak at least. gpox is heading home...
14:27 egonw ok, their loss :)
14:29 masak I'm in the middle of several somethings...
14:29 masak (and I don't drink coffee)
14:56 * olass loves coffee
15:07 edrin listening to orbital halcyon &on & on
15:08 CIA-30 bioclipse: carl_masak * r8449 /bioclipse2/trunk/_playground_/net.b​ioclipse.springBasedPrototypePlugin/ (2 files in 2 dirs): [n.b.springBasedProtoypePlugin] removed obs dep on CONSOLE
15:09 CIA-30 bioclipse: carl_masak * r8450 /bioclipse2/trunk/plugins/net.bioclipse.biomoby/ (2 files in 2 dirs): [n.b.biomoby] removed obs dep on CONSOLE
15:09 CIA-30 bioclipse: carl_masak * r8451 /bioclipse2/trunk/plugins/net.biocl​ipse.biomoby/src/net/bioclipse/biom​oby/business/BiomobyManager.java:
15:09 CIA-30 bioclipse: [BioMobyManager.java] cleanup
15:09 CIA-30 bioclipse: * organize imports
15:09 CIA-30 bioclipse: * remove unnecessary warn supression
15:09 CIA-30 bioclipse: carl_masak * r8452 /bioclipse2/trunk/plugins/​net.bioclipse.biomoby.ui/ (2 files in 2 dirs): [n.b.biomoby,ui] removed obs dep on CONSOLE
15:09 edrin do you know this song?
15:10 CIA-30 bioclipse: carl_masak * r8453 /bioclipse2/trunk/plugins/net.bioclipse​.biomoby.ui/src/net/bioclipse/biomoby/u​i/views/MobyViewContentProvider.java: [MobyViewContentProvider] cleanup
15:10 CIA-30 bioclipse: carl_masak * r8454 /bioclipse2/trunk/plugins/net.bioclip​se.biomoby.ui/src/net/bioclipse/biomo​by/ui/model/MobyViewServiceType.java: [MobyViewServiceType] removed obs dep on CONSOLE
15:11 CIA-30 bioclipse: carl_masak * r8455 /bioclipse2/trunk/plugins/net.bioclipse.moss/ (META-INF/MANIFEST.MF src/net/bioclipse/moss/MossRunner.java): [n.b.moss] removed obs dep on CONSOLE
15:11 * masak deluges #bioclipse with commits
15:44 edrin masak: remove the old js editor of svn trunk when adding the new one to bioclipse trunk?
15:44 masak edrin: uh, hold on... I thought you said yours was in the beginning stages?
15:45 CIA-30 bioclipse: shk3 * r8456 /bioclipse2/trunk/plugins/net.bioclipse.spectr​um/cheatsheets/GeneralSpectrumCheatSheet.xml: this was a c&p artifact
15:45 olass edrin: obviously this is not yet settled
15:45 edrin ok
15:46 olass masak, edrin: please make up your minds on the jseditor
15:46 masak well, I don't think we should be rash about it, at least.
15:46 olass which one should go in beta release?
15:46 olass and why?
15:46 masak that's not easy for two devs, each with their implementation of something, to decide upon.
15:46 egonw olass: updated the status reports of stable managers for CDKManager and CDKDebugManager
15:47 egonw they are releasable beta state
15:47 olass egonw: great
15:47 olass egonw++
15:47 edrin one thing of the old one is that it does not tell you where the script hangs (what line number, thus it will be difficult to debug the script)
15:47 olass masak: you did not write the old jseditor, that was an old SF project
15:47 masak olass: true.
15:47 masak edrin: could you describe how your editor does the line number thing?
15:48 * egonw is heading home
15:48 masak olass: I just wrote the binding to the js env, and even that I did with your help.
15:48 edrin masak: rhino tells it
15:48 masak edrin: ok, that sounds interesting, because it indicates that the "old" editor could do the same.
15:49 masak it uses rhino as well.
15:49 masak but maybe not in the appropriate way, then.
15:49 masak are line numbers the deciding issue between the two editors?
15:49 CIA-30 bioclipse: egonw * r8457 /bioclipse2/trunk/plugins/net.biocl​ipse.cdk.debug.test/src/testFiles/ (aromatic.mol atp.mol polycarpol.mol): Copied missing test files from cdk.busines.test
15:51 edrin masak: well, i have to admit that i think i would have kind of problems to debug a script if you do not get an indicator in what line the error is
15:51 edrin olass: what do you think?
15:51 olass about what?
15:51 masak my point is, judging from your description, it doesn't sound that hard to hack in line numbers in the "old" editor as well.
15:51 masak it runs on rhino too, see.
15:52 olass masak: sounds logical
15:52 olass edrin: I see no point in adding the new jseditor. It brings nothing new apart from line numbers
15:52 olass which is important btw
15:52 masak we'll have a look at that.
15:52 olass good
15:52 olass HOWEVER
15:53 olass I agree with edrin that the old jseditor is abandoned and his is newly written
15:53 masak I'm interested in how the new js editor does the line numbers thing.
15:53 edrin masak: i thought you set a WONTFIX to the line number bug?
15:53 masak olass: me too
15:53 olass edrin: will you work on extending the new jseditor more?
15:53 edrin masak: it's quite easy
15:53 masak edrin: yes, due to my understanding of the problem I put WONTFIX on it.
15:54 masak edrin: can you point to a line of code somewhere?
15:54 edrin olass: most likely, when xws4j project is ok
15:54 olass masak: is there anything in the old JSEditor that you would miss, should we switch to the new jseditor?
15:54 olass I mean, the editor is completely indepoendent of the script execution?
15:54 olass right?
15:55 edrin well, i would like to have tab completion ... but i would rather prefere to interact the js console with the editor as a "script logger, thus the tab completion might be easy
15:55 masak olass: I would miss the old color coding. :)
15:55 olass masak: I'm sure that can be adjusted...
15:57 * masak looks again at edrin's editor
15:58 masak I don't like the way the "Run" toolbar icon is done.
15:59 olass masak: that has nothing to do with the editor
15:59 olass feel free to change that
15:59 masak olass: it's connected to the editor.
15:59 masak and I don't feel free to change that.
15:59 edrin masak: here is the code how i run the script http://bioclipse.svn.sourceforge.net/viewvc/bi​oclipse/bioclipse2/trunk/plugins/net.bioclipse​.rhino/src/net/bioclipse/rhino/actions/ScriptE​xecution.java?revision=8420&view=markup
15:59 olass :-)
15:59 zarah edrin's link is also http://tinyurl.com/derhnq
16:00 edrin rhino reports the errors/exceptions + error lines properly back
16:00 masak edrin: that is _very_ easy to adopt in the other editor.
16:00 olass masak: please do that
16:01 edrin it has the following features: you can run the script in main thread, or as an eclipse job. in case of an eclipse job you can check within the script if the job was canceled by the user clicking the X in the progress view
16:01 masak edrin: I think "Run" should be a one-click action.
16:01 olass me too
16:01 masak more advanced alternatives are ok.
16:01 olass have two actions then
16:01 masak but they should not interfere with the most common action.
16:01 olass right
16:02 olass standard GUI guidelines ;-)
16:02 edrin masak: why dont you like the run button?
16:02 masak edrin: for the above reason.
16:02 olass pops up dialog?
16:02 edrin it is. you can just click it and it runs
16:02 masak no, menu.
16:02 olass masak: I don't understand
16:02 masak edrin: no, it has a pulldown menu with two alternatives.
16:03 olass is Run not a one-click action now?
16:03 edrin sure
16:03 edrin but
16:03 edrin the button can just do more then the one of the old editor
16:03 olass edrin: could we not have 2 buttons?
16:03 edrin if you just click it, it does the same as the old
16:03 masak edrin: I realize that.
16:03 olass like the run button in eclipse?
16:03 olass click fast=run
16:04 olass click and hold = popup?
16:04 edrin if you select context menu it offers just an additional option, (running script in its own thread)
16:04 masak olass: that's not called "popup"
16:04 olass pulldown?
16:04 masak aye.
16:04 edrin olass: but it is exactly as you say
16:04 masak hm.
16:04 olass masak: is that bad?
16:04 masak olass: I don't know...
16:05 olass sounds good to me
16:05 masak my head is spinning. :)
16:05 olass I'm geting quite used to eclipse buttons...
16:05 * olass would like to see masak's head spin 360 deg
16:05 * masak swivels in chair
16:05 olass edrin: this is late in release
16:05 masak olass: I think I'm ok with the current setup.
16:06 olass can we postpone this to next week?
16:06 olass edrin?
16:06 edrin ok
16:06 masak olass: I'm fine with that. until a few minutes ago, I thought that was the plan anyway.
16:06 olass indeed
16:06 olass decided then
16:06 edrin btw. if you are here i have something additional regarding this
16:06 olass yes?
16:06 olass to me or masak?
16:07 edrin regarding scripts and execution of them
16:07 edrin both
16:07 masak let's hear it.
16:07 edrin it is important because i think this was not considered properly yet
16:07 edrin it's how the script commands/ scripts / managers are run
16:08 masak yes?
16:08 edrin meaning run "in the main GUI thread" or in a separate thread to not block the GUI
16:08 masak edrin: why would one want to do the former?
16:08 olass jonalv should hear about this too...
16:08 edrin /similar to the two run options of  the js things
16:09 edrin masak: what is the former for you?
16:09 olass edrin: why run in GUI thread?
16:09 masak edrin: running script in main GUI thread.
16:09 edrin because if any GUI update occurs from the managers they must be run with certain care
16:09 edrin for example
16:09 olass sure
16:10 olass just do DIplsay.getDefault().asyncExec....
16:10 edrin + all managers have to be 100% thread save
16:10 jonalv joined #bioclipse
16:10 olass aah, jonalv
16:10 olass just in time
16:10 olass edrin is talking about threadsafeness
16:10 edrin olass: I am not sure if all devs are aware of this DIplsay.getDefault().asyncExec
16:10 masak I think running things in the GUI thread is a generally bad idea.
16:11 * olass leaves this discussion to jonalv, masak and edrin
16:11 jonalv yea we want a responsive client
16:11 olass edrin: can we not create a convenience method for that?
16:11 edrin olass: sorry, this is a very important thing
16:11 jonalv olass: please join in. You are the visionist when i comes to Bioclipse
16:11 * olass listens
16:11 masak olass: that's a bit hard since it involves anon subclassing...
16:11 jonalv a convenience method for what?
16:12 edrin convenience method?
16:12 masak jonalv: asyncExec
16:12 olass for updating GUI stuff
16:12 jonalv can someone sum up the topic for me?
16:12 jonalv there is such a thing
16:12 olass generally, scripts do not update GUI
16:12 jonalv You send a BioclipseUIJb with the method
16:12 olass GUI should listen to changes in model and update themselves
16:12 jonalv oh the scripts?
16:12 masak jonalv: we're talking about running in main GUI thread vs running in a separate thread.
16:12 olass jonalv: we are not talking managers but JS
16:12 edrin jonalv: sure, no i just wanted to mention again that I pretty assume not every developer is aware of the multithreading in managers
16:13 olass edrin: you are right
16:13 olass all need to learn
16:13 jonalv well scripts are not trun in gui thread
16:13 jonalv we DO NOT want to lock the gui thread
16:13 olass I agree
16:13 edrin jonalv: yes, i agree here
16:13 masak jonalv: in edrin's editor they _can_ be run in the GUI thread.
16:13 edrin not nice to do so...
16:13 edrin yes
16:13 edrin let me explain
16:14 olass please do
16:14 jonalv I don't like that. It's too confusing. I don't want our users to have to figure out how to run a script
16:14 edrin i would have preferred to set the default action to run it in a seprate thread
16:14 edrin but this is highly risky
16:14 jonalv I just want the script to run when you hit the play button
16:14 edrin sure
16:14 edrin i think so, too
16:15 olass edrin: why highly risky?
16:15 olass if the mix with GUI stuff, then the'd better learn the asyncExe() GUI update thingy
16:15 olass what else could go wrong?
16:15 masak *nod*
16:15 edrin because you never know what a manager method means to the IDE or anything object in general
16:16 olass edrin: please explain in more detail this
16:16 masak edrin: that's why manager methods should be thread safe.
16:16 olass I donät get it
16:16 masak .oO( mmm, donuts )
16:16 jonalv There is a major problem here which I don't think has gotten wnough attention
16:16 edrin olass: masak just said it: that's why manager methods should be thread safe.
16:16 masak that's not a problem, it's just a fact.
16:16 edrin jonalv: yes, that what i thought :)
16:16 jonalv We can not use listeners in the classic way when there are multiple threads involved
16:16 edrin masak: exactly it is a fact
16:17 olass ok
16:17 masak jonalv: oh? pray tell.
16:17 olass so how do we do this?
16:17 edrin olass: my fear is that people in general underestimate the necessity for making an application /in this case the manager) thread safe!
16:17 jonalv if the listener updates the gui then it needs to be run in the gui thread for example
16:18 jonalv but I am afraid there might be synchronization neeeded as well
16:18 edrin yes
16:18 jonalv but synchronization shouldn't be _so_ hard to do there.
16:18 jonalv stop me if I am wrong...
16:19 * olass is not stopping jonalv
16:19 edrin writing thread save code is damn difficult that even I have high troubles doing this...
16:19 edrin jonalv: synchronization is not enough
16:19 edrin at least the thing is much more difficult to achieve then you think
16:19 olass edrin: If we agree that running scripts in GUI is not an option, what do you suggest we do?
16:19 masak edrin: I think what you're saying is not reason enough to run JS code in the GUI thread.
16:19 jonalv edrin: I am not sure I agree
16:20 edrin hm
16:20 sneumann joined #bioclipse
16:20 olass edrin: can you give a good example when things break in current bioclipse setup?
16:20 jonalv the managers should not contain state. And should only work with the objects they have been handed. So not even synchronization should be needed...
16:21 olass jonalv: you can call other code than managers, but maybe that is not problematic?
16:21 olass eclipse code for example
16:22 jonalv I know that structuredb is not thread safe right now. But it is working with a database and francly it's a special case and hole I have dug for myself... :)
16:22 edrin olass: i dont think all eclipse API is thread safe
16:22 olass that I have never said
16:22 edrin olass: there is the example
16:22 olass edrin: no, that is no example
16:22 edrin structuredb
16:22 jonalv edrin: the gui parts are since they are only reachable from the gui thread...
16:23 olass edrin: a cncrete example
16:23 jonalv yea but the problem with structuredb is that I didn't think about thread safety when I started it and it's a very special case...
16:23 olass that eclipse code MIGHT not be thread safe is not an example
16:23 olass if that's what you meant
16:24 edrin ehm, the problem i mean is more this: to have a thread save manager means the developer must make it thread save. and i fear the most difficult part is that writing thread save code is that difficult that one must assume that managers will not be threadsafe in general, that's just the thing i wanted to mention
16:25 olass edrin: so what are you suggesting?
16:25 edrin olass: i meant exactly the thing jonalv just said
16:25 edrin olass: I'm not god, I don't know :)
16:26 olass ok, so if we get people to write thread safe managers, are we OK then?
16:26 olass edrin?
16:26 edrin yes, but this is very unlikely
16:26 jonalv edrin: how big a problem do you think that this is and what are you suggesting that we do about it? Since we most definitly don't want to lock up the gui and not even be able to cancel a long running job I don't see any alternatives...
16:26 jonalv having a gui that locks up and refuses to cancel things is not an option I think
16:27 olass I see only alternative to run scripts in another thread and try to keep managers thread safe.
16:27 edrin jonalv: i agree, there is a dilemma  :)
16:27 edrin maybe there is one
16:28 edrin all script code must run in one thread, but this is no warranty, too. because managers might still be used from the GUI
16:29 edrin olass, jonalv: in any case, if a manager has to be thread save this means you must tell this to the developers
16:29 olass edrin: yes
16:29 olass good point
16:29 jonalv yea very true
16:29 edrin and the tutorial for developers must start like this:
16:30 edrin "Managers must be 100% thread save. If you do not know what this means, do not write code. If you know what this means, it is still unlikely that you can achieve this, and do not write code" :/
16:30 olass edrin: I agree. I don't think I understood previously that this is such a big problem... I'm not good at threading issues
16:30 olass hmm
16:31 olass edrin: maybe not that harsh :-)
16:31 edrin yes, in know
16:31 edrin but i even have many problems writing thread safe stuff
16:31 olass edrin: do you want to run your scripts in GUI thread?
16:31 olass yes7no/somtimes?
16:32 edrin i think xws4j is thread safe, because it must be thread save the way I designed it, but i am not 100% sure
16:32 olass [17:37] < olass> edrin: do you want to run your scripts in GUI thread?
16:32 olass edrin?
16:32 edrin olass: no, i would prefer to do stuff not in GUI thread
16:33 edrin but that is just what i want
16:33 jonalv I guess we could have one worker thread which does everything. Would that help?
16:34 edrin this would help, but in this case one could not run scripts in parallel
16:34 masak I don't think that we should provide a "Run" alternative that lets the user run the script in the GUI thread.
16:34 jonalv we wouldn\t be able to walk and chew gum then but we would be able to look good while we are doing one the things...
16:34 edrin (and that's something i wanted to do as soon as there are some services available with xws)
16:34 olass jonalv: sounds good to me
16:35 masak the old editor never could run scripts in parallel, far as I know...
16:35 olass masak: does the editor run the script?
16:35 edrin masak: i thought jonalv tried to do it this way?
16:35 masak olass: I don't understand the question.
16:35 masak olass: the editor tells the js env to run the script.
16:35 olass is the editor so tightly connected to execution of scripts?
16:35 edrin jonalv: didn't you try to change something in a branch to make it multithreaded?
16:36 masak olass: how tightly?
16:36 masak so it can't multitask?
16:36 olass must scripts go via an editor to run?
16:36 masak no.
16:36 olass or can it go directly to JS env?
16:36 jonalv there is only one js thread
16:36 masak there's only one js env.
16:36 olass and is that in the same plugin?
16:36 masak it might be that this design is a mistake.
16:36 olass js env and jseditor?
16:37 masak I've been thinking about several js envs lately.
16:37 masak that could work.
16:37 olass interesting
16:37 masak basically as simple as creating new js threads.
16:37 masak the reason I hesitated to do this earlier was...
16:37 jonalv edrin: can you provide an example (not including a db) which is causing threading issues_
16:38 masak ...that I had this idea about js variables being shared across all of bioclipse.
16:38 masak i.e. _one_ single js env.
16:38 masak but I'm not sure anymore that the payoff is that great.
16:38 jonalv I would like to hear egonws and Gpoxs oppinions on this
16:38 masak so, let a hundred js flowers bloom.
16:38 edrin masak: sharing all stuff is not so good i fear
16:39 edrin my editor creates an environment for each script...
16:39 masak edrin: it wouldn't be a problem if there were only one env and only one thread.
16:39 edrin and that is important, for example when running the same script twice
16:39 masak but as I said, I'm reevaluating that design right now.
16:39 masak edrin: yes, I think creating a new js thread/env each time is the way to go.
16:39 edrin masak: but for a script in this case the environment is not predictable
16:40 edrin (i mean the global env)
16:40 jonalv edrin: you design means that you can't look at the results from a javascript execution using the console afterwards..
16:40 masak edrin: predictable?
16:40 masak jonalv: exactly.
16:40 masak you nailed it.
16:40 masak but under that design, that is not a problem.
16:40 jonalv I rather the like that you can do that...
16:40 jonalv it's the Matlab way I think
16:40 masak it's sort of the cost of having many threads.
16:41 edrin for example if you do i = 5; in one script and then you do i = i +3; in another script this would work and result in 8 but in fact the other script should fail because in its own logic this is an error!
16:41 edrin jonalv: you mix things up, you must separate the console from a script
16:42 jonalv edrin: you can't do that...
16:42 masak yes, that's the madness that shared memory creates.
16:42 masak let's not do that.
16:42 jonalv edrin: why?
16:42 edrin jonalv: what cant i do?
16:42 jonalv edrin: I think it works excellently to have one js thread
16:42 jonalv run multiple js threads in the same environment. That's crazy bad and ugly
16:43 edrin masak: thus each console, each run script SHOULD have its own environment
16:43 masak edrin: aye.
16:43 masak edrin: remove the "main GUI thread" option and we are in agreement.
16:43 * olass is happy this is leading somewhere
16:43 olass great
16:43 edrin jonalv: i think we are now talking two different thing?
16:43 shk3 olass: what are your plans for doing the release?
16:44 jonalv edrin: I don't thin kso
16:44 olass still working on the welcome page
16:44 olass getting there
16:44 shk3 I think you should publish year features first
16:44 shk3 and i will test speclipse against it
16:44 jonalv edrin: but feel free to point out which two things in that case
16:44 olass shk3: will have to do it after lunch tomorrow
16:44 shk3 and then update the update site
16:44 shk3 sure
16:44 edrin jonalv: so you want one global env? masak, you don't do you?
16:44 shk3 after lunch is fine
16:44 shk3 btw, do you know how context sensitive help in eclipse is working?
16:44 masak no, I don't believe in a global env anymore.
16:44 shk3 pressing f1 has no effect for me
16:45 olass no
16:45 jonalv edrin: I think it might be sane to keep it simple and just have one js environment
16:45 shk3 also not in normal eclipse
16:45 edrin masak: yeah
16:45 olass shk3: never tried that
16:45 olass and got no time
16:45 edrin jonalv: no
16:45 edrin not good
16:45 shk3 sure
16:45 shk3 just a question
16:45 shk3 we don't need it
16:45 jonalv edrin: why not?
16:45 shk3 not now, at least
16:45 masak jonalv: because it makes it impossible to run several js things at once.
16:45 edrin jonalv: because this would result in a confusing global "logic?"
16:46 edrin yeah, for example
16:46 masak jonalv: and our way of preventing that basically didn't work.
16:46 jonalv masak: you are right. You will have to wait for each thing to run sequenially
16:46 edrin this would mean you run *one* script from the beginning to the end of bc
16:46 masak jonalv: that solution is bad for several reasons.
16:46 jonalv edrin: what is a global "logic"?
16:46 edrin jonalv: and not eben this
16:46 masak jonalv: for one thing, tab completion uses the js env...
16:47 masak jonalv: so tab completion doesn't work when the js env is busy.
16:47 edrin jonalv: even if you run it sequentially it is critically unlogic
16:48 CIA-30 bioclipse: ospjuth * r8458 /bioclipse2/trunk/plugins/net.bioclipse.ui/ (17 files in 5 dirs): Completely rewrote welcome/intro page to make use of Eclipse universal intro.
16:48 jonalv masak: yuo shouldn't be bale to write anything at all when js is busy
16:48 olass shk3: that commit was a start
16:48 shk3 great
16:48 masak jonalv: that statement is based on your assumption of one and only one js env.
16:48 edrin jonalv: reagrding the thread safety of managers it might help (a little bit) to run stuff separate (in different js env, in my opinion) but this is not so good, too
16:48 olass shk3: will write a wiki page how to contribute
16:49 jonalv as things are now we basically have two threads, the gui thrad and the js-thread. If you want more threads than things are nog onna get easier...
16:49 masak I think they will.
16:49 edrin jonalv: not easier, but it's a must
16:49 edrin or maybe they will
16:49 edrin masak: i think that depends on the perspective of view
16:50 edrin :)
16:50 masak I guess.
16:50 jonalv masak: one js environment is not my assumption its what I am propsing
16:50 masak jonalv: I know.
16:50 masak we're disagreeing.
16:50 edrin regarding the thread safetiness it is more risky
16:51 jonalv but sure if you wanna use multiple js environments don't let me stop you.
16:52 masak thank you.
16:52 edrin jonalv: in fact this increases the multithreading risks
16:52 jonalv edrin: spot on
16:52 edrin spot?
16:52 jonalv but I guess it doesn't really matter.
16:53 masak in most cases, it won't matter.
16:53 edrin wyh spot?
16:53 jonalv either things are mulit threaded or not
16:53 edrin what do you mean
16:53 jonalv sry forget it
16:53 edrin ok
16:54 jonalv I was just trying to say: "exactly" in a different way... :/
16:54 edrin masak: my editor plugin runs the scripts in their own environments
16:54 edrin jonalv: ok
16:54 edrin masak: maybe you have objects to inject there for bc specific usage?
16:54 jonalv my next question is: Are you gonna have multiple consoles or just one?
16:54 edrin jonalv: multi
16:55 jonalv edrin: as editors or?
16:55 edrin well, it's up to masak, if he has time to implement them
16:55 edrin jonalv: editors?
16:55 jonalv you can't have multiple instances of a view but I guess with editors it's another thing
16:56 jonalv it's actually ratehr fitting. A new editor for each runnign javascript
16:56 edrin jonalv: it would be not multiple views  but tabs
16:56 edrin similar to the eclipse console
16:56 masak edrin: yes, I inject. that's what JsEnvironment does.
16:56 jonalv ah but then you wouldn't be able to show them next to eachother
16:56 masak edrin: if your editor doesn't use that, I don't think it belongs in bc2.
16:56 jonalv I think I would wanna be able to show them next to eachtother...
16:57 edrin masak: no, i mean in the js instanciated by my code
16:57 jonalv why not just use editors?
16:57 edrin masak: what should go in there? the managers are, anything else?
16:57 masak edrin: see the class JsEnvironment.
16:59 edrin is JsEnvironment a manager?
16:59 edrin do you have a link to this class?
17:00 masak edrin: it's not a manager, it's just a wrapper for rhino js.
17:01 masak hold on, I'll get you the URL.
17:01 jonalv edrin: shift-command-t
17:01 jonalv edrin: that brings up a dialog where you can enter class names
17:01 masak jonalv: thanks.
17:01 edrin jonalv: i amnot home
17:01 edrin would need a http svn link
17:01 jonalv edrin: you donät have eclipse? aha... :)
17:02 edrin masak: is there an advantage of running a script through the wrapper, compared to putting the script directly into rhino?
17:02 masak http://bioclipse.svn.sourceforge.net/view​vc/bioclipse/bioclipse2/trunk/plugins/net​.bioclipse.scripting/src/net/bioclipse/sc​ripting/JsEnvironment.java?view=markup
17:02 zarah masak's link is also http://tinyurl.com/dgefr3
17:03 masak edrin: yes. the whole bc2 design is built around this.
17:03 masak or was, until today.
17:03 edrin ;)
17:03 masak :)
17:04 edrin sorry, it's not my intention to reset the design of BC2
17:07 masak it's ok.
17:07 olass shk3: http://wiki.bioclipse.net/index.php?title​=How_to_contribute_to_welcome/intro_page
17:07 zarah olass's link is also http://tinyurl.com/ctdskz
17:07 masak we'll just have to discuss this further.
17:07 masak I'm leaning towards adopting a multi-thread/multi-env model.
17:07 olass masak: yes we must, and we shall
17:08 olass good that we have this discussion
17:08 masak but I still think it's a good idea to wrap things through JsEnvironment.
17:08 shk3 olass: will look at it and try to use it either today or tomorrow morning
17:08 edrin ok, good.
17:08 masak edrin: the JsEnvironment model was built as a reaction to the bc1 rhino solution.
17:08 edrin happy to hear..
17:08 olass shk3: we only need it for sharp 2.0 release so no rush
17:08 olass but now you have something to start from
17:08 shk3 true
17:08 masak edrin: I'm not overly happy about having a plugin called n.b.rhino back in bc2...
17:08 shk3 but needs to be done anyway
17:09 masak edrin: and the only commit comment to that commit reads 'xx'... :)
17:09 edrin well
17:10 edrin yeah
17:10 jonalv masak: what is the problem with the naming of that plugin?
17:10 edrin the name is not good
17:10 masak jonalv: first off, 'rhino' is an implementation, not a technology.
17:10 masak jonalv: second, JsEnvironment already does what that plugin does.
17:11 edrin regarding the "one environment" not...
17:11 edrin but the name is not good
17:11 edrin ranme it to jseditor
17:11 edrin at least i will
17:12 masak good.
17:12 masak edrin: but I don't want anything to call js through other means than JsEnvironment.
17:12 masak edrin: it can handle threading as well.
17:12 masak edrin: ...though _another_ wrapper called JsThread.
17:12 CIA-30 bioclipse: ospjuth * r8459 /bioclipse2/trunk/features/net.​bioclipse_feature/feature.xml: Added intro.universal to deps.
17:13 edrin ah ok?
17:13 masak edrin: that's how the js console does it, for example.
17:13 masak that design was arrived at precisely because we didn't want to run script commands in the main GUI thread.
17:15 edrin but the scripting environment is still missing, right?
17:15 masak edrin: please reformulate that question.
17:16 masak a JsThread wraps its own JsEnvironment.
17:16 edrin i mean the method to run a script in it's own environment with the Js*
17:16 masak eh?
17:17 masak that didn't make your question clearer.
17:17 jonalv olass: are you still there?
17:18 olass sort of
17:18 edrin ew, right now the editor just pushs a script to a js instance by itself. it's possible to remove this action and directly push it the the JsThreadEnvironment where it is pushed in the js instance concurrently created there
17:18 olass but signing off very soon
17:18 * olass is working on the release
17:18 jonalv olass: is the databases view in the net.bioclipse.services plugin now?
17:18 olass yes
17:18 olass and it should be
17:18 jonalv okey if you say so
17:19 olass since we will soon rename it to serviceview
17:19 jonalv ah
17:19 olass ok?
17:19 masak edrin: as far as I understand your question, yes, that's definitely possible.
17:19 jonalv okey then it's as it should be
17:19 olass :-)
17:19 edrin masak: ok
17:19 jonalv :)
17:19 masak edrin: and we'd want such a design if we have one thread per script execution.
17:19 edrin masak: we can do it like this
17:19 masak but I'd still like to go through JsEnvironment instances.
17:20 * jonalv is very happy that edrin and masak is talking this thourgh. You are our JS-guys after all
17:20 jonalv And I am sure that you will find a good solution
17:20 masak these are deep, important questions.
17:21 jonalv as for thread safe managers. I would like to hear what everyone thinks about it. And more spcially if anyone has another solution which we could agree on...
17:22 olass masak: this is at least one step on the way towards a solution
17:22 edrin masak: ok, so i'll make the button to inject the script within the editor to the JsThreadEnvironment as soon as it exists
17:23 masak olass: one that we won't take before the beta release.
17:23 olass masak: true
17:23 olass beta is tomorrow
17:23 masak edrin: with the current model, the JsEnvironment exists from bc2 startup.
17:24 edrin ok
17:24 edrin masak: but you will add a function for the next release?
17:24 masak edrin: a function?
17:24 masak what we're discussing is more complicated than that.
17:25 masak it's a revamping of the whole design around js and threads in bc2.
17:26 CIA-30 bioclipse: jonalv * r8460 /bioclipse2/trunk/plugins/​net.bioclipse.structuredb/ (11 files in 4 dirs): Worked on the databases view. It now shows the structuredb databases but no updating yet and no drag and drop.
17:27 edrin masak: maybe the manager code in bc.rhino can help you with this?
17:27 CIA-30 bioclipse: ospjuth * r8461 /bioclipse2/trunk/plugins/ (4 files in 4 dirs): Set version to 2.0.0.A20090123.
17:28 edrin i mean the job/progress code
17:28 masak edrin: with what? revamping?
17:28 masak yes, possibly.
17:28 masak I will have a look at it.
17:28 CIA-30 bioclipse: ospjuth * r8462 /bioclipse2/trunk/ (11 files in 11 dirs): Set version to 2.0.0.A20090123.
17:28 edrin i think it does many stuff you want or plan
17:33 egonw joined #bioclipse
17:37 olass egonw: I fixed the versioning of plugins
17:37 olass or rather, fixing it now
17:38 CIA-30 bioclipse: ospjuth * r8463 /bioclipse2/trunk/ (17 files in 17 dirs): Set version to 2.0.0.B20090123, not A. Grr.
17:39 masak going swimming.
17:39 * masak waves
17:40 CIA-30 bioclipse: ospjuth * r8464 /bioclipse2/trunk/plugins/ (9 files in 9 dirs): Set version to 2.0.0.B20090123 for beta release.
17:40 olass bye masak
17:40 olass bye jonalv
17:40 olass simma löugnt
17:40 olass swim calmly
17:40 jonalv jupp
17:40 edrin ciao
17:41 egonw olass: ok
17:41 olass bye edrin
17:41 olass bye egonw
17:41 * olass is signing off soon
17:41 * olass is signing off very soon
17:41 edrin good nioght, good way home
17:42 edrin left #bioclipse
17:42 * olass is already at home :-)
17:44 CIA-30 bioclipse: ospjuth * r8465 /bioclipse2/trunk/features/ (7 files in 7 dirs): Set version to 2.0.0.B20090123 for beta release.

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