Camelia, the Perl 6 bug

IRC log for #bioclipse, 2008-10-16

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

All times shown according to UTC.

Time Nick Message
05:29 thomasku joined #bioclipse
05:39 Gpox joined #bioclipse
06:55 thomasku joined #bioclipse
07:12 rojasm joined #bioclipse
07:27 |thomasku| joined #bioclipse
07:40 egonw joined #bioclipse
07:59 jonalv joined #bioclipse
08:03 jonalv hi
08:03 zarah hi jonalv
08:04 Gpox joined #bioclipse
08:09 CIA-30 bioclipse: egonw * r7299 /bioclipse2/branches/jmolFeature/: Setting up a branch for a Jmol feature and perspective
08:12 CIA-30 bioclipse: egonw * r7300 /bioclipse2/branches/ (jmolFeature/ jmolPerspective/): Renamed to make my local git happy
08:15 shk3 joined #bioclipse
08:20 CIA-30 bioclipse: egonw * r7301 /bioclipse2/branches/jmolPerspective/plugins/net​.bioclipse.cdk.jchempaint/src/net/bioclipse/cdk/​jchempaint/widgets/JChemPaintEditorWidget.java: Removed unused import
08:21 CIA-30 bioclipse: egonw * r7302 /bioclipse2/branches/jmolPerspec​tive/plugins/net.bioclipse.jmol/ (3 files in 3 dirs): Added a Jmol perspective
08:21 CIA-30 bioclipse: egonw * r7303 /bioclipse2/branches/jmolPerspectiv​e/plugins/net.bioclipse.cdk.debug/ (2 files in 2 dirs): Added missing help for scripting env
08:21 CIA-30 bioclipse: egonw * r7304 /bioclipse2/branches/jmolPerspective/features/ (5 files in 2 dirs): Added Jmol feature and have the chemoinformatics feature depend on the new Jmol feature and no longer on the Jmol plugin
08:22 CIA-30 bioclipse: egonw * r7305 /bioclipse2/branches/jmolPerspective/p​lugins/net.bioclipse.jmol/src/net/bioc​lipse/jmol/views/StatusListener.java: Attempt to output Jmol messages to the JS console (failed)
08:25 Julio joined #bioclipse
08:26 Julio left #bioclipse
08:30 CIA-30 bioclipse: egonw * r7306 /bioclipse2/trunk/plugins/net.biocli​pse.cdk.debug/META-INF/MANIFEST.MF: Import logging
08:31 CIA-30 bioclipse: egonw * r7307 /bioclipse2/trunk/plugins/net.biocli​pse.cdk.debug/src/net/bioclipse/cdkd​ebug/business/ICDKDebugManager.java: Added help info
08:31 CIA-30 bioclipse: egonw * r7308 /bioclipse2/trunk/features/net.bioclips​e.chemoinformatics_feature/feature.xml: Updated the feature for the 150 plugins
08:31 CIA-30 bioclipse: egonw * r7309 /bioclipse2/trunk/features/net.bioclips​e.chemoinformatics_feature/feature.xml: Added two missing plugins
08:32 CIA-30 bioclipse: egonw * r7310 /bioclipse2/trunk/plugins/net.bioclipse.cd​k.jchempaint/src/net/bioclipse/cdk/jchempa​int/widgets/JChemPaintEditorWidget.java: Removed unused import
08:42 gilleain joined #bioclipse
08:57 olas joined #bioclipse
08:57 olas hi
08:57 zarah hi olas
09:08 jonalv olas, egonw: Didn't we get the jmol perspective to show up in Bioclipse yesterday? It's not there for me when I try from that branch...
09:08 olas yes we did
09:09 egonw yeah, it should show up...
09:09 egonw any warnings in the console about resources not found?
09:10 jonalv no warnings int the console excpet exiting with unsaved changes in the previoius session
09:12 jonalv uhm, jmol requires cdk things...
09:12 jonalv olas: I thought you said it didn't
09:13 olas I might be wrong
09:13 olas what does it require?
09:13 olas hmm, I might have added some CDK stuff now that I think of it...
09:13 olas remove it and follow the red crosses!
09:14 * jonalv is trying to remove those dependencies and see what breakes...
09:14 egonw yeah, I think I saw CDK depedencies too...
09:15 egonw those would need to be removed
09:15 jonalv uhm they seem to be central to that plugin...
09:19 jonalv egonw: I am not sure this is going to be as easy as it seemed...
09:23 jonalv the jmolView has a cdk manager reference as instance variable. net.bc.cdk is used in the centrum of this plugin it seems...
09:25 egonw that needs to be factored out into net.bioclipse.jmol.cdk
09:25 egonw olas: that 150 is really a piece of cake :)
09:26 jonalv there is also something called net.bioclipse.cdk.providers.Ph​armacophoreDistanceConstraint
09:26 jonalv that is being referenced
09:26 olas that's my doing
09:26 egonw ok, maybe call it:
09:26 egonw net.bioclipse.jmol.pharmacore then
09:26 egonw net.bioclipse.jmol.pcore or just
09:26 egonw or just net.bioclipse.jmol.pcore
09:27 olas egonw: are you suggesting we refactor every plugin into at least 10 more?
09:27 olas more plugins is not always the best
09:27 olas somtimes we need to be pragmatic as well
09:27 jonalv olas: I would need your help with that
09:28 olas I know
09:28 olas next week then
09:28 olas structureDB has priority anyway over refactoring out jmol
09:28 egonw no, but it is need in this case, as we want a Jmol feature pur sang
09:28 egonw without having to depend on the full CDK
09:28 olas ok
09:28 jonalv but I wonder if this is the best way to go. What are the benefits of having jmol "free from cdk deps"?
09:29 olas egonw wants it
09:29 egonw much smaller download
09:29 egonw less CDK bug visibility
09:29 egonw less bloat
09:29 olas doenload size is not too important in bioclipse
09:29 egonw that's not what I hear from people
09:29 olas thnk of it from a users perspective
09:29 olas who says DL size is important?
09:30 gilleain But with the fragmented cdk plugins, can't you only have the subset of those that are needed?
09:30 olas I have never heard that
09:30 egonw "who says size is important?"
09:30 olas :-)
09:30 olas size does not matter
09:30 olas or?
09:30 jonalv gilleain: yea I guess...
09:30 egonw gilleain: no, because pcore uses most of the CDK plugins
09:30 gilleain oh.
09:30 jonalv I wonder what pcore is doing in the jmol plugin though...
09:31 olas egonw: seriously, I have never heard people care about DL size
09:31 olas is it true?
09:31 olas then we maybe need to take that seriously
09:31 egonw olas, jonalv: where would I check if there is work on the train track or other delays for train to Stockholm, then bus to Skavsta?
09:32 egonw olas, jonalv: I'll look into these jmol-on-cdk deps next week, and report what I think is possible
09:32 jonalv egonw: I don't know actually...
09:32 olas egonw: I don't know if it exists such a site
09:32 olas egonw: is it true you have gotten a lot of feedback that people actually care about size of Bioclipse?
09:33 olas this really surprises me
09:33 egonw that's what I remember, yes
09:33 olas I mean, I like to have Bioclispe as small as possible, but I have never put focus on it
09:34 olas maybe we should make that a priority then
09:34 olas for example we now include IDE
09:35 olas which could possibly be factored out
09:35 olas but it would require a large effort
09:35 gilleain What do you mean by "IDE" in this context? Integrated development environment?
09:35 olas the org.eclipse.ui.ide plugin
09:36 egonw how do we use that?
09:36 gilleain (sorry if that's a naive question :)
09:36 olas a lot of things depend on it
09:36 egonw perspectives?
09:36 egonw that sort of things?
09:36 olas no
09:36 olas more core things
09:36 olas XMLEditor
09:36 olas FileEditorInput
09:37 olas Subclipse requires IDE
09:37 olas many more things
09:37 olas I think we can work around all, but it would be very troublesome
09:37 jonalv FileEditorInput is something that we use alot, right Gpox?
09:37 olas we could extract only that
09:37 egonw then that seems like a rather hard depends...
09:37 olas that would be no problem
09:38 olas people say "Ship with IDE if you don't mind the added size"
09:38 egonw Gpox: please reply to my question on the JCP meeting too
09:38 egonw the jmol refactoring should be much, much simpler
09:38 jonalv olas: how much added size is the IDE?
09:39 olas no idea
09:39 olas the default Bioclispe DL is not small anyway
09:39 olas we can never get it to be small
09:40 egonw no, but depending on the (full) CDK is 14MB extra
09:40 olas right
09:40 egonw compared to 4MB for just Jmol
09:40 olas but it's split up now. right?
09:40 egonw yes...
09:40 egonw I guess nb.jmol doesn't depend on builder3d, so make that 10MB
09:41 jonalv so Bioclipse core is about 40 MB, and how big is the chemoinforamtics feature?
09:41 egonw at least the same
09:42 egonw 2x CDK ~= 30MB
09:42 egonw + Jmol
09:42 olas yes, it's big
09:42 egonw + a bit more, I guess
09:43 jonalv but really people are downloading movies 10 times taht size. I don't see the point of this. It's not like it is a big download compared to waht people is downloading "for fun"
09:43 olas I totally agree
09:43 egonw ah, but with movies know what they get
09:43 olas that's why I have never cared about size
09:43 egonw this is about a quick try
09:44 olas but it will never be less than 37 Mb anyway
09:44 olas and that is without any feature at all
09:44 egonw well, that beats a 100MB
09:44 olas right...
09:44 olas I think we should try to keep it as low as possible but IMO it is more important to get Bioclipse2 out as fast as possible
09:45 egonw btw, I think download sizes count twice on update sites
09:45 egonw to me it does
09:45 egonw right
09:45 olas yes, seen that too
09:45 olas no idea why
09:45 egonw olas: as said, I'll have a look at the deps, and see how much work it would be
09:46 olas ack
09:46 jonalv firefox is 17 MB, Eclispe vaires from 70 and up I guess....
09:46 jonalv We are in the same ballpark...
09:46 olas Eclipse core is rather large
09:47 jonalv Eclipse classic is 150 MB
09:47 olas egonw, jonalv: regarding pcores, it is not a priority now. If it bothers, you may remove it
09:47 olas it will not be in BC2 release
09:47 olas it will be a later issue
09:48 olas but what we really require is a solid way of sending selections to/from jmol
09:48 olas I made some tries
09:48 olas but they are not 100%
09:48 olas jmol is complex
09:49 olas and there are many ways for selecting things
09:49 edrin joined #bioclipse
09:49 edrin hi
09:49 zarah hi edrin
09:49 jonalv problems just keep popping up
09:50 * jonalv feels like bc2 is further away now then ever before...
09:51 * jonalv dives into structuredb instead for now...
09:53 gilleain Well, bioclipse is a grand plan, it's bound to be a little tricky :)
09:53 egonw gilleain, shk3, others: jonalv, olas and I discussed further development of BC2
09:54 egonw and we discussed to use branches much more, just like with CDK
09:54 egonw so that trunk is what leads to a release
09:54 egonw and that all porting work is primarily done in branches
09:54 egonw until it is stable
09:54 egonw porting, as well as general development
09:54 egonw modular development, on top of modular source code
09:54 gilleain Ah, ok. I saw your branch commits earlier.
09:54 olas yes, it sounds good to do it that way
09:55 gilleain Modular! :)
09:56 jonalv my question is: Should we start throing things out of trunk now?
09:57 egonw the alternative is to make a branch bioclipse-2.0.x
09:57 shk3 which things are you thinking about?
09:58 egonw unfinished plugins
09:58 olas we could sort out quite some stuff from trunk to _playground_
09:58 shk3 seneca might be a candidate
09:58 egonw olas: indeed
09:58 jonalv Btw this would mean that we should remove _playground_
09:58 olas I'd prefer to have trunk to work in towards 2.0
09:59 egonw olas: then again...
09:59 egonw we could even to this the real proper way...
09:59 egonw a trunk is basically the active development branch of something under development
09:59 jonalv egonw: which would be?
09:59 egonw e.g. the Bioclipse core download
09:59 olas I see
09:59 egonw while other features are really projects...
10:00 egonw which have a dependency, of course...
10:00 jonalv egonw: What are you suggesting?
10:00 egonw but this would have a number of advantages...
10:00 egonw and not just download size!
10:00 egonw :)
10:00 olas :-)
10:00 egonw but, more importantly, a well contained set of plugins defining a product
10:00 egonw tagging and branching makes much more sense then
10:01 egonw as MyNMRShiftDB and Seneca features might actually have a different release schedule
10:01 egonw which is the second advantage
10:01 egonw but it would be yet another big repository refactoring...
10:01 egonw but likely worthwhile
10:01 olas I like your proposal in principle
10:01 olas we must consider that some plugins are present in multiple features
10:02 egonw as each 'project' would could have a separate Master-of-the-BioclipseProject
10:02 olas yes
10:02 egonw of MotBP in short
10:02 egonw and be closer to the feature concept
10:02 egonw much closer.
10:02 egonw yeah, so basically: 1 feature == 1 bioclipse project
10:03 olas sounds logical
10:03 jonalv egonw: are you suggeswting that each feature gets's it's own folder under root in the svn repo?
10:03 egonw jonalv: yes, exactly that
10:03 * jonalv likes that
10:03 egonw so, bioclipse2/ would be the bioclipse_feature
10:03 olas how to do with plugins present in more than one feature?
10:04 egonw because we'd only have dependencies on features?
10:04 olas should we discourage that?
10:04 jonalv olas: I think we would have to
10:04 egonw good point, btw
10:04 olas hmm, not sure about just having deps to features
10:04 egonw this is where modular designs are not modular :)
10:04 olas will think b�bout this
10:04 jonalv olas: why not?
10:04 egonw yes, let's let this sink in over the weekend
10:04 * olas will think about this during lunch
10:05 egonw good idea
10:05 * egonw will get some coffee for with his sandwiches
10:05 * olas will cook
10:05 * jonalv will buy soemthing...
10:06 egonw brb
10:06 * gilleain will wait another hour :(
10:06 olas seems we have different opinions how lunch should be carried out
10:06 egonw :)
10:07 gilleain Luckily our lunches do not depend on each other's.
10:39 egonw git branch merging in action for git development:
10:39 egonw http://github.com/git/git/network
10:39 zarah egonw's link is also http://tinyurl.com/5y3muo
11:06 olas egonw: did you start a wiki page for chemoinfo use cases?
11:06 olas in that case, URL?
11:06 olas if not I will start one now
11:07 * olas is updating the wiki about bc 2.0 release
11:07 egonw no, have not started that
11:07 olas ok
11:14 rojasm joined #bioclipse
11:47 olas egonw: around?
11:47 egonw yes
11:48 olas join me on olasegonw
12:04 thomas_ku joined #bioclipse
12:05 thomas_ku joined #bioclipse
12:07 egonw Gpox: are you at UU?
12:08 Gpox yes
12:08 egonw Gpox: shall we meet at 15:00 to talk about JCP?
12:09 egonw and see where we are?
12:09 Gpox ok
12:46 CIA-30 bioclipse: jonalv * r7311 /bioclipse2/trunk/plugins/ (3 files in 2 dirs):
12:46 CIA-30 bioclipse: [Structuredb]
12:46 CIA-30 bioclipse: * net.bioclipse.structuredb.tests should compile now
12:46 CIA-30 bioclipse: * wrote internal business tests for creation of all Properties and Annotations
12:55 olas shk3, Gpox: I can no longer hoover in cdk10.jcp
12:55 olas changes must have broken this
12:55 olas please look at it asap
12:56 shk3 hoover?
12:56 shk3 what is this?
12:56 gilleain hovver
12:56 jonalv bad changes. changes are bad
12:56 shk3 jcp never was intended for vacuum cleaning
12:56 olas hold mouse over an atom
12:56 olas hoovering
12:56 shk3 hovering, you mean
12:56 olas :-)
12:56 gilleain indeed.
12:57 olas sry
12:57 shk3 sorry
12:57 olas IT DOES NOT WORK
12:57 olas anymore
12:57 shk3 what should happen when hovering
12:57 shk3 ?
12:57 shk3 sorry for being stupid
12:57 olas if you have set it to show something, then it should show it!
12:57 shk3 I see
12:57 shk3 sorry
12:57 shk3 will look at it
12:58 olas thx
12:58 shk3 this afternoon
12:58 olas great
12:58 olas realle really need it by tomorrow
12:58 olas new release :-)
12:58 shk3 ok, will try it least
12:58 shk3 shouldn't be too difficult, I suppose
12:58 olas I also get exceptions if I click on an atom in cdk10.jcp
12:59 olas Exception in thread "AWT-EventQueue-0" java.lang.ArrayIndexOutOfBoundsException: 0 >= 0
12:59 olas at java.util.Vector.elementAt(Vector.java:432)
12:59 olas at org.openscience.cdk.controller.AbstractController​2D.lassoSelection(AbstractController2D.java:700)
12:59 olas at org.openscience.cdk.controller.AbstractControlle​r2D.mouseReleased(AbstractController2D.java:558)
12:59 shk3 will look at it this as well
12:59 olas at org.openscience.cdk.controller.SimpleControlle​r2D.mouseReleased(SimpleController2D.java:59)
12:59 olas thx*2
13:03 egonw Gpox:
13:03 egonw coming?
13:04 Gpox yes
13:39 CIA-30 bioclipse: jonalv * r7312 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): [Structuredb] Wrote internal business tests for deleting all Properties and Annotations
13:49 malick joined #bioclipse
14:12 CIA-30 bioclipse: jonalv * r7313 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): [Structuredb] Wrote internal business tests for updating all Properties and Annotations
14:52 olas shk3: any news on the hovering?
14:56 gilleain He's not in the room.
14:56 gilleain He was just checking out the 150 new cdk plugins.
14:58 shk3 just updating bioclipse
15:10 olass joined #bioclipse
15:13 shk3 olas: what were you trying to do exactly?
15:14 shk3 seems to work for us
15:14 olass hmm ok
15:14 olass I'll look more into it then
15:24 gilleain By the way, this happens for me in Windows:
15:24 gilleain http://imagebin.ca/view/RUFuUQ.html
15:24 zarah gilleain's link is also http://tinyurl.com/53rrrs
15:25 CIA-30 bioclipse: jonalv * r7314 /bioclipse2/trunk/plugins/ (7 files in 4 dirs): [structuredb] A few more persistence tests now passes
15:25 jonalv gilleain: That looks familiar... last time I checked that was what happened to me too
15:26 gilleain Yeah, stefan says he gets that too.
16:18 shk3 where is olas(s)?
16:45 rojasm joined #bioclipse
16:48 CIA-30 bioclipse: shk3 * r7315 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: removed deprecated use
17:00 CIA-30 bioclipse: shk3 * r7316 /bioclipse2/trunk/plugins/net.bioclip​se.webservices/META-INF/MANIFEST.MF: it seems axis internally used apache.commons.logging, so this dependency is needed.
17:24 shk3 rojasm?
17:26 CIA-30 bioclipse: shk3 * r7317 /bioclipse2/trunk/plugins/net.bioclipse.​nmrshiftdb/src/net/bioclipse/nmrshiftdb/ (3 files in 2 dirs): added the pubchem name resolution
17:34 CIA-30 bioclipse: shk3 * r7318 /bioclipse2/trunk/plugins/net.biocli​pse.specmol/src/net/bioclipse/specmo​l/outline/SpecmolOutlinePage.java: the spectrum selection now works
17:42 jonalv joined #bioclipse
17:42 jonalv shk3: ping
17:45 shk3 hi jonalv
17:45 jonalv zarah: tell shk3 about revision 7316: Have you tried importing org.apache.commons.logging just like you do with org.apache.log4j I think you will get axis log in among the Bioclipse log then and that's probably what you want. I am not sure that it works though...
17:45 zarah Consider it noted.
17:47 shk3 jonalv: at least it runs then
17:47 zarah shk3: You have new messages. Write '@messages' to read them.
17:47 shk3 but I have no idea where logging ends up
17:47 shk3 I don't even know what axis is logging...
17:48 jonalv shk3: good, because if not org.apache.commons.logging is present it should use the correct implementation
17:48 jonalv when you do import
17:48 jonalv anyway I am of again. I just saw your commit message right now and thought I might be able to catch you... :)
17:49 jonalv bye
17:50 CIA-30 bioclipse: shk3 * r7319 /bioclipse2/trunk/plugins/net.bioclip​se.webservices/META-INF/MANIFEST.MF: apache.commons.logging imported, not dependency
17:55 edrin left #bioclipse
21:03 olas joined #bioclipse

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