Camelia, the Perl 6 bug

IRC log for #bioclipse, 2007-11-21

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

All times shown according to UTC.

Time Nick Message
06:37 egonw_ joined #bioclipse
08:13 masak joined #bioclipse
08:16 EskilA joined #bioclipse
08:23 olas joined #bioclipse
09:23 jonalv joined #bioclipse
09:57 CIA-18 bioclipse: ospjuth * r4308 /branches/bioclipse2/chemoinf​ormatics/net.bioclipse.jmol/ (7 files in 5 dirs):
09:57 CIA-18 bioclipse: Added scripting functionality for Jmol by hooking it into Js-console via the scripting EP. Can't write Jmol status messages (like error reports) to console yet, awaiting patch from masak.
09:57 CIA-18 bioclipse: Call jmol scripts in the following way: jmol.run("spin on")
10:59 olas joined #bioclipse
11:01 egonw_ olas: I got compile problems in bc2.cdk.ui
11:01 egonw_ and bc2.core.tests
11:02 olas tests I can't help you with
11:02 olas cdk.ui do has errors, I know
11:02 olas Have not gotten ther eyet
11:02 olas it is quite a big thing
11:03 olas and I don't know if I will do it or if someone else should (tobias, stefsn, egonw)
11:03 CIA-18 bioclipse: egonw * r4309 /branches/bioclipse2/chemoin​formatics/net.bioclipse.cdk/ (.classpath META-INF/MANIFEST.MF): Exported vecmath so that cdk.ui can use it
11:03 olas I have not written anything in JCPEditor
11:04 olas hope someone can help me out
11:04 egonw_ olas: oh, bad news...
11:04 olas why?
11:04 egonw_ the new CDK renderer code (and future editor)
11:04 egonw_ does not work on..
11:04 egonw_ OS/X
11:04 egonw_ because of a bug in Apples affineTransform java code
11:04 olas WHAT!?
11:04 olas oh no
11:04 egonw_ yeah, nice machines... crappy code :(
11:05 * olas is off for lunch
11:05 * egonw_ is to play Zelda
11:19 olas egonw_: Can we use an open source VM?
11:35 sneumann joined #bioclipse
11:51 olas joined #bioclipse
11:54 egonw_ olas: yeah maybe...
11:54 egonw_ not sure about the of icedtea on Os/X
11:54 egonw_ off for lunch
12:52 * jonalv felt the need to write, so: <http://bioclipse.blogspot.com/20​07/11/bioclipse2-and-spring.html>
12:53 egonw_ jonalv: excellent...
12:53 egonw_ next time, please add some links...
12:53 egonw_ e.g. to the Spring website...
12:53 jonalv right yes I forgot that... :)
12:53 egonw_ you can still edit it...
12:53 jonalv Will do.. :)
13:00 jonalv egonw, There, a lot of links :)
13:01 jonalv joined #bioclipse
13:11 * jonalv is checking out jfeature while awaiting some sort of answer from the Spring-OSGI team.
13:11 jonalv Although the answer probably won't come today...
13:38 CIA-18 bioclipse: egonw * r4310 /branches/bioclipse2/chemoinformatics/ (3 files in 3 dirs): OK, added lots of crap actually used by cdk.ui :( Ola: I know you don't like plugin explosions... but it would really be better to put all these third party libs in separate plugins...
13:40 jonalv egonw: I feel that we need a policy about this. I have felt the same thing about "crap" that Spring uses...
13:45 egonw_ :)
13:45 egonw_ olas: good you are around...
13:45 egonw_ do you have time next Thursday to go to France?
13:45 jonalv lol
13:45 egonw_ olas: seriously...
13:46 egonw_ apparently we made it to the next round for the Trophées du Libre
13:46 egonw_ there is 2000 euro funding...
13:46 egonw_ travel/housing will be arranged...
13:46 jonalv what's going on in france?
13:46 egonw_ all you need to do is give a presentation...
13:46 egonw_ otherwise we don't get the 'special prize' that is awarded to Bioclipse...
13:47 egonw_ only 30 min presentation...
13:48 olas egonw_: I don't think I can
13:48 olas is it impossible for you?
13:49 olas egonw_ ?
13:49 olas Do you have more info? Could you fwd the mail to me?
13:49 egonw_ yes, will forward it
13:49 olas you can't go?
13:50 olas NL is not far from FR
13:53 egonw_ ha...
13:53 egonw_ looked up some details...
13:53 egonw_ we got the 'special prize'...
13:53 egonw_ and article 18 reads:
13:53 egonw_ Details for the jury’s special prize
13:53 egonw_ The jury’s special prize rewards a candidate project that achieves exceptionnal quality but that is already too famous or too big to be rewarded by a prize in its category.
13:54 egonw_ problem is that next week there is a workshop really interesting to my work...
13:54 egonw_ my work on metabolomics...
14:06 jonalv AAh I hate my room maes
14:06 jonalv mates
14:07 olas why so?
14:07 olas at BMC or Akkis?
14:07 jonalv Akkis
14:07 jonalv We are 5 people here right now and people are talking and playing music
14:08 edrin joined #bioclipse
14:08 edrin hi
14:08 jonalv hi edrin
14:08 jonalv wassup?
14:08 edrin here?
14:08 jonalv just in general...
14:09 jonalv Now looking for an article? :)
14:09 jonalv s/Now/not
14:09 edrin doing final corrections of my thesis
14:09 jonalv sweet
14:10 edrin and will give it to my superviser/boss for his final look thursday of friday
14:10 jonalv I am checking jfeature out. It might be something to use in bioclipse developmnet...
14:10 jonalv exciting?
14:10 edrin lets see :)
14:18 masak joined #bioclipse
14:23 thomas_ku joined #bioclipse
14:29 egonw_ bbl
14:32 edrin do you want the rhino scripts to be part of bioclipse2 core, too?
14:33 edrin why is bioclipse2 in branch?
14:33 olas ask egonw
14:34 * olas is off to play bandy
14:34 masak edrin: because it is a brach of bioclipse 1
14:34 olas http://en.wikipedia.org/wiki/Bandy
14:34 masak ...one which redefines almost everything, but still, a branch
14:34 jonalv masak: sort of anyway..
14:34 masak jonalv: ack
14:35 edrin anyway but development should happen in trunk, no?
14:35 masak depends
14:35 masak are you a bioclipse 1 or bioclipse 2 developer? :)
14:35 edrin olas: you play this for a long time already?
14:36 jonalv hehe I am a bioclipse 2 devloper :)
14:36 * masak imagines bc2 will move to trunk when it's ready for production
14:36 olas 24 years
14:36 olas my team: http://www.helenelund-bandy.nu/
14:36 jonalv wow
14:36 masak olas++
14:36 olas http://www.helenelund-bandy.nu/show​Content.asp?type=page&amp;pageID=7
14:37 olas see you tomorrow
14:37 olas left #bioclipse
14:37 jonalv yep
14:37 jonalv bye
14:37 * masak waves
14:39 edrin well, i see no reason why to continue/start new plugins for bioclipse1 now
14:40 jonalv left #bioclipse
14:40 edrin who added the chemoinformatics folder to bioclipse2
14:41 jonalv joined #bioclipse
14:41 masak edrin: olas++, I think
14:41 edrin masak: what kind of project do you do again regarding you thesis?
14:41 edrin eh, why ++ ?
14:42 edrin i wonder why the chemoinformatics folder conatains plugins directly but bioinformatics has folders with features and plugins...
14:42 masak edrin: project: I'm writing a database plugin for a db called BASE2
14:43 edrin ok
14:44 masak edrin: re ++ : just a habit from other IRC channels
14:44 edrin i guess the bioclipse2 architecture is not fix yet... it's more like a in development thing.,..
14:44 masak true
14:44 jonalv definitly
14:45 * egonw_ wonders how to open a PDB file in BC2
14:45 * jonalv has no idea
14:45 edrin anyone minds if I add the bioclipse rhino plugin to bioc2?
14:46 masak edrin: no need
14:46 masak or do you specifically mean the running of script files?
14:47 edrin yes, running of scripts
14:47 masak that'd be nice, but it'd also be nice to tie it into the existing scripting capability
14:47 masak somehow
14:47 * egonw_ hopes masak and edrin work together on the new architecture
14:47 masak I'll have to think about it
14:47 egonw_ that both things will be integrated
14:47 masak but it should work
14:47 edrin what both things?
14:47 masak question is how best to do it
14:48 masak edrin: command line and file scripting
14:48 edrin do you think there is a problem of intercation?
14:48 edrin between files and console?
14:49 edrin masak: what is the existing scripting capability?
14:49 edrin do you mean the console?
14:49 masak yes.
14:49 CIA-18 bioclipse: carl_masak * r4311 /branches/bioclipse2/tools/plugins/: Initial import.
14:49 edrin ah
14:49 CIA-18 bioclipse: carl_masak * r4312 /branches/bioclipse2/tools/p​lugins/net.bioclipse.junit/: Initial import.
14:49 masak no problem of interaction really
14:50 egonw_ masak: why that plugin?
14:50 egonw_ why not reuse the junit plugin that comes with eclipse itself?
14:50 edrin i guess the console already contains the rhino .jar
14:50 CIA-18 bioclipse: carl_masak * r4313 /branches/bioclipse2/tools/p​lugins/net.bioclipse.junit/ (8 files in 4 dirs): Initial import.
14:50 masak egonw_: I'm mainly working on the orders of jonalv
14:50 masak maybe reusing from eclipse is a better idea
14:50 egonw_ sure it is
14:50 masak edrin: yes
14:51 masak it does
14:51 masak jonalv: egonw_ says it's better to use the junit jar file already in eclipse
14:51 jonalv egonw: How is that done?
14:51 egonw_ add a depends on bla.bla.junit?
14:51 egonw_ mom
14:52 egonw_ org.junit4
14:52 egonw_ or org.junit
14:52 egonw_ depending on which JUnit version you use
14:52 edrin i don't understand the relation of junit and scripting
14:52 jonalv I am tired of having to create a new testcase add choose add junit to build path every time I do pde tolols update classpath
14:52 * egonw_ don't think they are related at all :)
14:52 edrin ah
14:52 egonw_ jonalv: right...
14:52 egonw_ the plugins should add that dependency
14:52 masak edrin: no, just 2 discussions at once. no relation
14:52 jonalv but they don't...
14:53 masak back to scripting
14:53 edrin masak: can you point me to the folder where the scripting console is located?
14:53 egonw_ ummm... first remove that net.bioclipse.junit again ?
14:53 jonalv I have tried adding requirement to junit.jar but that doesn'r help
14:53 masak edrin: net.bioclipse.core.scripting
14:53 egonw_ jonalv: may I try?
14:53 egonw_ jonalv: which plugin ?
14:53 jonalv so if you know how to do it please do it in one plugin.tests and then I can see how it's done...
14:54 egonw_ core.tests you mean?
14:54 masak edrin: the console class is called JsEnvironment
14:54 jonalv yes do core.tests
14:54 edrin masak: where exactly, where can i find net.bioclipse.core.scripting in svn?
14:54 masak edrin: it's used in a view in net.bioclipse.core.views called RhinoConsoleView
14:55 masak I should probably rename that to JsConsoleView, becuase I'm using 'js' almost everywhere else
14:55 edrin masak: i still could not find it
14:56 edrin in what folder is this ?
14:56 edrin . /branch/bioclipse2/core... ?
14:56 jonalv scripting?
14:56 edrin yes
14:57 jonalv net.bioclipse.core.scripting
14:58 edrin ah, you added it to the bioclipse core plugin itself!?
14:58 edrin i understand
14:58 edrin may I suggest to create a rhino plugin itself?
14:59 edrin egonw: why is bioclipse developed in branch?
14:59 egonw_ jonalv: why does core.tests not have a plugin.xml?
14:59 masak edrin: branches/bioclipse2/core/p​lugins/net.bioclipse.core
15:00 masak edrin: yes, added to core
15:00 edrin masak: yes found it
15:00 masak edrin: no, it's supposed to be there
15:00 masak I understand the implications of taking it out of core, but I think it should be there
15:00 edrin masak: i would highly recommend to not add it to the main IDE/GUI plugin
15:00 masak ok. please explain
15:01 masak one of the main ideas of bc2 is that scripting should be core
15:01 jonalv core is not ui
15:02 egonw_ and there is a missing dep for core.scripting
15:02 edrin the IDE itself is likely to be completely functional without the scripting support. scripting itself will be independent from the main plugin. for distibuted development it is more clear if you put it in its own plugin. then ppl will not be confused with other files in the main plugin
15:02 masak egonw_: there is?
15:03 edrin masak: scripting is similar to the logging plugin...
15:03 masak edrin: not in bc2. scripting _is_ core in bc2
15:03 masak that's the idea
15:03 CIA-18 bioclipse: egonw * r4314 /branches/bioclipse2/tools/p​lugins/net.bioclipse.junit/: Removed plugin: use org.junit4 that comes with Eclipse instead
15:03 masak egonw_: thx
15:04 egonw_ ah, and the dep on .core is there...
15:04 egonw_ registry.getExtensionPoint("ne​t.bioclipse.core.scripting");
15:04 egonw_ that was confusing me...
15:04 egonw_ is the using the proper naming convention?
15:05 jonalv hang on masaks computer died...
15:05 edrin heh
15:06 CIA-18 bioclipse: egonw * r4315 /branches/bioclipse2/core/plug​ins/net.bioclipse.core.tests/ (.classpath META-INF/MANIFEST.MF): Added missing dep on JUnit4
15:07 egonw_ jonalv: I can now run the test...
15:07 egonw_ though is crashes with a NPE
15:07 edrin uh
15:07 jonalv egonw_: good I will try NPE?
15:07 jonalv what's NPE?
15:07 egonw_ on this line: IExtensionPoint extensionPoint = registry.getExtensionPoint("ne​t.bioclipse.core.scripting");
15:07 egonw_ NullPointerException
15:08 egonw_ seems the plugin was not loaded correctly?
15:08 edrin masak: not putting rhino in the core plugin does not mean it is not part of the basic distribution
15:08 egonw_ because only reason can be that the registry was null
15:08 jonalv I can confirm nullpointerException but it worked earlier
15:09 masak edrin: true
15:09 jonalv unclear what's going on...
15:09 egonw_ which is not too weird...
15:09 egonw_ because there is no plugin activator
15:09 masak but I believe it will be difficult to introduce scripting without a scripting engine
15:09 egonw_ masak: sure...
15:09 masak egonw_: hang on, looking
15:10 egonw_ masak: but wouldn't the rhino plugin be a provider of such an engine?
15:10 masak egonw_: maybe
15:10 egonw_ why would one otherwise have a core.scripting EP?
15:10 masak sounds sane, but I don't know if it's feasible yet
15:10 egonw_ the whole point of such an EP is to have tihngs like rhino in a different plugin
15:10 jonalv egonw_: how did you fix it?
15:10 egonw_ check the commit diff
15:10 edrin masak: i think ppl working on code of the scripting plugin within the core plugin might get confused with code that has actually no direct corelation with scripting/plugin code
15:10 masak egonw_: it's not what the EP is doing
15:11 egonw_ what is that EP doing then?
15:11 masak it's providing namespaces
15:11 masak though it might not be a bad idea to have an EP for scripting langs too
15:11 masak I just don't know yet
15:11 masak edrin: understood
15:12 jonalv Just so you know all this might change if I get any order to the Spring-osgi stuff...
15:12 masak I will think about a core design that's loosely coupled with rhino
15:12 masak jonalv: true
15:12 masak there are so many unknowns
15:14 jonalv ah so it's called org.junit4!
15:15 CIA-18 bioclipse: carl_masak * r4316 /branches/bioclipse2/core/p​lugins/net.bioclipse.core/ (3 files in 2 dirs):
15:15 CIA-18 bioclipse: [net.bioclipse.core]
15:15 CIA-18 bioclipse: * renamed RhinoConsoleView -> JsConsoleView
15:17 jonalv ah so there is an eclipse plugin called junit4 didn't realise that... thx egonw_
15:18 masak edrin: I already know that the current model is insufficient for what we want to do. right now, the view contains the js session, and no other part of bc2 can reach it
15:19 masak instead of having a session of its own, the view will probably need to subscrbe to an existing bc2-wide js session
15:19 masak but that's assumin that there needs to be one, and that it's core
15:20 CIA-18 bioclipse: jonalv * r4317 /branches/bioclipse2/bioinformatics/plugi​ns/net.bioclipse.biojava.business.tests/ (4 files in 3 dirs): added junit4 dep and an empty test for jfeature investigation in biojava plugin
15:22 CIA-18 bioclipse: carl_masak * r4318 /branches/bioclipse2/core/plugins/n​et.bioclipse.core/src/net/bioclipse​/core/scripting/JsEnvironment.java:
15:22 CIA-18 bioclipse: [JsEnvironment.java]
15:22 CIA-18 bioclipse: * now catches syntax errors during evaluation
15:23 edrin masak: hm
15:23 edrin masak: i guess that's not a problem, you can run other scripts in another object i guess
15:25 masak yes, perhaps
15:25 masak it's also possible that that would be very confusing and break the sense of a 'session' for users
15:25 masak because you define a variable in one session, and then it's not there in another
15:26 masak egonw_: I know what causes the NPE now
15:26 masak though I'm not sure how to fix
15:26 masak the offending line interacts with the plugin's plugin.xml file
15:27 masak so maybe the answer is copying details from core to core.tests
15:27 masak but that feels vaguely Wrong
15:27 egonw_ how does it interact?
15:28 masak it asks plugin.xml for an extension point
15:28 masak but since we're in core.tests now, the EP isn't there
15:28 masak there's not even a plugin.xml
15:28 egonw_ what about the 'Problem':
15:29 egonw_ An entry for plugin.xml is required in bin.includes.
15:29 egonw_ that a Problem for the .core plugin
15:29 egonw_ mom, I'll test that
15:29 masak please do
15:29 egonw_ no, does not seem to help...
15:29 masak ah, now I see the 'Problem' too
15:29 masak what's bin.includes?
15:30 egonw_ no idea...
15:30 egonw_ but including the plugin.xml there is new in 3.3 I think...
15:30 egonw_ have not seen that before, or maybe I just don't remember
15:30 masak huh
15:31 masak jonalv: I suspect we're being bitten by that neat package trick of ours
15:31 jonalv oh I haven't listen
15:31 jonalv please what's the problem?
15:32 masak short summary:
15:32 masak code that works in core gets called in core.tests
15:32 masak where it suddenly doesn't work
15:32 masak reason: plugin.xml exists in core but not in core.tests
15:32 masak and gets used in the code that stops working
15:33 masak this seems to be because the code still believes it is in the core plugin, and not in core.tests
15:33 masak so it accesses a plugin.xml that is not there
15:33 jonalv it's this damn eclipse idea of having one class path for each plugin
15:33 masak oh, it's that again? seems to show up often nowadays
15:34 masak so, er, whatdowedo?
15:34 egonw_ jonalv: got bless that decision
15:34 egonw_ (not that I believe in a god... just figure of speech)
15:34 jonalv I think it has given me more problems than anything else...
15:35 egonw_ jonalv: well... try working with several Xerces versions
15:35 jonalv what's Xerces?
15:35 egonw_ oh... wait... two axis versions...
15:35 jonalv I see.
15:35 egonw_ well.. Apache libraries in general
15:35 jonalv but his is broken
15:35 edrin hehe
15:35 masak yes.
15:35 masak feels like an error we could do without
15:36 jonalv If code in core is called from core.tests it shouldn't fo looking ion core.tests
15:36 edrin jonalv: Xerces is a the XML parser of apache
15:36 egonw_ http://maps.google.fr/maps?f=q&amp;hl=en&amp;geoc​ode=&amp;q=villeneuve+saint-germain&amp;sll=45.47​7827,3.185906&amp;sspn=0.006485,0.020084&amp;ie=U​TF8&amp;ll=49.381056,3.352745&amp;spn=0.001505,0.​005021&amp;t=k&amp;z=18&amp;iwloc=addr&amp;om=1
15:36 jonalv code in core should look in core
15:36 * masak has high thoughts of xerxes
15:36 masak jonalv: yes
15:36 masak so why doesn't it?
15:37 jonalv is it the null pointer exception?
15:37 masak yes
15:37 * egonw_ just thinks a plugin should always have a plugin.xml
15:37 edrin egonw: what do you suggest, have a trunk2 for bioclipse2 or move all plugins in trunk in a bioclipse1 folder and also create a bioclipse2 folder within trunk?
15:37 jonalv yes but they don't get generated do they?
15:37 jonalv Is that the problem?
15:37 masak egonw_: well, we could make the plugin.xml in core.tests always mimic the one on core
15:37 egonw_ why?
15:37 jonalv That there is nor plugin.xml in core.tests?
15:38 egonw_ jonalv: no
15:38 * jonalv is confused now..
15:38 masak jonalv: yes, might be
15:38 * jonalv more confused
15:38 masak confusion++
15:38 masak I'll try and create a plugin.xml
15:39 masak in core.tests
15:39 masak see if the problem goes away or changes
15:39 edrin what is villeneuve saint-germain?
15:39 jonalv This has nothing to do with class paths
15:39 jonalv I will soon tell you why it's not working
15:39 egonw_ Trophee de Libre
15:40 * masak is dying with expectation
15:40 egonw_ contest where Bioclipse participated
15:40 jonalv Okey we are running a test here
15:40 masak yes...
15:40 jonalv So we are not running Eclipse's OSGI framework
15:40 masak no?
15:40 masak oh.
15:40 masak so there's no registry...
15:40 edrin egonw oh? what kind of contest?
15:41 jonalv What this code is doing on line 46 is asking the framework for the ExtensionRegistry which doesn't exist
15:41 jonalv since we are not running bioclipse
15:41 masak jonalv: I can fix that
15:41 egonw_ edrin:  http://www.tropheesdulibre.org/
15:41 edrin cool
15:41 masak but that means that we can never test our namespaces
15:42 edrin who will go there?
15:42 jonalv not in this way no...
15:42 masak ...which we will definitely want in the long run
15:43 masak ok, fixed it
15:43 masak test now passes
15:43 masak committing
15:43 edrin many competitors, though
15:44 egonw_ yeah, we didnt' win 1st-3rd prize
15:44 ilogger2 joined #bioclipse
15:45 egonw_ http://www.tropheesdulibre.org/+​Finalists-projects+.html?lang=en
15:46 CIA-18 bioclipse: carl_masak * r4319 /branches/bioclipse2/core/plugins/n​et.bioclipse.core/src/net/bioclipse​/core/scripting/JsEnvironment.java:
15:46 CIA-18 bioclipse: [JsEnvironment.java]
15:46 CIA-18 bioclipse: * log and return if registry does not exist
15:46 CIA-18 bioclipse: * this makes test pass again
15:47 edrin be right back later
15:47 jonalv <http://www.technobuff.net/webapp/pr​oduct/showProduct.do?name=jfeature>
15:47 masak jonalv: thx
16:39 jonalv joined #bioclipse
16:48 jonalv jfeature is not working with Eclipse 3.3 :(
16:53 ilogger2 joined #bioclipse
22:32 edrin joined #bioclipse

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