Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-05-22

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

All times shown according to UTC.

Time Nick Message
05:17 masak joined #bioclipse
05:21 sneumann__ joined #bioclipse
06:06 egonw joined #bioclipse
07:45 rojasm joined #bioclipse
07:48 egonw joined #bioclipse
07:58 rojasm1 joined #bioclipse
08:11 shk3 joined #bioclipse
08:12 egonw shk3: ping
08:14 rojasm1 moin
08:14 zarah hello rojasm1, you fantastic person you
08:14 egonw hi rojasm
08:15 egonw rojasm: please add missing copyright/lincese templates to the files you recently added
08:15 shk3 morning
08:15 egonw hi shk3
08:15 rojasm1 oo thanks zarah. I didn't know you are a programmed computer I would be red
08:15 egonw shk3: did you commit your changes to make SpectrumManager NWO?
08:15 egonw because it currently is not
08:15 rojasm1 ok egonw
08:15 shk3 oops
08:15 shk3 well, I thought it would be
08:16 shk3 so I missed the point it seems
08:16 egonw no worries
08:16 egonw I can fix it...
08:16 egonw shk3: but can you please check first you have no local uncommited stuff regarding this?
08:16 egonw so that I don't duplicate work?
08:16 egonw this is my day off, and need start packing house things for our move next week
08:17 shk3 need to start eclipse
08:17 shk3 mom
08:17 egonw yes, no prob
08:21 shk3 looks like everything is committed
08:21 shk3 ie no local changes
08:22 egonw ok
08:25 CIA-40 bioclipse: shk3 * r10679 /bioclipse2/trunk/plugins/net.bioc​lipse.seneca/src/net/bioclipse/sen​eca/business/ISenecaManager.java: improved comments
08:26 edrin joined #bioclipse
08:40 egonw shk3: "Validates thi job"
08:40 egonw in your last commit
08:40 shk3 hm
08:41 shk3 the job specification better?
08:41 egonw just: thi -> the
08:41 egonw I guess
08:42 shk3 yes
08:43 CIA-40 bioclipse: shk3 * r10680 /bioclipse2/trunk/plugins/net.bioc​lipse.seneca/src/net/bioclipse/sen​eca/business/ISenecaManager.java: corrected typo
08:45 CIA-40 bioclipse: miguelrojasch * r10681 /bioclipse2/trunk/plugins/net.bioclips​e.reaction/src/net/bioclipse/reaction/ (52 files in 11 dirs): Added copyrights to the java classes
09:00 shk3 should we announce the update site problems on the mailing list? edrin asked for it yesterday, may be more people experience that.
09:03 rojasm left #bioclipse
09:05 egonw what happens exactly?
09:06 egonw right now, there is nothing to update
09:06 egonw ola said he would upload a new version this weekend or on monday
09:09 sneumann hi shk3
09:09 sneumann regarding the mzdata plugin
09:10 sneumann (or mzXML or mzML for that matter)
09:10 sneumann I said I could do it, and I definitely can do it.
09:10 shk3 sounds good
09:10 shk3 egonw: there is
09:10 sneumann Result would be a plugin to browse, view, and modify MS files
09:10 shk3 at least speclipse to install
09:10 shk3 but you can't do anything since you always get this missing core.patch
09:11 sneumann *but* that would only be an interface to the XML file
09:11 shk3 sneumann: I would use it for parsing
09:11 egonw right...
09:11 sneumann whereas stuff like jmzML
09:11 egonw shk3: but the speclipse update can wait until monday, perhaps?
09:11 shk3 sure
09:11 sneumann is doing the parsing and (possibly ?) provides a cleaner interface to the actual MS data model
09:11 sneumann i.e. some abstraction from the XML representation.
09:12 shk3 just thought we might want to announce it
09:12 shk3 what's jmzml?
09:12 sneumann http://code.google.com/p/jmzml/
09:12 zarah sneumann's link is also http://tinyurl.com/pbww9b
09:12 sneumann Done by Phil Jones across your corridor ;-)
09:12 sneumann (Not sure now, after the moves at EBI)
09:13 sneumann jmzML (should) also include a semantic validation,
09:13 sneumann which the EMF stuff can not easily do.
09:13 rojasm1 cool I didn't know it
09:14 shk3 thanks to having moved to the portacabins, I can say we have no corridor ;-)
09:16 shk3 jmzml is apache licence, it seems.
09:17 shk3 I just would like to include mzml in bioclipse and need a parser.
09:17 shk3 So what do you think would be better to use?
09:19 jonalv joined #bioclipse
09:19 jonalv oh hai
09:19 zarah oh hai jonalv
09:19 egonw hi jonalv
09:19 egonw shk3: NWO patch for ISpectrumManager being committed
09:19 egonw shk3: please update the specmol manager accordingly
09:20 egonw it's a fairly big patch, but the message explains the steps
09:20 egonw shk3: and more info in the wiki
09:20 * jonalv is downloading beta 5 on Windows. Gonna do some bug hunting :)
09:20 shk3 thanks
09:20 shk3 i will change specmol, spectrum etc. asap
09:20 edrin shk3: ola mailed me and said he removed the old updates for now!
09:20 shk3 sounds good
09:20 shk3 will test
09:21 sneumann__ joined #bioclipse
09:21 CIA-40 bioclipse: egonw * r10682 /bioclipse2/trunk/ (24 files in 15 dirs): (log message trimmed)
09:21 CIA-40 bioclipse: Converted to the New World Order:
09:21 CIA-40 bioclipse: * updated the META-INF/spring/context.xml
09:21 CIA-40 bioclipse: * added requires IJava/IJavaScript interfaces extending ISpectrumManager
09:21 CIA-40 bioclipse: * SpectrumManager no longer implements ISpectrumManager (no NWO manager should
09:21 CIA-40 bioclipse:  any longer extend the interface)
09:21 CIA-40 bioclipse: * updated the Activator to have get methods for both flavors IJava/IJavaScript
09:21 edrin jonalv: when cloding it i get sometimes SWT disposed excetions
09:21 edrin could be my fault
09:21 jonalv edrin: "cloding"?
09:22 egonw OK, me continues working in the house
09:22 egonw bbl
09:22 edrin not aborting print to console after shutdown... (in xws4j for example
09:22 jonalv oh closing
09:22 edrin yes
09:22 edrin closing
09:22 jonalv closing what?
09:22 jonalv Bioclipse?
09:22 edrin yes
09:22 shk3 egonw: that's spectrum, right?
09:23 shk3 sorry
09:23 shk3 I was thinking about seneca
09:23 egonw shk3: yes, that message meant to read as:
09:23 edrin Bug: the enter key of the numloc does not work in the jsconsole
09:23 shk3 my fault
09:23 egonw Convert the SpectrumManager to the NWO
09:23 shk3 I did not do anything to spectrum plugin
09:23 egonw yes, the SenecaManager may need updating too
09:23 edrin Bug: delete key does not delete files in the navigator
09:23 shk3 only seneca
09:23 egonw ah, ic
09:23 edrin these two bugs bug me most
09:23 shk3 well, I tried it on seneca
09:23 egonw that explains :)
09:23 shk3 but I get the problems I mentioned yesterday
09:24 shk3 yes, indeed, that explains
09:24 egonw I was asking about spectrum
09:24 egonw because that gave trouble for the medea feature
09:24 shk3 looking at your changes for spectrum it seems I did everything in seneca
09:24 shk3 ah, ok
09:24 shk3 sorry
09:24 shk3 I was so much thinking about seneca
09:25 egonw yes, that one looks good
09:25 egonw is there a secena feature ?
09:26 shk3 not yet
09:26 shk3 will do that today
09:26 jonalv @pz delete key navigator
09:26 zarah jonalv: I couldn't find anything on 'delete key navigator' in the database.
09:26 shk3 did you look at my problems with seneca?
09:26 jonalv @pz delete navigator
09:26 zarah [1051] Open file does not close when ... http://tinyurl.com/ptcodh
09:26 zarah Entire list at http://tinyurl.com/p85f9q
09:26 egonw shk3: no, not yet
09:26 shk3 ie this "wrong number of parameters"
09:26 shk3 and progress bar?
09:26 shk3 ok, thanks
09:27 egonw was still trying to get medea/speclipse to run properly via features
09:27 shk3 I can't see what's wrong
09:27 shk3 sure
09:27 shk3 no hurry
09:27 egonw but since that was still having problems, I did not get to the manager yet
09:27 shk3 much to do in seneca anyway
09:27 egonw cannot test it, if I cannot run it
09:27 shk3 yes, that sound logical
09:27 shk3 so i do the feature now
09:27 jonalv shk3: if you get "wrong number of argumets" or "could not find method to run" exceptions please file bugs to me
09:28 shk3 will do so
09:36 shk3 jonalv: filed you the bug report
09:36 shk3 doing the feature now
09:40 CIA-40 bioclipse: miguelrojasch * r10683 /bioclipse2/trunk/plugins/net.bioclipse.re​action/src/net/bioclipse/reaction/editor/ (ReactionEditor.java ReactionOutLinePage.java): Move the class ReactionOutLinePage out site of the ReactionEditor
09:43 CIA-40 bioclipse: shk3 * r10684 /bioclipse2/trunk/features/net.biocl​ipse.speclipse_feature/feature.xml: updated version of chemoinformatic dependency
09:48 CIA-40 bioclipse: shk3 * r10685 /bioclipse2/trunk/features/ne​t.bioclipse.seneca_feature/: Initial import.
09:50 CIA-40 bioclipse: shk3 * r10686 /bioclipse2/trunk/features/n​et.bioclipse.seneca_feature/ (.project build.properties feature.xml): Initial import.
09:51 shk3 ok, here is a seneca feature.
09:51 shk3 sorry, but since I can't run from features at all, I can't really test.
09:51 shk3 so if you give it a try, please report problems.
09:57 shk3 egonw: you did not update the spectrum tests, right?
10:12 jonalv what's up with the update site? It doesn't sem to be responding...
10:12 egonw shk3: no, I did not
10:13 egonw shk3: please follow:
10:13 egonw http://wiki.bioclipse.net/index.php?title=How_to_m​ake_a_manager_2#Setting_up_the_testing_frame_work
10:13 zarah egonw's link is also http://tinyurl.com/ouun2r
10:13 shk3 will do that
10:13 shk3 just wanted to avoid we do it twice
10:14 jonalv egonw, shk3: Do you guys know anything about the update site or should I start being worried?
10:15 shk3 I think the update site just needs an update
10:15 jonalv oh okey
10:16 jonalv I will file a bug ticket for it and then head down town to do some shopping and eating
10:16 jonalv I can't think of what else to test without having access to the update site..
10:17 shk3 the seneca stuff is not on the update site
10:17 rojasm1 At the moment saving a ReactionScheme I am not using the ICDKManager mainly because it doesn't exists a domain for IReactionScheme. If I want to create one make sense to put in reaction plugin instead of cdk.business?
10:17 shk3 if you want to look at that, it should be done from workspace
10:18 rojasm1 the ICDKManager can handle only ICDKReaction which is a single reaction
10:21 rojasm1 I think it is good that reaction contains a domain and business package to handle IReactionScheme
10:26 shk3 yes
10:26 shk3 sounds good
10:27 shk3 you might even move methods from cdk manager for reactions
10:27 shk3 to reaction plugin
10:27 rojasm1 ok!
10:27 shk3 if there is a manager in reaction plugin anyway
10:27 shk3 but ask egon
10:27 rojasm1 egonw: what do you think?
10:27 shk3 to put it in cdk manger was really because cdk is used for handling them
10:28 shk3 but might make more sense to put all reaction stuff in one
10:28 shk3 if we have got a variety of methods
10:28 shk3 egonw seem to be gone
10:28 egonw yes, day off
10:28 * jonalv is thinking that it might be time to work on the project to split the cdkmanager over mulitple plugins. That is an intresting project...
10:28 egonw but check screen now and then :)
10:28 rojasm1 but in the same mean we should create the same for molecule
10:29 jonalv question is only if it's worth it...
10:29 egonw jonalv: yes, I think managers sharing namespace will become increasingly interesting
10:29 rojasm1 I will move all reaction stuff from cdk plugin to reaction plugin
10:29 jonalv but I was going for food...
10:29 egonw rojasm1: please set up a bioclipse.domain.IReactionScheme in net.bioclipse.reaction
10:29 egonw rojasm1: make sure to have the reaction plugin depend on Spring AOP
10:29 egonw rojasm1: check the spectrum plugin
10:30 egonw there are three plugins you should depend on
10:30 egonw to get the domain object recordable
10:30 rojasm1 ok!
10:30 egonw rojasm1: I'll review your patches later, so just commit when you think it works
10:30 rojasm1 yes!
10:31 jonalv oh just a quick question. When you say "domain object" what are you guys referring to?
10:31 egonw gone again
10:31 egonw oh
10:31 egonw umm
10:31 jonalv for me a domain object is something that implements IBioObject...
10:31 egonw like bc2's IMolecule
10:31 egonw right
10:31 jonalv good
10:31 jonalv just checking
10:31 egonw that's what I am talking about too
10:31 egonw those things are recordable
10:31 egonw and if people copy/paste the NWO stuff from other plugins
10:31 egonw it seems to assume that
10:32 egonw I remember getting problems...
10:32 egonw anyways...
10:32 jonalv well yes as recordable as anyhting is..
10:32 egonw back to home stuff
10:32 jonalv I am out of here. cya later
10:40 CIA-40 bioclipse: shk3 * r10687 /bioclipse2/trunk/plugins/n​et.bioclipse.specmol.test/: added .settings to svn ignore
11:02 CIA-40 bioclipse: shk3 * r10688 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (3 files in 2 dirs): added a content provider for spectrum files
11:08 CIA-40 bioclipse: shk3 * r10689 /bioclipse2/trunk/plugins/net.bioclipse.specmol/ (3 files in 2 dirs): added labels to new wizard, only spectrum files show on spectrum selection page
11:08 CIA-40 bioclipse: shk3 * r10690 /bioclipse2/trunk/net.bioclipse​.speclipse-updatesite/site.xml: update site updated for beta5
11:14 CIA-40 bioclipse: shk3 * r10691 /bioclipse2/trunk/plugins/net.bioclipse.spe​ctrum/src/net/bioclipse/spectrum/fileconten​tprovider/SpectrumFileContentProvider.java: using new header
11:17 CIA-40 bioclipse: shk3 * r10692 /bioclipse2/trunk/plugins/​net.bioclipse.nmrshiftdb/: added .settings to svn ignore
13:05 rojasm1 one question! why doesn't CDKManager  implement the interface ICDKManager?
13:05 egonw because it no longer should
13:05 rojasm1 mmm
13:05 egonw that's one confusing part, I agree
13:05 rojasm1 what is the difference?
13:06 egonw the ICDKManager interface is what software should use
13:06 egonw Bioclipse will convert this into calls to the actual CDKManager
13:07 rojasm1 but and principle you can miss some method in CDKManager
13:07 egonw the proxy will complain
13:07 egonw if you miss crucial methods
13:07 rojasm1 but not when you are writing the method
13:10 rojasm1 there are some plugin which is still using that e.g. specmol plugin
13:10 rojasm1 and spectrum plugin
13:12 rojasm1 ok
13:13 egonw spectrum not
13:13 egonw fixed that this morning
13:13 egonw specmol, yes
13:16 rojasm1 ups
13:16 rojasm1 ok
13:19 rojasm1 not updated yet
14:06 CIA-40 bioclipse: shk3 * r10693 /bioclipse2/trunk/plugins/net.bioclipse.spectrum​/src/spok/parser/JcampToCMLSpectrumMapper.java: hz-ppm-conversion is indicated
14:08 jonalv joined #bioclipse
14:08 * jonalv is back
14:18 CIA-40 bioclipse: shk3 * r10694 /bioclipse2/trunk/ (12 files in 3 dirs): Cleaned up the IJudge interface a bit, data processing now done by judge
14:27 egonw hi jonalv
14:27 egonw welcome back
14:28 egonw jonalv: ping
14:30 jonalv egonw: pong
14:36 egonw good
14:37 egonw I realized that we need a library that allows us to retrieve information about a manager
14:37 egonw e.g. your proxies need to get a list of equivalent methods
14:37 egonw to decide which one should be called
14:37 egonw I need that for the unit tests too
14:37 egonw to see *if* there are proper equivalent methods
14:38 egonw this library will also allow us to make nice developers documentation
14:38 egonw listing the API of a manager
14:38 egonw indicating for each methods
14:38 egonw ..
14:38 egonw what characteristics that method has
14:38 egonw since we do not hardcode that with annotation anymore
14:41 CIA-40 bioclipse: shk3 * r10695 /bioclipse2/trunk/plugins/net.biocl​ipse.specmol/src/net/bioclipse/spec​mol/wizards/NewSpecMolWizard.java: removed unnecessary line
14:49 jonalv egonw: I don't think I understand...
14:49 egonw something like:
14:49 egonw IManagerAnalyzer
14:49 egonw .getMethodInterfaces
14:49 egonw .getMethodImplementation
14:50 egonw with returns a list of
14:50 egonw IManagerMethod
14:50 egonw (or so)
14:50 egonw which defines the API
14:50 egonw and provides an API to learn what the method does
14:50 egonw could be used by the JSConsoleView 'man' too
14:51 egonw quite like reflection
14:51 egonw but
14:51 jonalv sounds advanced and very nice.
14:51 egonw aggregating stuff from the IManager and Manager
14:51 egonw IManagerMethod.supportsProgressMonitor()
14:51 jonalv but almost too nice. What are the added benefits that we don't already have?
14:52 egonw IManagerMethod.supportsIterating()
14:52 egonw benefit:
14:52 jonalv hm yes very nice indeed
14:52 egonw simple to use API
14:52 egonw one location that knows about the current manager design
14:52 egonw focuses code used by several bits of code:
14:53 egonw JSConsoleView, Spring proxy, unit testing later
14:53 egonw and possible creation of (inline) developer help
14:54 egonw oh
14:54 egonw and could be used later by net.bioclipse.sdk
14:54 egonw for autocompletes
14:54 jonalv yea I guess
14:54 jonalv but this sounds like 2.1 stuff maybe even 2.2
14:54 egonw perhaps
14:55 egonw I could use it very much for the unit testing
14:55 egonw and that is 2.0
14:55 egonw I need to write these bits of code anyway
14:55 jonalv egonw: okey do you have a direct example for what you neeed for unit testing?
14:55 egonw but the .sdk, develop help certainly 2.1
14:55 egonw unit tests I have in mind:
14:56 egonw 1. manager implements a method from the interface
14:56 egonw (sort of coverage)
14:56 jonalv yea that is very needed but covered by the plugin tests
14:56 egonw 2. all methods with a certain pattern have a match in the interface
14:56 egonw true...
14:57 egonw when the AbstractCoverageTest works again
14:57 jonalv ah the other way around is not tested
14:57 egonw (currently broken, see bug report)
14:57 egonw right
14:57 egonw the creation of devel docs is useful too
14:57 jonalv so you want a test checking that all manager methods are reachable from the interface?
14:57 egonw as it would indicate what features are supported
14:57 egonw so the author can ensure that what he intends, is actually understood by BC2
14:58 egonw e.g. he might intend method A to have a callback feature
14:58 egonw the docs would indicate if it had so
14:58 egonw jonalv: yes, certainly
15:03 * egonw is heading to Stockholm to pick up sister in law
15:03 egonw in about 15 minutes
15:03 jonalv okye
15:04 egonw will think about it
15:04 egonw have you found many problems with beta5 on win?
15:04 jonalv egonw: It sounds very nice but also ambitious...
15:04 jonalv egonw: no not really. The update site didn't work so I got stuckthere...
15:05 egonw ok, that's a known problem
15:05 egonw fixed in SVN
15:05 egonw but needs a new upload to SF
15:05 jonalv Now I am over att pollax with a phd student there. Trying to close bugs...
15:05 egonw pollax?
15:05 jonalv polacksbacken
15:05 jonalv MIC
15:05 egonw MIC?
15:05 jonalv next to Ångström
15:06 egonw ok
15:06 egonw is this phd using bioclipse?
15:06 jonalv he is phd student in computer science so, no
15:07 egonw but he is looking for a cool project to test his research on?
15:07 egonw so, found bioclipse?
15:07 jonalv but I try to test Bioclipse on him from time to time
15:07 jonalv nah he is an old class mate who left bio science for computer science...
15:08 egonw ok, have to pack up now...
15:08 jonalv okey
17:06 sneumann__ joined #bioclipse
17:38 edrin left #bioclipse
19:57 CIA-40 bioclipse: miguelrojasch * r10696 /bioclipse2/trunk/plugins/net.bioclips​e.core/src/net/bioclipse/core/domain/ (IReaction.java IReactionScheme.java Reaction.java):
19:57 CIA-40 bioclipse: Add new IReactionScheme which group complex reactions.
19:57 CIA-40 bioclipse: Remove class Reaction which was created as miss error
20:00 CIA-40 bioclipse: miguelrojasch * r10697 /bioclipse2/trunk/plugins/ (8 files in 4 dirs): Removed all methods and classes handling with reactions. Now these methods are moved to reaction plugin.
20:14 CIA-40 bioclipse: miguelrojasch * r10698 /bioclipse2/trunk/plugins/net.bioclipse.reaction/ (26 files in 12 dirs): Create new Domain (IReaction and IReactionScheme) and theirs corresponding Managers. Update plugin for dependence with Spring AOP.
20:18 CIA-40 bioclipse: miguelrojasch * r10699 /bioclipse2/trunk/plugins/ne​t.bioclipse.reaction.test/: Initial import for testing reaction plugin
20:22 CIA-40 bioclipse: miguelrojasch * r10700 /bioclipse2/trunk/plugins/n​et.bioclipse.reaction.test/ (25 files in 11 dirs): Initial Import
21:02 sneumann_ joined #bioclipse

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