Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-04-08

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

All times shown according to UTC.

Time Nick Message
04:46 masak joined #bioclipse
06:26 Gpox joined #bioclipse
06:46 masak joined #bioclipse
06:48 egonw joined #bioclipse
07:00 olass joined #bioclipse
07:06 olass moin
07:06 zarah ni hao olass
07:06 masak ni hao olass
07:08 olass ni hao masak
07:08 olass and to all of you
07:08 olass from all of me
07:08 olass a very
07:08 olass ..very..
07:08 olass ni hao!
07:08 masak awww :)
07:09 * masak smiles
07:16 CIA-58 bioclipse: ospjuth * r9660 /bioclipse2/trunk/plugins/ (4 files in 4 dirs): Added about.ini and feature branding icons to build.props.
07:28 egonw moin
07:28 zarah oh hai egonw
07:29 egonw http://code.google.com/eclipse/
07:29 zarah egonw's link is also http://tinyurl.com/cm222y
07:30 CIA-58 bioclipse: ospjuth * r9661 /bioclipse2/trunk/plugins/ (5 files in 5 dirs): Updated classpaths.
07:31 CIA-58 bioclipse: ospjuth * r9662 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Cleaned up deprecated use of bioobjectdescriber for JCP.
07:45 jonalv joined #bioclipse
07:51 * olass is building beta4
07:51 CIA-58 bioclipse: ospjuth * r9663 /bioclipse2/trunk/features/ (2 files in 2 dirs): Set version to 2.0.0.B20090407.
07:54 jonalv @pz 711
07:54 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=711
08:01 edrin joined #bioclipse
08:01 edrin hi
08:02 zarah privet edrin
08:08 shk3 joined #bioclipse
08:09 CIA-58 bioclipse: ospjuth * r9664 /bioclipse2/trunk/plugins/net​.bioclipse.balloon.business/ (5 files in 2 dirs): Added branding for balloon.
08:12 CIA-58 bioclipse: ospjuth * r9665 /bioclipse2/trunk/plugins/ (5 files in 5 dirs): Set version to 2.0.0.B20090407.
08:12 CIA-58 bioclipse: ospjuth * r9666 /bioclipse2/trunk/plugins/net.biocl​ipse.balloon/META-INF/MANIFEST.MF: Set version to 2.0.0.B20090407.
08:22 CIA-58 bioclipse: ospjuth * r9667 /bioclipse2/trunk/plugins/n​et.bioclipse.data.drugbank/ (5 files in 2 dirs): Added branding for drugbank feature.
08:25 CIA-58 bioclipse: ospjuth * r9668 /bioclipse2/trunk/plugins/net.biocli​pse.data.drugbank/build.properties: Added about.[files] to build.properties.
08:25 CIA-58 bioclipse: ospjuth * r9669 /bioclipse2/trunk/plugins/net.biocli​pse.data.sampledata/about.mappings: Updated copyright to 2009.
08:28 CIA-58 bioclipse: ospjuth * r9670 /bioclipse2/trunk/plugins/net​.bioclipse.data.blueobelisk/ (5 files in 2 dirs): Fixed branding: Resize feature image to 32x32 and added feature branding text.
08:28 jonalv olass: release ETA?
08:28 olass today
08:28 olass any commits in the pipie?
08:29 masak the what? :)
08:29 CIA-58 bioclipse: goglepox * r9671 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.jchempaint/src/net/bioclipse/cdk/j​chempaint/editor/JChemPaintEditor.java: Remove partListener when dispose
08:29 masak "you're going to stock your commits... where?"
08:29 jonalv olass: how much do I have time to do before? I am feeling a crawing for bug hunting and thought that I should do that in the exported version...
08:30 masak s/stock/stick/
08:30 olass jonalv: good thinking
08:30 olass I suspect I need another hour
08:30 jonalv I just closed a bunch of fixed bugs. I think I need to open a few new ones... ;)
08:30 masak wohoo, release before lunch!
08:30 olass jonalv: please write example scripts
08:31 olass that is a good testing procedure
08:31 jonalv olass: any special area?
08:31 olass cheminformatics, please
08:31 olass generate 2D/3D
08:31 olass well
08:31 olass jonalv: if you can, I'd like to have substructure searching of SDFiles
08:32 olass it is important for 2.0 IMO
08:32 olass but is not a bug but a feature request
08:32 olass but since you know how it works...
08:32 jonalv olass: oh I could look into that, should be possible to write a crazy slow script for that as things are now
08:32 jonalv hm or maybe not
08:32 olass jonav: please look at bug 816
08:32 shk3 shouldn't that be part of a more general framework?
08:33 olass yes
08:33 shk3 I once talked about similarity in sdfs
08:33 shk3 if you remember
08:33 olass see bug 816
08:33 shk3 it's similar
08:33 jonalv @pz 816
08:33 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=816
08:33 olass shk3: please file a bug
08:33 olass if not already done
08:33 * masak sets target platform
08:33 shk3 well, I even got code
08:34 shk3 but the sdf editor isn't ready to accept it
08:34 jonalv shk3: I don't know how to append things to an sdf file. Do you?
08:34 shk3 no
08:34 shk3 the problem is
08:35 shk3 right now, things are always read from file
08:35 shk3 you can't add columns in memory
08:35 jonalv shk3: what "things:?
08:35 shk3 the fields
08:35 jonalv uhm I have a feeling we are not talking about the same things..
08:35 shk3 the data displayed in the columns
08:35 shk3 ah ok
08:36 jonalv I was wondering how to append another molecule last to an sdf file...
08:36 shk3 you mean "append additional structures"?
08:36 shk3 yes
08:36 jonalv yes
08:36 shk3 sorry
08:36 shk3 I meant "append more fields to the existing structures"
08:36 jonalv adding properties basivcly means reweriting the entier fiel right?
08:36 shk3 right now, it would mean, yes
08:36 olass shk3, jonalv: Gpox has bugs for this
08:37 shk3 but I think it would be an important feature for the editor
08:37 shk3 to append in memory
08:37 jonalv But when doing that for a 1 GB file we cna't keep the entier file in memeroy at the same time so need to read from onw and append to another...
08:37 CIA-58 bioclipse: ospjuth * r9672 /bioclipse2/trunk/plugins/net​.bioclipse.chemoinformatics/ (about.ini about.mappings about.properties build.properties): Added feature branding text.
08:37 masak are we still using cdk10?
08:37 olass jonalv: Gpox has bugs for this
08:37 shk3 no
08:37 olass masak: NO
08:37 shk3 not cdk10
08:37 olass :)
08:37 olass cdk 1.2.0
08:37 masak olass: it's still in my workspace, and there's errors in it.
08:37 jonalv could we please remove it if we are not using it?
08:37 egonw @pz cytoscape
08:37 olass cdk10 plugins should be dead by now
08:37 zarah [818] Test Cytoscape3 in Bioclipse2      http://tinyurl.com/cozuxl
08:37 zarah Entire list at http://tinyurl.com/d98fk8
08:37 masak olass: what jonalv said.
08:38 masak zarah++
08:38 olass jonalv: yes we can
08:38 olass please do
08:38 jonalv masak: do you want the honours?
08:38 olass masak: please
08:38 * egonw notes this will remove plugins which are not obsolete yet
08:39 shk3 which ones?
08:39 jonalv egonw: like which ones?
08:39 masak hokay.
08:39 egonw try svn ls
08:39 masak so, I just remove it from SVN?
08:41 jonalv egonw: huh? (try what, where, and why?)
08:41 olass egonw: what is not obsolete?
08:41 egonw well, whatever you do, make sure to email to bioclipse-devel
08:41 CIA-58 bioclipse: ospjuth * r9673 /bioclipse2/trunk/plugins/net.bioclip​se.data.sampledata/about.properties: Updated copyright to 2009.
08:41 jonalv egonw: What plugins where you refering to that where not obsolete? If they depend on cdk10 then they have not been working for quite some time....
08:41 egonw there are plugins there of Miguel for metabolomics
08:42 olass indeed
08:42 olass egonw: they cannot work in bioclipse
08:42 egonw yes, they have not been worked on for a while
08:42 egonw sure they can
08:42 egonw cdk10 is just another dep
08:42 masak I would also be very happy if we could kill off the entire directory outdated/ in the process. (and not bring it back.)
08:42 jonalv could we move all of it to a branch or something?
08:42 egonw jonalv: that's an idea I like more
08:42 * masak likes the branch idea
08:43 olass egonw: please email miguel and ask
08:43 jonalv I don't like having old stuff hanging around in trunk. In branches I c an live with it...
08:43 masak jonalv++
08:44 shk3 We should try to discuss stuff with miguel at the cdk workshop.
08:44 shk3 I would like to see his stuff in bc2
08:44 masak the more release-ready the trunk is, at all times, the better.
08:44 egonw it's not old stuff
08:44 egonw there is functionality there that is unique
08:45 egonw it's just called outdated/ because we wanted to show it that we like to see it ported to the new CDK
08:45 masak egonw: will it be part of 2.0?
08:45 olass no
08:45 egonw masak: you suggest to kill a lot more plugins in trunk?
08:45 CIA-58 bioclipse: ospjuth * r9674 /bioclipse2/trunk/plugins/net.bioclipse.cdk.debug/ (6 files in 2 dirs): Fixed branding: Resize feature image to 32x32 and added feature branding text.
08:45 egonw like structuredb?
08:45 egonw jmol.cdk?
08:46 egonw they are not part of 2.0
08:46 olass egonw: be nice
08:46 masak egonw: ideally, yes.
08:46 egonw ah, good
08:46 masak I think work should be done in branches.
08:46 egonw that's called branching
08:46 egonw make a bioclipse2.0 branch
08:46 egonw and delete whatever you like
08:46 masak :)
08:46 egonw that's surely the way forward
08:46 egonw and suggested this several times in the past
08:47 egonw making releases from branches is a very good idea
08:47 egonw and actual practice in many projects
08:47 egonw and you do not even need git for that
08:48 masak egonw: I'm not sure I mean exactly what you mean.
08:48 masak egonw: I was talking mainly of trunk purity, and only merging in things you want to release from the branches.
08:48 egonw masak: funny you say that... as git adept :) in git no branch has any special meaning
08:49 masak egonw: I think you might possibly have misunderstood that policy if you say it like that.
08:49 masak egonw: on the technical level, you are right.
08:49 masak on the social level, there's something to be said for making trunk/master special.
08:50 egonw yes, I like to call my next branch jedimaster
08:50 CIA-58 bioclipse: ospjuth * r9675 /bioclipse2/trunk/plugins/net.bioclipse.ui/ (about.mappings about.properties plugin.properties): Updated version and about text.
08:52 masak egonw: IMHO that does not address what I said. but never mind.
08:54 jonalv egonw: please do not call you next branch jedi master. It brings back terrible memories from the basic programming course where people call their variables all kind of strange names without any meaning. Please call things after what they are
08:54 CIA-58 bioclipse: ospjuth * r9676 /bioclipse2/trunk/plugins/net.bioclipse.qsar.ui/ (5 files in 2 dirs): Fixed branding: Resize feature image to 32x32 and added feature branding text.
08:54 egonw jonalv: like 'master' ?
08:54 jonalv egonw: yup
08:55 masak the master branch.
08:55 masak the one more socially significant than the others.
08:55 masak egonw: I'm actually very unsure about your point right now.
08:56 egonw point:
08:56 egonw in the SVN world is is better to release from a branch
08:56 egonw and not from trunk
08:56 egonw in the Git world it tends to be the other way around
08:56 egonw which is not entirely correct
08:56 egonw as one can have many masters
08:57 olass so git is not correct?
08:57 egonw unlike with SVN
08:57 egonw what makes you say that?
08:57 * olass heard a negative word about git
08:57 masak egonw: I have no argument there.
08:57 olass :)
08:57 masak olass: you're just not reading carefully.
08:57 * egonw sees no negative word
08:57 olass oh
08:57 egonw just the distinction between trunk and master
08:57 olass :(
08:58 olass could we have this discussion after 2.0 and after move to git?
08:58 masak sure.
08:58 masak I still think outdated/ should die.
08:58 masak in trunk.
08:58 masak in branches, you're free to call your folders whatever you like.
08:59 * egonw suggests renaming it to cdk10-based/
08:59 egonw the name outdated is wrong
09:00 egonw it refers to a process, not a fact
09:00 * jonalv wants that stuff out of trunk
09:00 masak indeed.
09:00 jonalv please could we move it to a branch
09:00 masak egonw: renaming is a good first step.
09:00 masak I agree with jonalv that it ideally should be in a branch.
09:01 egonw jonalv: yes, please make a branch of trunk
09:01 masak but at least it won't be called outdated/ if we rename it.
09:01 jonalv masak: If you call a pig a cow it's still a pig...
09:01 egonw but let's rename it first indeed
09:01 egonw then you can kill it, make it die, vaporate is, use your light saber, whatever you like
09:01 jonalv :)
09:01 masak jonalv: indeed, but at least it won't invide people to miscategorize more plugins if we call it cdk10-based
09:01 jonalv masak: okey, is that what you are afraid of?
09:02 masak jonalv: it's a risk, yes.
09:02 masak but I also agree that things not slated for the next release shouldn't be in trunk.
09:02 * egonw really dislikes code future without the author involved...
09:02 egonw really bad management
09:02 egonw does not show any respect for the effort
09:02 masak "code future"?
09:03 egonw removing code?
09:03 egonw that indicates some time line? not?
09:03 masak egonw: I believe someone has to have responsibility for the code quality of trunk.
09:03 egonw sure
09:03 egonw but up front then, not siltently afterwards
09:03 masak egonw: we don't have a common set of standards (yet).
09:04 egonw we do
09:04 egonw not for everyting
09:04 egonw but for commit messages at least
09:04 masak egonw: but I think "only release-ready things in trunk" is a nice guideline.
09:04 jonalv masak: Could youplease talk to the author of this code? (I thought that was Egon...)
09:04 egonw oh, and whitespace, of course
09:04 masak egonw: I'm talking about code quality.
09:04 masak jonalv: seems not.
09:05 egonw jonalv, masak: next time, please do check where code comes from, and what you actually suggest killing
09:05 jonalv egonw: before even talking about it?
09:05 masak egonw: sorry if I offended you somehow.
09:05 jonalv egonw: I think there is a point in talking about things...
09:05 masak I just don't think we should keep unstable code in trunk.
09:06 masak egonw: that's not a criticism of you or anyone else.
09:06 masak it's a discussion of policy.
09:06 * egonw sees a subtile difference between: "let's kill that code" and "mmm... why is that code in outdated/?"
09:06 masak I haven't said anything about the code in question.
09:06 egonw subtile?... spelling correct? not sure
09:06 masak I just don't think we should have an outdated/ directory in trunk.
09:07 masak it sends the wrong signals.
09:07 masak namely, that trunk doesn't contain stable things.
09:07 egonw yes, I misnamed that folder, agreed
09:07 egonw trunk has never contained just stable code
09:08 masak I know.
09:08 egonw half of the managers are not even stable
09:08 masak but I think it's a good policy, eventually.
09:08 grandm44 joined #bioclipse
09:08 egonw yes, in the Git world that is possible indeed
09:08 masak if "stable" is "things we intend to release".
09:08 egonw ok, let's stop now...
09:08 egonw Plan:
09:08 egonw rename the folder
09:08 egonw send bioclipse-devel email about it
09:08 egonw (because everyone will have to clean up there Eclipse workspace accordingly)
09:08 egonw make a branch
09:09 egonw for post-2.0 development
09:09 egonw and clean up trunk
09:09 masak +1
09:09 * jonalv would like it to eather compile or go live in a branch...
09:09 egonw send email to bioclispe-devel explaining why there code is not longer found in trunk
09:09 egonw there code -> their code
09:10 CIA-58 bioclipse: ospjuth * r9677 /bioclipse2/trunk/features/ (2 files in 2 dirs): Set correct update sites in features.
09:12 masak jonalv: I suspect that's what egonw meant by 'clean up trunk'
09:12 jonalv masak: oki
09:14 CIA-58 bioclipse: ospjuth * r9678 /bioclipse2/trunk/plugins/se​.uu.farmbio.hivpred.client/ (8 files in 2 dirs): Added branding: Resize feature image to 32x32 and added feature branding text.
09:16 shk3 jonalv: with respect to appending structures to sdf
09:16 shk3 shouldn't be too hard
09:16 shk3 to make a cdkmanager method for that
09:17 jonalv shk3: could you do that for me?
09:17 olass with unit tests please
09:18 shk3 sure
09:18 shk3 will try
09:22 CIA-58 bioclipse: ospjuth * r9679 /bioclipse2/trunk/plugins/net.bioclipse.xws4j/ (7 files in 3 dirs): Set version for beta4 release, resized feature image to 32x32 and added feature branding text.
09:23 olass damn, so much housekeeping for a release
09:23 CIA-58 bioclipse: ospjuth * r9680 /bioclipse2/trunk/plugins/ (3 files in 3 dirs): Set version for beta4 release: 2.0.0.B20090407
09:23 shk3 just looking at Refactor SpectrumManager into the "New World Order" stuff
09:24 shk3 how is this IFile/path translation done?
09:24 shk3 what do I need to do?
09:24 shk3 just have the two methods?
09:24 jonalv shk3: have you read the wiki page?
09:24 shk3 yes
09:24 shk3 but it's not clear for me
09:25 shk3 do I need to implement both methods?
09:25 egonw shk3: please tell jonalv which sections are not clear
09:25 egonw by telling him what you do not understand about that section
09:25 shk3 the one The translation from a String representing a file path to an IFile
09:25 shk3 egonw: I am doing this
09:26 shk3 right here
09:26 egonw good :)
09:26 shk3 So I repeat:
09:26 shk3 Do I need both methods in the inferface?
09:26 shk3 If yes, both need to be implemented, right?
09:26 jonalv shk3: the interface needs all methods
09:26 shk3 what to put in the methods then?
09:26 shk3 ok
09:26 CIA-58 bioclipse: ospjuth * r9681 /bioclipse2/trunk/plugins/n​et.bioclipse.ui/about.ini: Use transparent icon for bioclipse core feature branding.
09:27 jonalv shk3: http://wiki.bioclipse.net/index.php?title=How_to_​make_a_manager#Then_implement_the_manager_methods
09:27 zarah jonalv's link is also http://tinyurl.com/dypvah
09:27 olass *phew* I think I have fixed all feature brandings now.
09:28 shk3 ah
09:28 masak olass++
09:29 shk3 but this IllegalStateException will not actually be thrown?
09:29 egonw olass: maybe write a script for this?
09:29 egonw olass: at least updating all plugin versions should be easy enough for a qualified Perl hacker
09:29 olass that is not the thing
09:30 shk3 jonalv: and what to do if the method is not void?
09:30 jonalv shk3: what method is not void?
09:30 shk3 if I leave it empty then eclipse will complain, won't it?
09:31 shk3 the one taking String or IFile
09:31 shk3 but wait, if I throw the exception it should be all right, shouldn't it?
09:31 shk3 ah, no errors
09:31 shk3 going to test it
09:32 jonalv shk3: If the method is not void that means it returns something. A method returning something means that you have to wait for it. THat means that you can't start it from the gui thread and then go and do other stuff but that the gui thread would have to wait for the method to end.
09:33 jonalv There is no point in runnign a method that returns as a Job since you can't leave it in a seperate thread anyway..
09:34 shk3 yes
09:35 CIA-58 bioclipse: ospjuth * r9682 /bioclipse2/trunk/plugins/net.biocl​ipse.qsar.ui/META-INF/MANIFEST.MF: Renamed as this name also gives name to feature.
09:36 masak got myself a nice juicy error in JChemPaintEditor.java
09:37 olass oh no
09:37 masak Gpox: anything you might be able to help with?
09:37 olass file a bug?
09:37 masak "cannot resolve ModuleState"
09:37 CIA-58 bioclipse: ospjuth * r9683 /bioclipse2/trunk/plugins/ne​t.bioclipse.data.sampledata/ (META-INF/MANIFEST.MF icons/dbs_32.png): Resized feature branding image to 32x32 and set name to Bioclipse Sample Data.
09:37 masak olass: it makes my Bioclipse not run. :/
09:38 shk3 jonalv: I am now in the And finally we hook in all the Spring AOP stuff-section
09:38 olass sounds bad
09:38 shk3 which file is this actually?
09:38 shk3 I mean: what's the file name?
09:38 shk3 can't find this.
09:38 jonalv META-ING/Spring/*.xml
09:38 jonalv shk3: good point
09:39 masak Gpox is helping me now.
09:39 shk3 hm
09:39 shk3 there used to be two
09:39 shk3 context.xml and context-osgi.xml
09:40 shk3 is there only one left?
09:40 CIA-58 bioclipse: goglepox * r9684 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.jchempaint/src/net/bioclipse/cdk​/jchempaint/handlers/ModuleState.java: Added COMMAND_ID
09:40 jonalv shk3: it varies
09:41 shk3 that's not a good answer
09:41 jonalv shk3: only point in having 2 is that one can be used for test runnign without OSGI
09:41 masak Gpox: that's better. thanks.
09:41 shk3 ah
09:41 jonalv but since we are not doing that I have tarted merging them into one
09:42 shk3 and what would that file be called?
09:42 CIA-58 bioclipse: ospjuth * r9685 /bioclipse2/trunk/plugins/net.bioclipse.cdk.debug/ (META-INF/MANIFEST.MF about.properties): Set name of net.bc.cdk.debug to CDK Extras.
09:42 jonalv shk3: they can be called anything that ends with .xml
09:42 shk3 ah
09:42 shk3 so context.xml would be fine?
09:43 jonalv yes I think that is what it is called in the example right?
09:43 shk3 I think it's not called there at all
09:43 jonalv oh
09:43 shk3 I will try to get this working
09:43 * jonalv heads to look at that text now
09:44 shk3 it seems there is much more in the file now than there used to be?
09:45 jonalv shk3: yes now there is one bean for the JS version and one for the other version
09:45 jonalv not quite double the amount but almost
09:45 * masak cannot launch Bioclipse
09:45 masak just wanted to let you know that.
09:46 masak spending the rest of the day fiddling with icons instead of banging head against wall.
09:46 olass masak: what is the error?
09:46 olass JCP-related?
09:46 masak "Exception launching the Eclipse Platform"
09:46 masak olass: no, the JCP thing is fixed now.
09:47 * masak has plenty of icons to draw
09:47 olass what is the problem?
09:47 masak olass: when launching, I get that exception.
09:47 masak it contains a ClassNotFoundException
09:48 masak the class that cannot be found is called org.eclipse.core.runtime.adaptor.EclipseStarter
09:48 olass mom
09:48 olass committing a fix to product
09:48 masak :)
09:48 CIA-58 bioclipse: ospjuth * r9686 /bioclipse2/trunk/plugins/net.b​ioclipse.ui/bioclipse.product: Updated product with new versions of selected features.
09:48 olass Try starting with this product
09:48 olass that works for me
09:48 * masak tries
09:49 masak olass: works fine now. gracias.
09:49 olass sorry
09:49 olass was my fault not to have that updated
09:50 masak olass: np. you fixed it, that's good.
09:50 olass :)
09:50 masak olass++
09:50 masak half of my commits are fixes to previous mistakes. :)
09:50 masak that's how software moves forward.
09:50 jonalv shk3: Are things working for you?
09:51 shk3 mom
09:51 shk3 still editing the xml file
09:51 jonalv ah
09:51 shk3 starting bc
09:52 shk3 the spectrum manager is not available
09:52 jonalv errors?
09:53 shk3 hm, looks like the xml is corrupt
09:53 shk3 hang on
09:54 shk3 org.springframework.beans.factory.x​ml.XmlBeanDefinitionStoreException: Line 1 in XML document from URL [bundleentry://21407/META-INF/spring/context.xml] is invalid; nested exception is org.xml.sax.SAXParseException: The processing instruction target matching "[xX][mM][lL]" is not allowed.
09:54 shk3 first line is just
09:54 shk3 <?xml version="1.0" encoding="UTF-8"?>
09:54 jonalv huh? :)
09:55 shk3 strange
09:55 shk3 wait a second
09:55 * jonalv has no idea at all
09:55 shk3 a space at the start
09:55 shk3 that might be a problem
09:56 shk3 ah indeed
09:56 shk3 hm, calling
09:57 shk3 spectrum.loadSpectrum("Sample Data/spectra/spectrum1.xml")
09:57 shk3 gives me
09:57 shk3 java.lang.IllegalStateException: This method should not be called
09:57 jonalv okey then we have somewhere to start from :)
09:57 shk3 so that seems not to work
09:57 * jonalv has a bad feeling about htis...
09:58 jonalv @pz 741
09:58 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=741
09:58 shk3 shouldn't be there some advisor for this in interceptorNames
09:58 shk3 ?
09:59 shk3 or is JSJobCreatorAdvice doing it?
09:59 jonalv shk3: spot on, JSJobCreatorAdvice is not doing it. But it probably could...
10:00 jonalv I need to fix this...
10:00 shk3 ok
10:00 shk3 shall I commit?
10:00 jonalv it has not been a problem until now
10:00 jonalv shk3: hm is this a plugin that will be included in todays release?
10:00 shk3 yes
10:01 shk3 will do a speclipse feature release later
10:01 jonalv shk3: would not be ideal to introduce a known bug just before the release...
10:01 shk3 yes
10:01 shk3 I can do the release and later we look at this
10:01 shk3 no problem
10:02 jonalv let's save this to after this beta release then I say
10:02 CIA-58 bioclipse: ospjuth * r9687 /bioclipse2/trunk/plugins/ (6 files in 5 dirs): Fixed broken line endings in about.properties.
10:02 * jonalv looks forward to the day we starts using git, then we can just place it in a beanch I suppose...
10:02 CIA-58 bioclipse: ospjuth * r9688 /bioclipse2/trunk/plugins/net.bio​clipse.qsar.ui/about.properties: Fixed broken line endings in about.properties.
10:03 shk3 I hate git
10:03 shk3 I hope we will never change to git
10:03 jonalv shk3: why? This is a perfect example of when git would have helped us...
10:03 CIA-58 bioclipse: ospjuth * r9689 /bioclipse2/trunk/plugins/ (2 files in 2 dirs): Fixed broken line endings in about.properties.
10:04 shk3 the problem with git is I can't use it
10:04 jonalv shk3: you can learn, can't you?
10:05 shk3 I have some cdk changes and have no idea what to do now
10:05 * jonalv is also learning git...
10:05 shk3 but git is so hard to learn
10:06 CIA-58 bioclipse: ospjuth * r9690 /bioclipse2/trunk/plugins/net.bioclip​se.balloon.business/about.properties: Fixed broken line endings in about.properties.
10:06 egonw shk3: I can talk you through it
10:06 egonw it's really not that different from CVS and SVN
10:06 egonw in particular not the things you can only do in CVS
10:07 shk3 lets have a look at what I would do with my cdk stuff later
10:09 jonalv lunch
10:14 CIA-58 bioclipse: shk3 * r9691 /bioclipse2/trunk/ (5 files in 5 dirs): changed version to 2.0.0.B20090407
10:34 CIA-58 bioclipse: ospjuth * r9692 /bioclipse2/trunk/plugins/net.bioclipse.ui/ (4 files in 2 dirs): Externalized about text to plugin.properties.
10:48 CIA-58 bioclipse: ospjuth * r9693 /bioclipse2/trunk/plugins/net.bioc​lipse.data.sampledata/plugin.xml: Commented out spectra sample data that should be moved to a separate plugin, see bug #823.
10:48 zarah bug #823 | http://tinyurl.com/cs7ahv
11:07 olass ok
11:07 olass anyone with pending commits?
11:07 olass beta4 is on the way
11:08 olass Gpox? Anything?
11:08 olass well, it can go on update site
11:09 * olass is building beta4 now
11:09 olass hmm
11:21 olass damn
11:22 olass when installing balloon, the exec/balloon is not executable
11:22 olass how can I make it that?
11:22 olass it is in SVN
11:22 olass when you package/export, the *nux file permissions are lost...
11:22 olass damn
11:22 * egonw suggests to have the balloon plugin check and fix it on runtime
11:23 olass ?
11:23 olass how can I fix it on runtime?
11:27 masak joined #bioclipse
11:27 olass ok, find a solution
11:27 masak rehi
11:27 olass found
11:27 olass wb masak
11:27 masak ty olass
11:52 olass managed to solve the file permissions issue
11:59 jonalv Gpox, egonw: Can I export a png image file from jchempaint somehow?
12:00 olass jonalv: that is an important task indeed
12:00 olass file a bug
12:00 jonalv so the feature is not there?
12:01 olass shk3: I urgently need your assistance
12:01 olass Creat2D and Create3D does not work
12:01 olass filing a bug now
12:01 Gpox @pz 370
12:01 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=370
12:02 olass egonw: around?
12:03 egonw sort of
12:03 olass egonw: do we have a cmlcontenttype5d?
12:03 egonw yes
12:03 olass ok
12:03 CIA-58 bioclipse: ospjuth * r9694 /bioclipse2/trunk/plugins/net.bioclipse.c​dk.ui/src/net/bioclipse/cdk/ui/handlers/ (Create2dHandler.java Create3dHandler.java): Added nullProgressMonitor() to overcome loading error in New World Order.
12:07 olass egonw: I need your help
12:07 olass please look at and comment on bug #825
12:08 olass then assign to shk3 (it's his Action)
12:08 olass but you can probably make out what is wrong fast
12:08 olass egonw: I need this in the release
12:08 olass else I must remove CDK generate 3D action
12:08 olass egonw: can you please assist with this ASAP?
12:13 olass shk3: I urgently need your asisstance with bug #826
12:13 egonw comment added on 825
12:14 olass and reassigned to shk3?
12:14 egonw no, please see comment...
12:14 egonw I am not entirely sure what the problem is yet
12:14 egonw and cannot test it myself right now
12:15 egonw it might be my task really
12:15 egonw that is...
12:15 egonw not sure yet if the action is the fault
12:15 egonw and that it just forgot to trigger a refresh in the navigator
12:15 maclean joined #bioclipse
12:15 egonw or that there is something wrong with content type detection...
12:16 maclean jonalv : you can export png from bioclipse-jcp, yes.
12:16 jonalv maclean: I can? How?
12:16 jonalv maclean: there is a bug ticket open about it...
12:16 * maclean forgets, but there is a way
12:16 olass egonw: you commented on 1 of 3 questions
12:17 olass added a comment again to 825
12:17 maclean there's a manager method - aha! called 'snapshot'
12:17 * maclean is memory-dredging
12:17 egonw olass: did not see any question marks
12:17 maclean so you can do like "jcp.snapshot("/path/to/file")
12:17 maclean +"
12:17 olass egonw: please read the bug description again
12:18 olass egonw: Opening in JCP gives ERROR
12:18 olass that is no question
12:18 olass it's a statement
12:18 olass the wuestion is, why is this?
12:18 olass thought that question was obvious
12:18 egonw good quesiton
12:18 egonw that is one though
12:18 egonw not thre
12:18 egonw e
12:18 olass :)
12:18 jonalv maclean: jcp.snapshot does not exist. However jmol.snapshoot does...
12:18 maclean what!!?
12:18 maclean !!
12:19 olass shk3: around?
12:19 olass maclean: where is shk3?
12:19 maclean <-
12:20 maclean (sorry : not sat down - shall I ask him to?)
12:20 shk3 here I am
12:20 maclean now he's ->
12:20 olass hi shk3
12:20 shk3 hi olass
12:20 olass I have bugs about the generate 2D and 3D
12:21 olass they are urget
12:21 olass do you have time to fix them now?
12:21 olass urgent
12:21 olass else I must strip them from beta4
12:21 olass so need to know
12:21 shk3 will look at them
12:21 olass ok
12:21 olass bug 825 and 826
12:21 olass and one more coming up
12:21 jonalv olass: strip from beta4?
12:21 shk3 we have a meeting at 1.30, though
12:22 shk3 would do it afterwards
12:22 olass hmm
12:22 jonalv @pz 825
12:22 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=825
12:22 shk3 will start now
12:22 jonalv @pz 826
12:22 zarah http://pele.farmbio.uu.se/cgi-b​in/bugzilla/show_bug.cgi?id=826
12:24 shk3 when I try to create 3d coords,
12:24 shk3 i get this error
12:24 shk3 This manager method should not be called, loadMolecules(IFile, BioclipseUIJob)DEBUG - java.lang.UnsupportedOperationException: This manager method should not be called, loadMolecules(IFile, BioclipseUIJob)
12:24 shk3 which I had yesterday
12:25 olass shk3: please update!!
12:25 shk3 who removed the loadMolecule method?
12:25 olass I already fixed that one
12:25 olass shk3: always work on the latest version
12:25 olass [14:06] < CIA-58> bioclipse: ospjuth * r9694 /bioclipse2/trunk/plugins/net.bioclipse.c​dk.ui/src/net/bioclipse/cdk/ui/handlers/ (Create2dHandler.java Create3dHandler.java): Added nullProgressMonitor() to overcome loading error in New World Order.
12:25 shk3 if I would not have come across it in spectrum yesterday, it would have been unnoticed
12:25 shk3 could such major changes please be announced?
12:25 shk3 or better,
12:25 olass shk3: please read jonalv's doc on New World Order
12:26 shk3 could the person doing it take care of method calls?
12:26 olass shk3: I fixed it for you
12:26 shk3 thanks
12:26 olass there are other errors...
12:26 shk3 yes
12:26 shk3 checking these
12:26 olass shk3: can you fix them really fast?
12:27 shk3 not sure
12:27 shk3 just updating
12:27 * egonw is confused...
12:28 egonw have their been new changes to the New World Order in the last 3-4 weeks?
12:28 masak we have always been at war with Oceania.
12:28 shk3 i think the fix shouldn't be  problem
12:29 shk3 but we have the meeting now.
12:29 olass how long?
12:29 * egonw will run the CDKManager unit test
12:29 * olass is in the last stages of release
12:29 shk3 some time left
12:29 shk3 I try
12:29 olass shk3: also note bug #830
12:29 egonw and it better be there are no failing ones
12:29 olass yes, it better
12:30 * olass have not run it
12:30 olass if errors, hunt the guilty down...
12:30 masak guilty until proven innocent.
12:30 jonalv kill kill burn!
12:31 masak ex-TERM-inate!
12:31 egonw fortunately, I have been busy so much with the VR applications, there is little chance I was the guilty one
12:31 * jonalv just realises that he probably is the guilty one
12:31 egonw jonalv--
12:31 olass is there a guilty one?
12:31 masak ex-TER-minate!
12:31 egonw kill
12:31 shk3 the problem is the saveMolecule method
12:31 egonw no unstable code in trunk
12:31 egonw :)
12:31 masak LOL
12:31 shk3 which explicitly says it saves in format it was loaded
12:31 shk3 looking at it
12:32 olass shk3: please create a test for it
12:32 masak egonw: you seem askew on purpose today. :P
12:32 egonw I what?!?!?!
12:32 masak exactly.
12:32 egonw :)
12:32 egonw bloody grant applications...
12:32 egonw makes you braindead
12:32 masak they should call them 'grunt applications'.
12:33 olass shk3: if you haven't fixed the generate2D/3D until 16.00 swedish time (that is in 1.5 hours) I have to release without CDK generate2D/3D actions
12:33 CIA-58 bioclipse: gilleain * r9695 /bioclipse2/trunk/plugins/net.bioclipse.jmol/sr​c/net/bioclipse/jmol/business/JmolManager.java: Whitespace correction.
12:33 shk3 well, if this fucking meeting wouldn't be there
12:33 egonw olass: I suggest that if there are unit test failures in CDKMnager, there will be no release
12:33 CIA-58 bioclipse: jonalv * r9696 /bioclipse2/trunk/plugins/net.bioclipse.jmol/src​/net/bioclipse/jmol/business/IJmolManager.java: added params info to @publishedMethod in jcp.snapshot
12:33 jonalv olass: if you release without those actions I am afraid that scripts are goign to fail
12:33 olass egonw: are there errors?
12:33 egonw dunno yet
12:33 egonw updating
12:33 jonalv olass: oh without the gui actions showing?
12:33 olass yes
12:33 * jonalv thinks he understands
12:34 olass egonw: agreed, if errors no release
12:34 shk3 the guessFormatFromExtension gives null for x.mdl
12:35 egonw shk3: checking that...
12:36 egonw shk3: .mdl is, AFAIK, part of what CDK defines for that format...
12:36 egonw but checking...
12:36 egonw no, .mdl is not defined for MDLV2000Format
12:36 egonw which would explain it
12:37 shk3 the problem is:
12:37 shk3 guessFormatFromExtension
12:37 shk3 has:
12:37 shk3 if (aFormat == MDLFormat.getInstance()) {
12:37 shk3 // never match this one
12:37 shk3 } else if (file.endsWith("."+aFormat.g​etPreferredNameExtension())) {
12:37 shk3 return aFormat;
12:37 shk3 }
12:37 shk3 so it never matches MDL
12:37 shk3 why so?
12:37 egonw because that format is no longer used
12:37 egonw MDLFormat != MDLV2000Format
12:39 shk3 ok
12:39 shk3 if extension is mol
12:39 shk3 it works fine
12:39 shk3 if extension is mdl,
12:39 shk3 the format is not found
12:39 shk3 that's the problem
12:39 egonw right
12:39 shk3 so egonw to blame
12:39 egonw depends
12:39 shk3 no
12:39 shk3 never
12:39 olass shk3: please fix it
12:39 shk3 egonw to balme
12:39 shk3 blame
12:39 shk3 blame
12:40 shk3 blame
12:40 zarah blame
12:40 olass shk3: please fix it
12:40 shk3 thanks zarah
12:40 * egonw feels some retaliation in the air
12:40 egonw :)
12:40 olass :)
12:40 shk3 well, it looks like
12:40 shk3 matching can be done to one extension only
12:40 shk3 right now
12:40 egonw olass: but I agree with shk3... this is not his task to fix
12:40 olass is it yours?
12:40 egonw possibly...
12:40 olass egonw: could you please?
12:41 egonw maybe the .mdl is assigned to the BC2 content type
12:41 egonw but it is not in the CDK content type...
12:41 * olass feels that egonw knows best about the contenttypes
12:42 egonw shk3: btw, I do see a distinction between blame and responsibility
12:42 shk3 we could remember type when loading
12:42 shk3 and reuse that later
12:42 shk3 would be better than relying on file extensions anyway
12:42 egonw subtle but significant, IMHO
12:42 shk3 responsibility
12:42 shk3 responsibility
12:42 shk3 responsibility
12:42 zarah responsibility
12:42 shk3 thanks zarah
12:42 egonw shk3: indeed, and that's why I like peer review so much
12:43 shk3 I have a look if that's possible
12:43 egonw I like to peer-review everything I take responsibility for
12:43 egonw not needed
12:43 egonw already looking at it
12:43 * jonalv is deeply stuck in a merge conflict
12:43 jonalv I hate bioclipse.product
12:44 CIA-58 bioclipse: carl_masak * r9697 /bioclipse2/trunk/plugins/net.bioclipse.scrip​ting.ui/src/net/bioclipse/scripting/ui/views/ (JsConsoleView.java ScriptingConsoleView.java):
12:44 CIA-58 bioclipse: [JsConsoleView] made line-breaking a lot saner
12:44 CIA-58 bioclipse: Seems I had tried to solve the same problem with different methods in two
12:44 CIA-58 bioclipse: different places. The result was an unholy mess. Removed one of the solutions,
12:44 CIA-58 bioclipse: making the result readable again. Was the solution I removed used somewhere?
12:44 CIA-58 bioclipse: Hope not. If you track your problems down to this commit, it probably was...
12:44 shk3 ok
12:45 olass jonalv: why are you messing around in product? That's my responsibility!
12:45 olass stay off it
12:46 masak "This product is too small for the two of us."
12:46 olass *morr*
12:46 jonalv olass: I have no idea... I hues it's because I want to try running things...
12:46 jonalv |gues
12:46 jonalv |guess
12:48 olass egonw, shk3: ETA for the Generate 2D/3D actions?
12:50 shk3 what's eta?
12:50 shk3 egonw said he looks at a fix
12:50 olass Estimated Time for Arrival
12:50 shk3 I am making a test right now
12:50 olass excellent
12:50 * olass is happy then
12:50 olass Then I'll hold beta4 a while
12:51 CIA-58 bioclipse: ospjuth * r9698 /bioclipse2/trunk/plugins/ (7 files in 4 dirs): Added executable file permissions on PDE build using a permissions.properties file.
12:51 * jonalv thinks he has managed to merge the shit now
12:51 jonalv no I get merge conflict each time I update
12:53 egonw sure
12:53 egonw bioclipse.produce changes each time you need to run BC2
12:53 egonw because you need to remove all features, save, add them again, save again
12:53 egonw bioclipse.product should not be in SVN, I think
12:54 egonw but only a template
12:54 egonw bioclipse.product.template
12:54 egonw which we expect people to copy when setting up the workspace
12:54 egonw or maybe an ant build.xml to do it
12:56 jonalv egonw: yes I know
12:56 jonalv egonw: I had dirt in the stash...\
12:58 * maclean has implemented jcp. snapshot, need to test it, though.
12:58 maclean To the meeting!
12:58 shk3 I committed a test
12:58 shk3 which fails
12:58 shk3 for this problem
12:58 shk3 egonw: good luck
12:58 * shk3 is off for the meeting
12:58 CIA-58 bioclipse: shk3 * r9699 /bioclipse2/trunk/plugins/net.b​ioclipse.cdk.business.test/src/ (2 files in 2 dirs): added a test for bug #826
12:58 zarah bug #826 | http://tinyurl.com/dxywj8
13:00 CIA-58 bioclipse: shk3 * r9700 /bioclipse2/trunk/plugins/net.bioclipse.cdk.​business.test/src/testFiles/polycarpol.mdl: added a test for bug #826
13:00 zarah bug #826 | http://tinyurl.com/dxywj8
13:03 CIA-58 bioclipse: ospjuth * r9701 /bioclipse2/trunk/features/net.bioc​lipse.balloon_feature/feature.xml: Wrote license, copyright, and deps for balloon feature.
13:05 CIA-58 bioclipse: ospjuth * r9702 /bioclipse2/trunk/features/ (2 files in 2 dirs): Updated feature information, e.g. added description.
13:06 egonw on the phone with Jarl
13:07 olass egonw: let me know how progress goes
13:07 * olass will get coffee
13:13 CIA-58 bioclipse: jonalv * r9703 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: Changed default name generated by cdk.loadmolecules from filename-index to molecularformula. If cdk finds a CDKconstants.TITLE that one should still be used instead.
13:14 CIA-58 bioclipse: goglepox * r9704 /bioclipse2/trunk/plugins/net.bioclipse​.cdk.jchempaint/src/net/bioclipse/cdk/j​chempaint/editor/JChemPaintEditor.java: Added check for if widget is disposed when wanting a redraw.
13:14 egonw olass: ping
13:15 olass pong
13:15 olass egonw: pong
13:15 egonw I'm having trouble running the PluginManagerTest
13:15 egonw saying something along the lines:
13:16 egonw Missing Contraint: Require-Bundle: bla
13:16 olass in run config
13:16 egonw where bla can be several things, like
13:16 olass remove all plugins
13:16 egonw ok...
13:16 olass add required
13:16 olass target.platform could also be the case
13:17 egonw reran set target plarform earlier
13:17 egonw killed the run config
13:17 egonw it now seems to be runnign the tests again
13:17 olass good
13:17 egonw except for some time out...
13:18 egonw mmm
13:19 egonw it cannot allocate the CDK manager
13:21 egonw and I missed the StARLite schema walkthrough...
13:22 jonalv there was a StARLite schema walktrhough? ANd we missed it? :(
13:22 jonalv Oh noose
13:23 egonw well, I guess we did our walkthrough ourselves already...
13:23 egonw what would have been interesting...
13:25 egonw jonalv: java.lang.IllegalStateException: Could not get the CDK manager.
13:26 egonw jonalv: ping
13:26 jonalv egonw: pong
13:27 jonalv egonw: what did you do to get that? It works for me... :)
13:27 egonw I have trouble running the CDKManagerPluginTest
13:27 jonalv oh test
13:27 * jonalv has nut run tests for alomst an eternity...
13:27 egonw please try
13:27 jonalv oki
13:29 olass egonw: CDKManagerTest runs for me
13:29 olass with 27 errors and 2 failures :(
13:29 olass must go away for laundry
13:29 olass be back in a while
13:29 jonalv egonw: has CDKManagerPlugintest worked earlier?
13:29 olass oh yes
13:30 * jonalv looks into it
13:31 egonw olass: that one runs for me too, but without errors
13:31 egonw jonalv: sure!
13:31 egonw does anyone actually care about the unit tests??
13:32 jonalv egonw: I stopped caring when the nightly unit test runnign project got movbed to post 2.0
13:32 stain :-(
13:33 egonw jonalv--
13:34 jonalv egonw: I think that was quite natural way for me to react. After all it can't be very important when it is not prioritized...
13:34 egonw jonalv: about commit 9703
13:34 egonw why that change?
13:34 egonw are you sure no one is depending on that?
13:35 jonalv egonw: depending on it? How? If someone is depending on that they are on deep water...
13:35 egonw sure, but that's not an argument
13:35 egonw there  must have been a reason for the code
13:35 jonalv egonw: Do you read the entire commit comment?
13:36 egonw yes, I actually did
13:36 jonalv egonw: then you know that if CDK finds another title than that one will be used
13:36 * egonw hates functional changes on release day (period)
13:36 jonalv egonw: aha yea good point. We need branches...
13:37 egonw so, does CDKPluginManagerTest not work for you either?
13:37 egonw then don't worry about that commit
13:37 egonw because we will very likely not have a release today anyway
13:38 olass egonw: sorry, CDKManagerPluginTst runs here
13:38 olass with 27 errors and 2 failures :(
13:38 jonalv egonw: I am looking at trying to get that test to run now
13:38 egonw ah, ic...
13:38 egonw very good
13:38 egonw so no release
13:38 olass guess not :(
13:39 olass many are errors in the save()
13:39 olass jonalv: did you not run tests after you changed CDKmanager?
13:39 olass jonalv--
13:39 jonalv olass: what change are you referring to?
13:39 CIA-58 bioclipse: egonw * r9705 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: Added comment on why that one should not be matched
13:39 egonw not that his CDKManager has anything to do with saveAs() :)
13:39 olass now world order I guess
13:40 egonw but I guess the NWO may have something to do with it
13:40 CIA-58 bioclipse: egonw * r9706 /bioclipse2/trunk/plugins/net.bioclipse.cdk.busine​ss/src/net/bioclipse/cdk/business/CDKManager.java: CDK does not define .mdl for MDLV2000, but Bioclipse2 does for the MDL molfile content type; this fixes that incompatibility by having Bioclipse test for that format hardcoded
13:40 * jonalv didn't think we where running tests any longer. I thought that was for post 2.0 like everything else... :/
13:40 egonw no, just the nightly service was postponed
13:40 olass jonalv: of course we are running tests!
13:41 olass how many times must I say this: we CAN run tests even if nightly is post 2.0!
13:41 olass we just need to do it manually
13:41 olass *sigh*
13:41 egonw s/CAN/MUST/
13:44 egonw damn... cannot get the plugin test to work
13:44 * jonalv get errors in Biojava when he tries to run that test...
13:44 egonw otherwise I could have tried to fix things
13:45 egonw yeah, Biojava is outdated :)
13:45 jonalv it sure is...
13:45 egonw but not outdated/
13:46 * jonalv can run the test now
13:46 jonalv 69 test, 27 erros and 2 failures
13:47 jonalv egonw: Check Run configurations and run only with the needed plugins seems to help...
13:48 jonalv it seems to me like something is very broken here...
13:49 * jonalv ducks and covers
13:49 egonw ok, will try that...
13:49 egonw maybe the default clean config does not do the job either...
13:49 egonw ah, sorry, should have known
13:50 egonw Eclipse default has all plugins in the workspace :(
13:50 egonw olass: sorry for thinking that deleteting the failing config == as doing the required plugins stuff
13:51 jonalv hm it's actually not as bad as it looks
13:51 egonw OK, *clearly* the CDKManager has *not* been converted to the NWO yet
13:51 jonalv what is not working is cdk.loadMolecule
13:51 jonalv which of course is bad...
13:51 jonalv *lol*
13:52 egonw ok, reading the wiki now
13:55 jonalv egonw: this test class is not up to date with New world order. I am sorry...
13:56 jonalv I also realised that the New World order is sort of borken but that Bioclipse is not using the broken parts except in the this test...
13:56 egonw jonalv: so, which tests should the test being testing?
13:57 jonalv egonw: sry?
13:57 egonw NWO is broken?
13:57 egonw well, the unit tests test the methods which used to be needed to be tested
13:57 egonw apparently the unit tests now need to test different method
13:57 jonalv egonw: well I am afraid the methods taking a String might not work so well on the gui flavoured manager. At least that is what it looks like to me from these tests
13:57 egonw moreover
13:58 egonw I have a strong feeling more is at work here...
13:58 jonalv they probably should work on that manager... aswell
13:59 jonalv Most of the failing tests fail because the calls to cdk.loadMolecule(String) is not redirected to the IFile version
14:00 egonw what is is BioclipseUIJob? Is that replacing IProgressMonitor?
14:00 jonalv egonw: no
14:00 jonalv however that method works from the JavaScript environemnt
14:00 egonw what method?
14:00 egonw and why does the exception say at lin 174
14:00 jonalv cdk.loadMolecule
14:01 egonw one should use loadMolecule(IFile, BioclipseUIJob)
14:01 jonalv egonw: what exception?
14:01 egonw and so does the method itself
14:01 egonw line 184
14:01 egonw that is
14:01 egonw in CDKManager at link 173 and 184
14:02 jonalv uhm the UnsupportedOperationException is thrown at line 183..
14:02 egonw an Exception is thrown that says the method loadMolecule(IFile, BioclipseUIJob) should be used
14:02 egonw yes, correct
14:02 egonw but why is method A throwing an exception saying that method A should be used??
14:02 egonw well...
14:03 egonw actually even worse...
14:03 jonalv egonw: no it is saying that it is method A that should not be used
14:03 egonw it is saying A' needs to be called
14:03 egonw que??
14:04 jonalv It is saying: This method should not be called and then it says which method this is refering to
14:04 egonw OK, please fix the messages
14:04 egonw they are wrong and confusing
14:04 jonalv egonw: I think they are good. Please explain what you want instead
14:04 egonw OK, look at them carefullly
14:04 egonw I now understand that they should say something like:
14:04 egonw this method A should not be used
14:04 egonw where is now says This manager method should not be called, S
14:04 egonw where is now says This manager method should not be called, A
14:04 egonw correct?
14:05 egonw now compare A with the actual method signature from where it is called
14:05 egonw and, it is, IMHO, more useful to actually say which method should be called instead
14:05 jonalv egonw: no
14:05 egonw to which of the three problems?
14:05 egonw my assumption?
14:06 egonw OK, please explain
14:06 jonalv egonw: it says which method should no be caleld so that we know whre to look when trying to fix the bug if that exception turns up on the JAvaScript console
14:06 egonw right, so A is the method that is reported to not to have been called, correct?
14:07 egonw that's what I assumed above
14:07 jonalv yes
14:07 egonw OK, then the messages *are* wrong
14:07 egonw and confusing
14:07 egonw and neither do the help the one who is supposed to fix it
14:07 egonw by not saying which method should be used instead
14:07 jonalv egonw: the messages are not for the end user but for the debugger. The end user can't do anything. If that excxeption turns up there isn't much to do...
14:07 egonw ad 2): yes, I think they are not just mildly confusing
14:08 egonw debugger == me
14:08 egonw I'm the responsible one for this manager
14:08 egonw and someone broken 29 unit tests
14:08 jonalv egonw: yes...
14:08 egonw without informing me
14:08 jonalv that was me
14:08 egonw :)
14:08 jonalv okey, I DID IT
14:08 egonw I'm not angry
14:08 jonalv happy now?
14:08 jonalv good
14:08 egonw no, just want to understand what I need to do to fix it
14:08 egonw and the exception messages are still wrong
14:09 egonw that you need to fix
14:09 egonw or should have done right in the first place
14:09 jonalv egonw: could we do it the other way around?
14:09 jonalv you fix the error messages and I try to figure out what to do to make tests passs?
14:09 egonw you complain to me on what is wrong about those exeptions?
14:10 jonalv egonw: I did not understand that sentence
14:10 jonalv ah
14:10 jonalv *LOL*
14:10 jonalv communicating over IRC really sucks...
14:10 egonw OK, I'll change them according to what I think why that code is broken
14:10 egonw and ask you to review that to see if you agree
14:11 jonalv egonw: oki, what would you want them to say?
14:11 egonw ok, 173:
14:11 egonw changeset 1
14:11 egonw "This manager method should not be called, " +
14:11 egonw "loadMolecules(IFile, BioclipseUIJob)");
14:11 egonw ->
14:12 egonw "The manager method 'loadMolecules(IFile, BioclipseUIJob)' should not be called." +
14:12 egonw secondly
14:12 egonw )                "This manager method should not be called, " +
14:12 egonw "loadMolecules(IFile, BioclipseUIJob)");
14:12 egonw ->
14:12 egonw + "loadMolecule(IFile)");
14:13 egonw to have it actually have the same method *and* signature as the method from which it is called
14:13 egonw thirdly:
14:13 egonw "The manager method 'loadMolecules(IFile, BioclipseUIJob)' should not be called. It should use the method Y instead"
14:13 egonw and you tell me what Y is
14:14 jonalv egonw: or scratch that about it should use instead and just say: "The AOP part has messed up"
14:14 egonw yes, very debugger friendly
14:14 egonw why not "File a bug report for developer Z" ?
14:15 jonalv are you serious?
14:15 jonalv yea why not...
14:15 egonw oh, you were?
14:15 jonalv I guess Z="jonalv"
14:15 egonw no idea
14:15 * jonalv is still serious yes
14:15 egonw I have not clue what you meant with AOP has messed up
14:16 jonalv egonw: that is our problem. I seem to be the only one...
14:16 jonalv anyways, my judgement is that we can make the release today and that fixing this test class is going to take some time
14:17 jonalv but I will get right on it
14:17 jonalv I must think though
14:17 jonalv and explore the errors
14:18 jonalv the test class is testing a bigger universe than what Biocipse is using. Not bad but not excatly great right now either.
14:18 jonalv Oh and Bioclispe could very well use that very same universe it's jsut that if people has followed the customs it doesn't
14:19 jonalv egonw, olass: Can you live with that?
14:19 egonw live with what?
14:20 jonalv releasing beta 4 although these test cases fails
14:20 egonw no, I cannot live with that
14:20 jonalv they are not testing the same things that are used in Bioclispe...
14:20 egonw because apparently code downstream is failing too
14:20 egonw such as, create2d/create3d
14:20 egonw and who knows what else
14:20 jonalv egonw: hold on
14:20 egonw I am not convinced they are not used...
14:21 jonalv egonw: create 2d and create 3d fails because of this?
14:21 * jonalv fires up Bioclipse and tests create 2d
14:21 egonw loadMolecule(IFile) is used at least 5 times
14:21 egonw loadMolecule(IFile, BioclipseUIJob) is called 4 times
14:21 egonw so, this is not limited to just the unit tests
14:22 jonalv create 3d works for me
14:22 * egonw is going for a strole... to clear his mind
14:23 jonalv egonw: coudl we talk?
14:23 egonw yes
14:23 jonalv do you have skype?
14:23 egonw mom, let me check
14:24 jonalv olass: are you there?
14:25 * jonalv can find nothing wrong with generate 2d and generate 3d. What must I do to see this bug?
14:28 olass jonalv: yes
14:28 olass but on phone
14:30 * jonalv is on skype with egon
14:31 CIA-58 bioclipse: ospjuth * r9707 /bioclipse2/trunk/plugins/net.bioclipse.statisti​cs/src/net/bioclipse/statistics/Activator.java: Removed bioclipse1 logging.
14:33 shk3 how are things going?
14:34 maclean joined #bioclipse
14:43 CIA-58 bioclipse: ospjuth * r9708 /bioclipse2/trunk/plugins/net.bioclipse.statistic​s.jama/src/net/bioclipse/plugins/Activator.java: Renamed Bc_Jama to Activator and adapted to 2.0, e.g. removed bc1 logging.
14:43 CIA-58 bioclipse: ospjuth * r9709 /bioclipse2/trunk/plugins/net.bioclipse​.statistics.jama/META-INF/MANIFEST.MF: Use new Activator.
14:44 CIA-58 bioclipse: ospjuth * r9710 /bioclipse2/trunk/plugins/net.bioc​lipse.statistics.jama/src/net/bioc​lipse/plugins/Bc_jamaPlugin.java: Renamed Bc_Jama to Activator and adapted to 2.0, e.g. removed bc1 logging.
14:48 egonw olass: ping
14:50 olass egonw: pong
14:50 jonalv olass: skype
14:51 olass is it important?
14:51 egonw yse
14:51 olass ok
14:51 jonalv olass: yes if you think the beta release is...
14:51 olass add me
14:57 shk3 egonw: success with the coordinate generation stuff?
15:10 egonw shk3: I fixed the .mdl issue as you asked
15:10 shk3 good
15:10 shk3 will update and test and close bugs
15:10 shk3 if working
15:13 olass egonw: attached file to bug #825
15:13 egonw thanx!
15:14 CIA-58 bioclipse: ospjuth * r9711 /bioclipse2/trunk/plugins/net.bioclipse.qsar.ui/sr​c/net/bioclipse/qsar/ui/builder/QSARBuilder.java: Do not generate CSV file if QSAR dataset is <2 rows or <2 columns
15:23 jonalv olass: I amtrying to get the Matrix editor to weok
15:24 olass it works here...
15:24 jonalv olass: I get a NullpointerExcpetion on row 86 in the Activator
15:24 egonw @pz matrix
15:24 jonalv basically it can't find the URL Platform.getBundle(PLUGIN_ID).getEntry("/" + LOG_PROPERTIES_FILE);
15:25 egonw zarah: ?
15:25 olass jonalv: please update
15:25 olass [16:46] < CIA-58> bioclipse: ospjuth * r9709 /bioclipse2/trunk/plugins/net.bioclipse​.statistics.jama/META-INF/MANIFEST.MF: Use new Activator.
15:25 jonalv I am a bit confunded as to what the configurelogger is doing...
15:25 jonalv olass: ah
15:25 olass [16:47] < CIA-58> bioclipse: ospjuth * r9710 /bioclipse2/trunk/plugins/net.bioc​lipse.statistics.jama/src/net/bioc​lipse/plugins/Bc_jamaPlugin.java: Renamed Bc_Jama to Activator and adapted to 2.0, e.g. removed bc1 logging.
15:25 zarah [831] Opening an incorrect CSV opens ... http://tinyurl.com/betckj
15:25 zarah Entire list at http://tinyurl.com/bs37l3
15:25 zarah egonw: yes?
15:28 * jonalv is still trying to get this matrix thing to work...
15:29 CIA-58 bioclipse: egonw * r9712 /bioclipse2/trunk/plugins/ne​t.bioclipse.cml/plugin.xml: Fixed name to be human readable
15:31 jonalv egonw, olass: I seem to have found an infinite loop in your MatrixREsource... :(
15:32 egonw line?
15:33 jonalv egonw: I will fine a bug repport. Do you want it or should I give it to Ola?
15:34 olass jonalv: you cannot solve it?
15:35 olass if you can, please do
15:35 jonalv olass: I found another problem which I can not wolve. I am moving towards managing to look at your...
15:35 jonalv olass: I didn't write the code. It seems complicated. String parsing usually is...
15:37 CIA-58 bioclipse: egonw * r9713 /bioclipse2/trunk/plugins/ne​t.bioclipse.jmol/plugin.xml: Also allow opening of 5D CML files in Jmol
15:38 jonalv now, how does a correct csv file look? :)
15:38 jonalv because what I thought was one clearly was not...
15:39 shk3 the net.bc.data.drugbank plugin contains a new wizard
15:39 shk3 which points to a class which does not exist.
15:39 shk3 is this intended?
15:39 shk3 some clever trick I overlooked?
15:40 egonw mom, let me check
15:42 maclean jonalv : it has values separated by commas?
15:42 maclean I think the CSV format is not well-defined.
15:43 jonalv maclean: yes but my file (which I thought had that) chrached Bioclipse totally (or rather frooze it up) so I am not so sure...
15:43 egonw shk3: which class are you missing?
15:43 maclean wikipedia has a nice page on CSV, actually.
15:43 egonw I think the class mentioned is from the .data plugin
15:44 maclean "No general standard specification for CSV exists. Variations between CSV implementations in different programs are quite common and can lead to interoperation difficulties."
15:44 shk3 net.bc.data.drugbank.wizard​s.NewDrugbankProjectWizard
15:44 maclean :(
15:44 jonalv lovely
15:44 egonw shk3: right, that from .data
15:44 egonw no secret trick
15:44 egonw it's just available from the classpath
15:45 egonw via plugin deps
15:45 egonw gone for dinner now
15:45 shk3 aha
15:55 olass jonalv: you want a csv that works in matrixeditor?
15:56 jonalv olass: I think I figured out how they need to look bow
15:56 olass bow?
15:56 jonalv yes how things look
15:57 jonalv I can't figure bow to do something about that Exception though
15:57 jonalv olass: right now I can't even catch it...
15:57 jonalv :(
15:58 jonalv seems something is swalloing it internally...
15:59 jonalv yup that's what's going on
15:59 jonalv olass: and it's in your MatrixREsource again
15:59 jonalv olass, egonw: I don't think I should change in your code...
16:00 olass yes please do
16:00 olass else it will not be fixed
16:00 olass I give you full permission :)
16:00 jonalv olass: no it's egons code as well. He get's mad if I do these sort of changes. besides I am not too fond of doing it in code which I don't know better...
16:01 olass jonalv: then it will not be fixed :(
16:01 jonalv olass: hang on I will see how wrote those lines... :)
16:01 olass jonalv: why chase after him?
16:01 jonalv olass: because he get's so mad if I don't
16:02 olass I give you full permission if I wrot ethe code
16:02 olass and if Eskuil wrote the code
16:02 olass if Egon wrote the code, assign to him
16:02 jonalv oh lol, last one to edit was a ceratin masak
16:03 jonalv I gues that was in the tab to spaces peration...
16:03 jonalv olass: is the Matrixresource used by anything else than the Matrixeditor?
16:03 olass no
16:03 jonalv okey then I will only introduce error in that...
16:03 jonalv okey here goes...
16:11 CIA-58 bioclipse: jonalv * r9714 /bioclipse2/trunk/plugins/net.biocli​pse.statistics/src/net/bioclipse/sta​tistics/model/MatrixResource.java:
16:11 CIA-58 bioclipse: Changed from swallowing exception and just printing them to STDOUT to re
16:11 CIA-58 bioclipse: throwing them wrapped in RuntimeException. This should fix bug 831 and hopefully
16:11 CIA-58 bioclipse: not introduce too much other errors since this code should not be used by much
16:11 CIA-58 bioclipse: other stuff according to olass++.
16:12 zarah bug #831 | http://tinyurl.com/betckj
16:14 olass jonalv++
16:14 jonalv olass: have you tried? :)
16:14 olass nope
16:14 jonalv lol
16:14 shk3 does the Drugbank sample data project work for you?
16:14 olass think so
16:14 olass mom
16:15 shk3 for me, it says "The selected wizard could not be started"
16:16 olass shk3: works like a charm
16:16 shk3 hm
16:17 * jonalv is leaving for today. cya
16:21 shk3 egonw: creating 3d coordinates now works also on *.mdl, but testBug826 in CDKManagerTest still fails.
16:21 shk3 could you have a look at that?
16:23 maclean egon's eating, I think.
16:26 egonw shk3: ping
16:26 egonw shk3: in CDKManagerTest or in CDKManagerPluginTest?
16:27 shk3 cdkmanagertest
16:28 shk3 the name of the method is misleading, i just found
16:28 shk3 it's more testing the saveMolecule
16:28 shk3 feel free to change
16:28 egonw 'who wrote that test?
16:28 egonw I think he/she forgot to do something
16:29 shk3 me
16:29 egonw can you please check the test
16:29 shk3 was in a hurry before the talk...
16:30 egonw at least lines 805-817 are clearly wrong :)
16:30 shk3 why?
16:30 shk3 yes
16:30 shk3 you are right
16:31 shk3 the test still fails because output is cml
16:31 shk3 at least on my machine
16:31 egonw yes, agreed
16:31 egonw it already fails on the first assert
16:32 shk3 that was the core issue behind that problem with coord generation
16:32 CIA-58 bioclipse: shk3 * r9715 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business.test/src/net/bioclipse/​cdk/business/test/CDKManagerTest.java: fixed the test
16:35 CIA-58 bioclipse: shk3 * r9716 /bioclipse2/trunk/plugins/net.bioclipse.c​dk.ui/src/net/bioclipse/cdk/ui/handlers/ (Create2dHandler.java Create3dHandler.java): resetting the other coordinate set to null, since in output, they might override each other
16:35 egonw shk3: what? why?
16:36 shk3 if a file has 2d and 3d and you write a mol file, it will always contain th 3d
16:37 shk3 so creating 2d coords on a mol file wiht 3d will not actually change the file.
16:37 shk3 I had this before
16:37 egonw but what if we want 5d?
16:37 shk3 hm
16:37 shk3 we can't have that in mdl
16:37 egonw true
16:37 shk3 so we would need to do it according to file format
16:37 egonw but that would suggest 2 action types
16:37 egonw one for CML files
16:37 egonw one for MDL files
16:37 shk3 well
16:38 shk3 yes
16:38 egonw yes, they could share code
16:38 egonw maybe even the same class
16:38 egonw not you cannot easily get the file format from the IMolecule
16:38 egonw can Action's have parameters?
16:38 shk3 no
16:39 egonw then maybe two action types is the simplest approach
16:39 shk3 yes
16:39 egonw now it breaks 5D generation in CML files
16:39 egonw or
16:39 shk3 that might have been the reason why somebody removed this
16:40 egonw have you checked which commit removed it?
16:40 egonw if so, which one?
16:40 egonw then we can see why it was done and who did it... (me?)
16:42 shk3 couldn't find it
16:42 shk3 do we handle other files than mol and cml
16:42 shk3 ?
16:42 shk3 molecule files, i mean?
16:43 egonw for reading yes, for writing not
16:43 egonw I found the problem why the unit test failes...
16:43 shk3 no, doesn't look like
16:43 egonw being a MockFile
16:43 egonw it does not do content type recognition
16:43 egonw moreover...
16:44 egonw it cannot detect the file format from the file name either (being done as back up)
16:44 egonw because MockFile always returns "Mock.mock" as name
16:45 shk3 hm
16:45 egonw so, I would suggest moving the test to the CDKManagerPluginTest
16:45 shk3 so still using the extension
16:45 egonw and load a molecule from the workspace
16:46 egonw so that content type detection is properly used
16:46 egonw and no IMockFile stuff
16:46 shk3 feel free
16:46 egonw shk3: that's a back up for a bug in OS/X
16:46 egonw feel free?
16:47 egonw I feel very busy
16:47 shk3 well, then I will do it
16:47 shk3 btw, I found net.bioclipse.contenttypes.mdlMolFile is based on org.eclipse.core.runtime.text
16:48 shk3 shouldn't it be based on net.bioclipse.cdk.ui.moleculeType
16:48 shk3 ?
16:48 egonw dunno what that last is...
16:48 egonw but first impresssion sounds like it is CDK oriented
16:48 egonw in which case I would say no
16:48 egonw that said...
16:49 egonw there is something to say about moleculeType and so...
16:49 egonw unfortunately, content types does not allow multiple inheritance
16:49 egonw so we cannot depend both on file format and on type type
16:49 shk3 hm ok
16:50 shk3 but would that be a case of multiple inheritance?
16:50 shk3 shouldn't all the molecule types inherit from that moleculeType?
16:50 shk3 hm, might be a problem with cml
16:50 shk3 that should inherit from this and from xml, probably
16:50 shk3 anyway
16:50 shk3 I will not think about this
16:51 shk3 right now
16:55 egonw right, CML is the situation where I realized the limitation of the Eclipse content type model
16:56 egonw actually... I should blog about that in Planet Eclipse
17:10 maclean to the blogmobile!
17:24 olass egonw: Gpox discovered how to get the inheritance model for contenttypes in JCPEditor
17:24 olass or was it moltable?
17:24 olass yes, moltable
17:24 olass to discover if input was SmilesType or SDFileType
17:24 olass maybe not what you were after but anyway wanted you to know about it
17:25 olass in short: he was able to get the inheritance structure of a content type programmatically
17:27 CIA-58 bioclipse: shk3 * r9717 /bioclipse2/trunk/plugins/net.bioclipse.cdk.ui/ (7 files in 2 dirs): All the possibilities are now properly handled with respect to reset. More classes, but cleaner code.
17:38 egonw olass: sounds interesting...
17:59 egonw joined #bioclipse
18:06 egonw joined #bioclipse
18:12 edrin left #bioclipse
19:33 edrin joined #bioclipse
19:57 egonw joined #bioclipse
20:34 edrin hi egonw
20:34 egonw hi edrin
21:03 edrin egonw: i added some text to the paragraph "The XMPP architecture is a good scaffold for achieving service federation"
21:04 edrin do you like it this way?
21:10 egonw I don't like the distinction of HTTP/HTTPS
21:10 egonw isn't HTTPS an extension of HTTP?
21:11 egonw XMPP has extensions...
21:11 egonw so why require encryption to be part of HTTP itself... it's not for XMPP either, or is it?
21:11 egonw also, I think HTTP has several extensions that support strong authentication
21:12 egonw way beyond IP addresses...
21:12 egonw HTTP server support username based login, LDAP and Shibboleth
21:12 egonw when you start to contrast this with XMPP
21:12 egonw you become fairly vague...
21:12 egonw "XMPP servers on the other hand can connect server-to-server and form networks of XMPP servers"
21:13 egonw for example
21:13 egonw is not fairly informative as XMPP services are primarily client-server communication, not?
21:13 egonw then
21:13 egonw restricting services to a dedicted domain sounds very much like picking a few IP addresses in the HTTP case
21:14 edrin hm
21:14 edrin yes
21:14 egonw in short
21:15 egonw the current text doesn't really convinde me XMPP is better than HTTP
21:15 edrin the difference is
21:16 edrin XMPP is like a Mc Donalds Super Menu, and HTTP-based services is like a super market where you must buy all ingredients yourself
21:16 edrin and cook a meal with it
21:17 edrin unfortunately I cant write it like this
21:17 egonw :)
21:17 egonw no, because no scientist will like McDonalds
21:18 egonw seriously...
21:18 edrin hm, there is always an exception...
21:18 egonw let's dive into this...
21:19 egonw XMPP servers can turn of encryption, right?
21:19 edrin they can
21:19 egonw XMPP servers can turn of support for encryption, right?
21:19 edrin yes
21:19 egonw so, its an extension
21:19 egonw any apache does HTTPS by a simple switch too
21:19 egonw AFAIK
21:19 edrin sure
21:19 edrin that's true
21:19 egonw so, no difference
21:20 edrin but it's a difference with authentication
21:20 egonw (and now I am hoping you say: but wait...)
21:20 egonw ah, there it is :)
21:20 egonw OK, let's try that...
21:20 egonw HTTP allows authentication in various ways: LDAP, Shibboleth, simple username/pwd...
21:20 egonw the service can simply pick what to use
21:21 egonw turnkey technology...
21:21 egonw hey, even I managed to do it :)
21:21 edrin that's true, too. you can even transmit divx videos via http and watch them with a player
21:22 egonw ah, even big files is no problem
21:22 edrin indeed
21:22 egonw no buts?
21:23 edrin well, the but is
21:23 edrin that there is no generally accepted way to do this
21:23 edrin if you would want to write a library for this...
21:23 egonw as in?
21:23 egonw the mentioned solutions are widely used
21:23 egonw that sounds like generally accepted
21:24 edrin give me a second
21:24 egonw sorry
21:25 edrin http://ws.apache.org/axis/java/secur​ity.html#IsSOAPFundamentallyInsecure
21:25 zarah edrin's link is also http://tinyurl.com/4pftc
21:26 edrin and then: http://ws.apache.org/axis/java/sec​urity.html#AuthenticatingTheCaller
21:26 zarah edrin's link is also http://tinyurl.com/4pftc
21:27 edrin egonw: do you see what I mean?
21:27 edrin there is a lot "does not (yet) support..."
21:28 egonw the first link I agree with the author: SOAP is as secure as the Apache server you use
21:28 edrin And I think with all the thousand parallel standads
21:28 edrin they will probably have to remove the (yet)
21:28 egonw this is about SOAP itself
21:28 egonw not HTTP
21:28 egonw SOAP deliberately does not security
21:28 egonw as that is done by the transport layer
21:28 edrin yes of course
21:29 egonw and HTTP does that more than sufficiently
21:29 edrin so, why is this not supported by Web Service libs?
21:29 edrin or is it?
21:30 egonw because it is in a different layer
21:30 egonw WS lib does the SOAP level
21:30 egonw Apache the HTTP level
21:30 edrin but that's what I mean
21:30 egonw security is done in the SOAP/HTTP at the transport level
21:30 edrin they end up in chaos
21:30 egonw no, sounds like clean design
21:30 egonw also
21:31 egonw xws4j does not security either
21:31 edrin with too many people involved
21:31 egonw it leaves that to the transport layer too
21:31 edrin sure
21:32 edrin egonw: the difference is: to get these secure SOAP services one needs to follow the advices of some web service working groups
21:32 edrin not sure if there are best practice rules for secure http soap services
21:33 edrin maybe there will be some
21:33 egonw in our world, they have to follow the advices of some XMPP working group
21:33 edrin when the ~1000 involved people agree on them
21:33 edrin egonw: yes
21:33 edrin there is one XMPP working group
21:33 edrin with http there is w3c
21:34 edrin the w w w consortium
21:34 egonw so, the argument is that XMPP has fewer people involved it is better?
21:34 edrin i would not go so far
21:34 edrin rather faster
21:35 edrin more flexible
21:35 edrin that's what scientists would need
21:35 edrin in my opinion
21:35 egonw can you back that up with some show case
21:35 edrin sure
21:35 egonw where it was shown that security of XMPP is better than that of HTTP?
21:35 edrin as you can see the w3c is not fast enough, thus the embrace defines "best practice rules"
21:36 edrin on the other hand embrace does not develop the soap libs
21:36 egonw I don't see it that way
21:37 egonw the W3C was fast enough
21:37 egonw but the library implementers like Apache fucked up
21:37 egonw ummm
21:37 egonw messed up
21:37 edrin meaning, while in principle w3c is responsible for defining stuff, embrace must define it themselve, and this is just a hack as they are bound to existing libs
21:37 egonw but that's not W3C's fault
21:37 edrin it's nobodies fault
21:38 edrin it's just slow
21:38 egonw and has nothing to do with size or amount of working groups
21:38 egonw what is slow?
21:38 egonw not W3C
21:38 egonw the axis project, CFX team have been slow with coming up with a proper implementation
21:38 edrin the mix of it together
21:38 egonw implementing the full spec
21:38 edrin yes
21:39 egonw and what does that have to do with security?
21:39 edrin that there is no generally accepted way to built up a generally accepted security environment for cloud services over http/soap
21:40 egonw what is wrong with having multiple standards? we are introducing a new one... just adding to the mess then?
21:40 edrin while in XMPP, there is a "you must do it this way"
21:41 egonw ok, so what happens if XMPP decides no on IO-DATA and chooses SOAP over XMPP?
21:41 egonw then it is exit IO-DATA?
21:41 edrin they should choose what is implemented
21:42 edrin at least that's how they think
21:42 edrin afaik
21:43 egonw so what happens if XMPP decides no on IO-DATA and chooses SOAP over XMPP?
21:44 edrin then we will end up in history of doomed?
21:45 egonw like a certina HTTP security protocol which was never generally accepted?
21:45 edrin yes
21:45 edrin the question is
21:46 edrin is it more likely that this happens when doing it with HTTP or with the XMPP?
21:47 edrin egonw: i think it is much easier to build a trusted environment with XMPP than with HTTP
21:47 egonw I don't feel much for putting nostradamical predictions in papers
21:47 egonw yes, that
21:48 egonw but it's not about what you think
21:48 edrin the military industries uses XMPP to make combat participants to communicate with each other
21:48 edrin i.e. EADS
21:48 egonw it's about what we can convince people with arguments and facts
21:48 egonw they also use dolphins
21:48 edrin the new "Quake live" uses XMPP authentication mechanisms
21:48 egonw yes, bioinformatics uses SOAP
21:48 edrin several computer games do such stuff
21:48 egonw these are not arguments
21:49 egonw the world used VHS, even though Betamax was better
21:49 egonw they use microsoft instead of linux
21:49 egonw popularity is not an argument we can use in a scientific paper
21:49 egonw only convincing arguments
21:50 egonw only convincing and factually true arguments
21:50 edrin yes
21:50 edrin so?
21:50 egonw please provide me arguments that show the XMPP has better security features than HTTP
21:50 egonw and generally accepted is not going to do it
21:51 egonw then we should just remove that section
21:51 egonw as we did not do it yet and therefore can't be part of results
21:51 edrin i dont think so
21:51 egonw and neither has any place in the discussion then
21:52 egonw we can only observe that XMPP provides methods for federation
21:52 egonw fine, so does SOAP
21:52 edrin yes
21:52 egonw that's not a discussion of features of the method
21:52 egonw and should move downwards to the outlook paragraphs
21:52 edrin SOAP provides methods for federation? yes, but nowhere or not possible to implement
21:53 edrin i can make a dolphine fly
21:53 egonw that will certainly not fit in this paper
21:53 edrin when putting it in a rockt
21:53 edrin ok
21:53 egonw so, come up with a good argumentation why it is impossible to implement
21:54 egonw that would be a valid argument
21:54 edrin egonw: to me implementing federation for http/soap just appears as a workaround hack
21:54 edrin in XMPP it's all standard
21:54 egonw btw, I just ran into SOAP over the Java Messaging Service
21:55 egonw no, sorry... that's popularity again
21:56 egonw look, we're fighting against the SOAP standard...
21:56 egonw just saying XMPP is more standard is not working
21:56 edrin ok, then just keep it like ola wrote it
21:57 edrin no?
21:57 egonw I don't know what he wrote
21:58 edrin now it's like before
22:03 egonw I have no clue what was there in between
22:09 egonw OK, going to bed now
22:11 edrin ciao

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