Camelia, the Perl 6 bug

IRC log for #bioclipse, 2008-09-01

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

All times shown according to UTC.

Time Nick Message
07:26 Gpox joined #bioclipse
07:27 rojasm joined #bioclipse
07:51 edrin joined #bioclipse
07:53 rojasm1 joined #bioclipse
08:03 jonalv joined #bioclipse
08:03 jonalv moin
08:03 zarah hi jonalv
08:05 jonalv @seen olas
08:05 zarah olas was last seen 2008-08-29T13:33:48 in #farmbio saying 'trefvlig helg på er!'
08:24 edrin hi
08:24 zarah hi edrin
09:04 egonw joined #bioclipse
09:14 edrin hi egon
09:18 egonw_ joined #bioclipse
09:27 egonw hi edrin
09:27 egonw edrin: did I reply? I haven't, have I?
09:36 Gpoks joined #bioclipse
09:59 thomas_ku joined #bioclipse
10:11 CIA-30 bioclipse: egonw * r6827 /bioclipse/trunk/plugins/ (2 files in 2 dirs): Makes it work with Eclipse 3.4
10:15 egonw jonalv: ping
10:34 edrin egonw: no, i think you did not yet
10:35 egonw edrin: ack, sorry about tha
11:03 olas joined #bioclipse
11:03 olas moin
11:03 zarah hi olas
11:06 edrin hi olas
11:09 olas hi edrin
11:09 olas edrin: I've seen your progress on XMPP
11:09 olas great job!
11:10 olas edrin++
11:10 olas no time this week, hope fully I can get some work doner on it next week
11:10 egonw lunch time
11:10 egonw bbl
11:19 edrin thanks
11:20 edrin olas: it's now possible to do function.getIoSchemata()
11:21 edrin although right now users would have to do the XML Schema 2 Java Class conversions on their own
11:22 edrin olas: I will be on holidays the next week and the week after and on a congress the last week of september...
11:29 jonalv egonw, pong
11:44 egonw back
11:44 jonalv egonw, wb
11:44 egonw edrin: the design requires the client implementors to support the various kinds...
11:44 egonw don't see how we can change that without changing the XEP
11:46 jonalv egonw, you pang :)
11:54 egonw jonalv: did you solve the CML lib problem?
11:56 jonalv egonw: no not really
11:57 jonalv egonw, did you see the bug report?
11:58 olas edrin: I think it's fair that users can do their own XSD 2 java conversion
11:58 olas I'd like to use EMF and maybe someone else would like to use some other technique
11:59 jonalv egonw, http://sourceforge.net/tracker/inde​x.php?func=detail&aid=2078045&a​mp;group_id=51361&atid=463005
12:06 egonw ok, noted
12:07 jonalv egonw, can you help? :)
12:08 egonw I hope so
12:08 egonw sitting down here with Jim now
12:08 jonalv sounds good
12:16 jonalv egonw, did you just comment on the bug?
12:17 egonw yes
12:17 egonw I note that there actually are two jumbo jars in BC2
12:17 egonw one in the CDK project
12:17 egonw which is the 'new' one...
12:17 jonalv did not the stack traces at the pastebin work?
12:18 egonw and an older in net.bioclipse.cml
12:18 egonw sorry about that.... overlooked those links
12:18 egonw but better indeed put them inline or as attachment
12:18 olas net.bc.cml is not in production
12:18 jonalv they where crazy long.. :)
12:18 olas hwncw CDK does not depend on it
12:19 olas hence
12:20 egonw jonalv: ok, this is waht I thikn goes wrong
12:20 * jonalv is listening
12:22 egonw for one reason or another...
12:22 egonw sorry, distracted by the cool ubiquity scripts in my blog
12:22 egonw the code is searching the types.xsd in the jumbo5.2.jar in net.bioclipse.cml
12:23 egonw which got for some reason in your case precedence over the jumbo jar in the cdk plugin
12:23 jonalv But that thread has the wrong classloader
12:23 egonw I can confirm that the jumbo5.3 jar indeed does not have this file
12:23 jonalv the thread was created in some other plugin at some other place...
12:23 egonw whereas the cdk jumbo jar does
12:23 egonw yes, agreed...
12:23 egonw but it's the only reasonable explanation left
12:24 egonw that something got fucked up with the classloading
12:24 egonw and, honoustly...
12:24 egonw with all the CDK versions, jumbo, version, managers, spring, osgi...
12:24 egonw I'm not bewildered to see problems like this eitehr
12:24 jonalv yea Eclispe uses multiple classloaders so you can't just go around beliieving that theres only one and that you can get it from any thread...
12:25 jonalv yea it's an osgi thing... :)
12:26 jonalv as far as I hav understtoo it anyway...
12:28 egonw so, let's drop the two jumbo jars in favor of only one...
12:31 jonalv egonw, sure but I doubt that is related to my problems...
12:31 jonalv although it might be..
12:57 egonw jonalv: ok, going to try a quick ahck
12:57 egonw going to put the types.xsd in the nb.cml jar
12:57 egonw jonalv: I need to you test it afterwards...
12:57 jonalv egonw, have you been able to reproduce my error?
12:58 egonw no
12:58 jonalv ah okey
12:58 egonw maybe I got a better idea
12:58 egonw I'll try to copy the CDK version into net.bioclipse.cml
12:58 jonalv hu wasn't that not used?
13:00 egonw no
13:00 egonw trying to make that point
13:01 jonalv sry?
13:03 egonw net.bioclipse.cml is using an ancient CML lib
13:05 jonalv egonw, I am not even running that plugin but I still get that error
13:05 egonw well, cdk10 has the old version too
13:05 jonalv k
13:09 egonw ok, I'll upload a new net.bioclipse.cml
13:10 egonw please try .cdk *without* .cdk10 as a plugin too
13:10 egonw commit pending
13:11 jonalv egonw, do you want me to try something? (I am working on brunn today so I am not 100% into this...)
13:11 egonw yes
13:11 jonalv egonw, I only had that problem from structuredb
13:11 egonw please update with the upcoming commit
13:11 egonw and then try to reproduce the bug
13:11 jonalv or hm perhaps some cdk method too
13:11 egonw well, not necessarily now
13:12 egonw but would be helpful, as we could further explore thing if you could
13:12 jonalv but as I told you I am not even runnign the cml plugin...
13:12 egonw then just the .cdk
13:12 egonw no .cdk10, no .cml
13:13 jonalv I think the cdk tests depends on cdk10 so that might be tricky...
13:13 egonw which cdk tests?
13:13 egonw of .cdk ??
13:13 jonalv cdk.business
13:13 egonw no, they don't depend on .cdk10
13:13 egonw cdk.business is around .cdk, not
13:14 egonw ?
13:14 egonw mom
13:14 egonw ok, well... there's the source of your problems
13:15 egonw OK, so who can we blame for that?
13:15 * egonw runs 'svn annotate'
13:17 CIA-30 bioclipse: egonw * r6828 /bioclipse2/trunk/net.bioc​lipse.cdk.business.test/: Share project "net.bioclipse.cdk.business.test" into "https://bioclipse.svn.sourceforge​.net/svnroot/bioclipse/bioclipse2"
13:20 egonw WTF??
13:20 egonw sorry about that commit...
13:20 egonw no idea what eclipse has just been doing...
13:20 jonalv lol
13:33 egonw ok, here's the commit finally getting along, I think
13:33 jonalv good
13:34 jonalv I am just gonna start a Brunn build :)
13:34 egonw it's commiting now
13:35 CIA-30 bioclipse: egonw * r6829 /bioclipse2/trunk/plugins/net.bioclipse.cml/ (5 files in 4 dirs): Upgraded to the Jumbo version used by .cdk
13:35 jonalv brb
13:37 jonalv b
13:51 olas egonw: around?
13:52 olas is this the preferred way of depicting sybyl atom types for a molecule?
13:52 olas http://paste2.org/p/68611
13:53 olas no convenience method for an atomcontainer?
13:58 egonw mom, checking
14:00 egonw what do you mean with 'depict'?
14:15 olas infer
14:15 olas I have an AC and want to have atom types inferred
14:15 olas on all atoms
14:21 jonalv joined #bioclipse
14:21 jonalv hi
14:21 zarah hi jonalv
14:34 olas hi jonalv
14:34 jonalv hi olas
14:42 masak joined #bioclipse
15:19 olas left #bioclipse
15:24 shk3 joined #bioclipse
15:26 maclean joined #bioclipse
15:39 jonalv joined #bioclipse
15:51 jonalv shk3: is net.bioclipse.spectrum supposed to compile at current revision? (because it doesn't)
15:52 shk3 it does for me
15:53 shk3 which error do you get?
15:55 jonalv net.bioclipse.spectrum.editor.SpectrumEditor == The import org.xmlcml cannot be resolved
15:56 maclean Could this be related to changes egon was making to the cml plugin today?
15:56 jonalv very possibly so yea
15:57 shk3 probably
15:57 shk3 I will update and fix
15:57 jonalv good
15:57 shk3 packe names in jumbo changed
15:57 shk3 that should be the reason
15:57 * jonalv likes when everyhting compiles :)
15:59 maclean It's better than when everything breaks, that's the honest truth.
16:04 rojasm1 left #bioclipse
16:33 edrin left #bioclipse
16:36 CIA-30 bioclipse: shk3 * r6830 /bioclipse2/trunk/plugins/net.bioclipse.cml/ (.classpath META-INF/MANIFEST.MF build.properties): made egons changes to actually work
16:55 CIA-30 bioclipse: shk3 * r6831 /bioclipse2/trunk/plugins/net.bioclipse.spectrum/ (6 files in 5 dirs): using new jumbo
17:08 rojasm joined #bioclipse
17:18 olas joined #bioclipse
17:25 CIA-30 bioclipse: ospjuth * r6832 /bioclipse2/trunk/plugins/net.bioclipse.cd​k.business/src/net/bioclipse/cdk/business/ (CDKManager.java ICDKManager.java): Implemented Sybyl atom typing and test for sybyl atom typing + mol2 saving. Tests fails, see bug 2087218.
17:25 olas @tell egonw: Urgent bug: http://sourceforge.net/tracker/inde​x.php?func=detail&aid=2087218&a​mp;group_id=150681&atid=778609
17:25 zarah Consider it noted.
17:28 CIA-30 bioclipse: ospjuth * r6833 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business.test/src/net/bioclipse/​cdk/business/test/CDKManagerTest.java: Added Sybyl and Mol2 serialization tests. Bug 2087218.
17:31 olas left #bioclipse
17:31 olas joined #bioclipse
17:58 CIA-30 bioclipse: shk3 * r6834 /bioclipse2/trunk/plugins/n​et.bioclipse.spectrum/src/ (5 files in 2 dirs): bugfixes in metadata editing
20:13 olas @msg
20:13 zarah olas: You have no new messages.
20:13 olas left #bioclipse
21:48 m4r1in joined #bioclipse

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