Camelia, the Perl 6 bug

IRC log for #cdk, 2007-11-07

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

All times shown according to UTC.

Time Nick Message
05:15 CIA-16 jmol: hansonr * r8575 /trunk/Jmol/src/org/jmol/ (21 files in 7 dirs):
05:15 CIA-16 jmol: version=11.3.42_dev ramachandran/quaternion/frameTitle
05:15 CIA-16 jmol: # bug fix: ramachandran/quaternion fix for assignment of structure
05:15 CIA-16 jmol: # bug fix: isosurfaces not taking their default name properly
05:15 CIA-16 jmol: # new feature: frame title "this a title for THIS frame only"
06:32 CIA-16 jmol: hansonr * r8576 /trunk/Jmol/test/org/jmol/api/TestScripts.java: test update
06:33 CIA-16 jmol: hansonr * r8577 /trunk/Jmol-datafiles/tests/scripts/ (2 files in 2 dirs): test upate
06:34 CIA-16 jmol: hansonr * r8578 /trunk/Jmol/src/org/jmol/viewer/Jmol.properties: (log message trimmed)
06:34 CIA-16 jmol: version=11.3.42 VERSION
06:34 CIA-16 jmol: # bug fix: ramachandran/quaternion fix for assignment of structure
06:34 CIA-16 jmol: # bug fix: isosurfaces not taking their default name properly
06:34 CIA-16 jmol: # new feature: frame title "this a title for THIS frame only"
06:34 CIA-16 jmol: # bug fix: small fix for povray cap at back side
06:34 CIA-16 jmol: # bug fix: overly long titles now again substringed in the popup menu.
07:56 masak joined #cdk
07:57 CIA-16 cdk: miguelrojasch * r9389 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (15 files in 8 dirs): Different classes to handle molecular formula concept
08:26 jonalv joined #cdk
08:40 CIA-16 bioclipse: jonalv * r4170 /trunk/playground/plugins/net.bioclipse.usermanage​r/src/net/bioclipse/usermanager/UserManager.java: last(at least for now) usermanager test passes
08:43 CIA-16 bioclipse: carl_masak * r4171 /trunk/plugins/net.bioclipse.rhino/​src/net/bioclipse/plugins/bc_rhino/ (2 files in 2 dirs):
08:43 CIA-16 bioclipse: [RhinoConsoleView.java]
08:43 CIA-16 bioclipse: * restored the variable initialization
08:43 CIA-16 bioclipse: [RunRhinoScriptInThreadAction.java]
08:43 CIA-16 bioclipse: * made installScriptingTools() public
09:49 jbrefort joined #cdk
10:06 CIA-16 cdk: egonw * r9390 /trunk/cdk/src/org/openscience/cdk/ (10 files in 5 dirs): Reverted patch by Miguel intended(?) for his branch, and Rajarshi's work on getting it compiling again (commits 9386-9388)
10:07 CIA-16 cdk: miguelrojasch * r9391 /branches/miguelrojasch/mf/src/org/op​enscience/cdk/MolecularFormula.java: The IMolecularFormula don't accept add IAtomContainer but yes IMolecularFormula. That should be work of MolecularFormulaManipulator. Put again some corrections of rajarshi
10:22 CIA-16 bioclipse: carl_masak * r4173 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
10:22 CIA-16 bioclipse: [RhinoConsoleView.java]
10:22 CIA-16 bioclipse: * basic command history
10:23 CIA-16 cdk: miguelrojasch * r9392 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (5 files in 3 dirs): Clean up JavaDocs
10:27 CIA-16 bioclipse: jonalv * r4174 /trunk/playground/plugins/ (3 files in 3 dirs): Commit before attempt with rewriting the HsqldbUtil into stop current server and create new server each time new database is needed. Nothing works now and this (although it feels wrong) seems like the only way to go...
10:46 CIA-16 bioclipse: jonalv * r4175 /trunk/playground/plugins/net.bioclipse.hsql​db/src/net/bioclipse/hsqldb/HsqldbUtil.java: Creating a new server when creating a new database seems to work although it feels wrong
10:49 CIA-16 bioclipse: carl_masak * r4176 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
10:49 CIA-16 bioclipse: [RhinoConsolveView.java]
10:49 CIA-16 bioclipse: * separated out key bindings into hash table
11:02 CIA-16 cdk: miguelrojasch * r9393 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (3 files in 2 dirs): New tests
13:22 CIA-16 cdk: miguelrojasch * r9394 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (2 files in 2 dirs):
13:22 CIA-16 cdk: Class to manipulate the MolecularFormula. I has the function to convert from
13:22 CIA-16 cdk: IAtomContainer to IMolecularFomula or backwards. Furthermore, it extracts the
13:22 CIA-16 cdk: string. I should supply the MFVAnalyser which contains a lot of methods. And
13:22 CIA-16 cdk: respective tests.
13:33 CIA-16 bioclipse: jonalv * r4177 /trunk/playground/plugins/net.bioclipse.hsql​db/src/net/bioclipse/hsqldb/HsqldbUtil.java: removed debug System.out.println
13:53 CIA-16 bioclipse: carl_masak * r4178 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
13:53 CIA-16 bioclipse: [RhinoConsoleView.java]
13:53 CIA-16 bioclipse: * factored out console prompt into variable
13:53 CIA-16 bioclipse: * console prompt now un-erasable
13:53 CIA-16 bioclipse: * also impossible to navigate out of command line using arrow keys
13:58 CIA-16 bioclipse: carl_masak * r4179 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
13:58 CIA-16 bioclipse: [RhinoConsoleView.java]
13:58 CIA-16 bioclipse: * beautified error messages as per suggestion from jonalv++
14:04 CIA-16 cdk: miguelrojasch * r9395 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (2 files in 2 dirs): Manipulator of MolecularFormulaSet. Added two methods for extract the maximim and minimum occurrence of IElements from a set of IMolecularFormula. And tests.
14:05 jpansanel hi
14:05 egonw hi jpansanel
14:06 jbrefort hi all
14:16 CIA-16 bioclipse: carl_masak * r4180 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
14:16 CIA-16 bioclipse: [RhinoConsoleView.java]
14:16 CIA-16 bioclipse: * copying (with Ctrl+C or Cmd+C) now works
15:05 CIA-16 bioclipse: carl_masak * r4181 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
15:05 CIA-16 bioclipse: [RhinoConsoleView.java]
15:05 CIA-16 bioclipse: * context menu with cut/copy/paste
15:45 hyalu joined #cdk
15:45 hyalu hi
15:49 egonw hi hyalu
15:53 CIA-16 cdk: rajarshi * r9396 /trunk/cdk/src/org/openscience/cdk​/atomtype/CDKAtomTypeMatcher.java: Fixed a bug which was checking that an atom had hyb unset as well as being SP2 which is impossible! Fixes a number of failing smarts parser tests
15:53 hyalu hi
15:53 CIA-16 bioclipse: carl_masak * r4182 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
15:53 CIA-16 bioclipse: [RhinoConsoleView.java]
15:53 CIA-16 bioclipse: * fixed up-arrow-when-caret-elsewhere bug
15:54 hyalu i sw your email about the 'continous' nightly build
15:54 hyalu good idea, but will take some coding
15:54 hyalu possible done in the next few weeks
15:55 hyalu is their an XML feed for CIA?
15:55 egonw yeah, no hurry...
15:55 egonw was just something I had been thinking about a couple of times now
15:55 egonw yes, I think so...
15:55 egonw mom
15:56 hyalu btw, regarding the issue of N in N-pyridine oxide  -it seems that it'd be safe to ignore the O=n case and just recognize the [O-]=[n+] case
15:56 egonw http://cia.vc/stats/project/cdk/.rss
15:56 egonw yes
15:56 egonw no
15:56 egonw you mean [O-]-[n+] correct?
15:56 egonw that is ... single bond
15:56 hyalu oh, right
15:57 egonw yes, SMILES and most other format allow input of inaccurate chemistry...
15:57 egonw that's one advantage of this new atom type perception...
15:57 hyalu true
15:57 egonw run that first on your input prior to anything else (e.g. QSAR) and see if you coded your molecules up correctly
15:58 hyalu right
15:58 hyalu the desc calc no catches lots of weird/wrong cases
15:58 egonw indeed
15:59 hyalu but the atom typing is nice now. Just 5% failures and that too the high at num eleemnts mainly
16:00 egonw right... not bad at all
16:00 hyalu regarding continous builds
16:00 hyalu were you think triggering a build everytime a committ occurs?
16:00 egonw no...
16:00 hyalu or only for committs with say a keyword (nightly) in them?
16:01 egonw I was thinking that (after 12 hours) Nightly would run "svn log -r LATESTCOMMITFORLASTRUN:HEAD trunk/" to see if anything changed once per hour
16:01 egonw the use of "-r LATESTCOMMITFORLASTRUN:HEAD" makes it a fairly bandwidth friendly call
16:02 egonw and running it on just trunk/ is important
16:02 egonw so, I would suggest to create a new crontab to run once every hour that does:
16:02 egonw 1. has it already been 12 hours? No, then stop
16:02 hyalu I was thinking: pull the CIA RSS feed, get all items since the last run, if any commit explicitly include the keywrod then run, other wise wait till 12 hours have passed
16:03 egonw 2. check the log for trunk
16:03 egonw 3. if found a new commit, then run Nightly
16:03 egonw the RSS feed contains too few items...
16:03 egonw you might miss interesting commits...
16:03 hyalu ??
16:03 egonw e.g. the Bioclipse commits are in there as well...
16:03 hyalu that's easily filterable
16:04 egonw use http://cia.vc/stats/project/cdk/cdk
16:04 egonw that's just teh CDK commits...
16:04 egonw but even then...
16:04 egonw it only lists 20-30 items or so...
16:04 hyalu hmm
16:04 hyalu true
16:04 egonw better use the log option, I think...
16:04 hyalu I would like to acvoid triggering builds every hour
16:05 hyalu when we have a large load on that machine the build can take upto an hour
16:05 hyalu rather, if you think a commit needs a rebuild, add a keyword to the descriptionm
16:05 egonw no, it will not build every hour
16:05 egonw it will just check if there are changes every hour...
16:05 hyalu and if there are changes every hour?
16:05 egonw but it will not rebuild more often than every 12 hours
16:05 hyalu aah, got it
16:05 egonw [2007-11-07 17:03:03] <egonw> 1. has it already been 12 hours? No, then stop
16:06 hyalu so how does it differ from what we have now?
16:06 hyalu well it does differ
16:06 hyalu but not very significantly
16:06 egonw because it will not rebuild if there have been no changes...
16:06 egonw major difference...
16:06 egonw so if nothing happens in trunk/ for 3 days, it will not rebuild for 3 days
16:07 hyalu aah, ok
16:07 egonw because the results will be accurate anyway
16:08 hyalu hmm, should be easy to implement
16:09 hyalu more excuse to not write a paper :)
16:09 egonw hehehe
16:11 CIA-16 cdk: egonw * r9397 /branches/egonw/remove-non-cdkatom​type-code/src/org/openscience/cdk/ (6 files in 6 dirs): Synched with trunk/ r9396
16:12 egonw ok, me is rerunning JUnit on his branch...
16:12 egonw to see how much fewer failing tests he has :)
16:13 hyalu :)
16:14 egonw oh, btw... would like to add the time it needs for running the tests to the output table too
16:14 egonw because I'm hoping that performance goes up, with the new code
16:15 hyalu ok, can you put in a feature request
16:15 egonw something else... do you have a list of CDK-based SOAP/BioMoby webservices that run at IU?
16:15 hyalu one sec
16:16 hyalu http://rguha.ath.cx/~rguha/​cicc/wsrss/wsdlrss/getFeed
16:16 hyalu each RSS item contains the link to the WSDL
16:16 hyalu for a service
16:16 hyalu also includes some non-IU services
16:17 hyalu actually, use http://www.chembiogrid.org/ch​eminfo/wsrss/wsdlrss/getFeed
16:18 egonw OK, thanx
16:18 egonw OK, nightly reports 99+22 JUnit fails
16:19 egonw my branch has: 75+20
16:19 hyalu voerall?
16:19 hyalu oh right
16:19 hyalu the fix I put a few minuites back will reduce the smarts fails
16:19 egonw by how much?
16:19 hyalu to 20 I think
16:20 egonw 99+20 ?
16:20 hyalu it should go down to 80+22
16:20 egonw got that one applied in my branch
16:20 egonw ah, ok
16:20 egonw so, my branch has 5+2 extra fixes...
16:20 hyalu maybe
16:20 hyalu the fix I commited probably affects other things apart from SMARTS
16:20 egonw which includes a couple of failing HueckelAromaticity tests...
16:21 hyalu hmm, let me rerun nightly
16:25 egonw oh, btw... your commits still contains formating changes
16:25 hyalu btw, if you look at the SMARTS module results you'll see that the two errors are due to ring finding timeouts
16:25 hyalu damns :(
16:26 hyalu different machine, settings did'nt transfer
16:26 hyalu do those tests run fine locally? I can run them with no error. But nightly doesn't seem to work on those tests
16:27 hyalu regarding formatting changes, which commit? CDKAtomTypeMatcher seems OK
16:28 egonw yes, that one was fine
16:29 egonw not sure...
16:29 egonw I often have those kind of problems in Eclipse...
16:29 egonw seems that a lot of garbage collections is being done...
16:29 egonw which makes it pass the time out
16:30 hyalu hmm
16:30 hyalu well that explains the 2 errors
16:35 egonw grmph...
16:35 egonw adding a lot of AtomContainerManipulator.perci​eveAtomTypesAndConfigerAtoms() calls to calculate() methods...
16:35 hyalu btw, regarding the annotation based tests, did we decide whether we wanted @estClass and @TestMethod
16:35 egonw ... caching... much .... nee.... ded...
16:35 hyalu :)
16:35 egonw no, we did not decide yet
16:36 hyalu it'll help clean up test coverage issues (e.g. ConformerContainer)
16:36 egonw so, what is the decision to make?
16:36 egonw use it or not?
16:39 CIA-16 bioclipse: carl_masak * r4183 /trunk/plugins/net.bioclipse.rhino/src/net/biocli​pse/plugins/bc_rhino/views/RhinoConsoleView.java:
16:39 CIA-16 bioclipse: [RhinoConsoleView.java]
16:39 CIA-16 bioclipse: * text field now scrolls with caret after long output
16:43 egonw OK, added a *lot of* missing atom perception calls...
16:50 egonw OK, *now* I am at 84 + 26 problems :(
16:50 egonw smarts is 29:3
16:53 * egonw is running tests for trunk too now...
16:53 egonw and will issue the following commands after that:
16:54 egonw cd cdk/
16:55 egonw grep FAILED branches/egonw/remove-non-cd​katomtype-code/reports/*.txt > branch.results
16:55 egonw grep FAILED trunk/cdk/reports/*.txt > trunk.results
16:55 egonw diff -u trunk.results branch.results
16:58 hyalu nightly is showing:  smarts80120397.13
16:58 egonw huh?
16:58 egonw how do I read that?
16:58 hyalu http://cheminfo.informatics.indiana.edu/~​rguha/code/java/nightly/junitsummary.html
16:58 egonw 20+3 in 97secs?
16:58 hyalu be back in a sec
16:59 egonw OK, 88 + 20 overall
17:00 egonw trunk versus my branch:
17:00 egonw 16 fixes
17:00 egonw 12 new fails :)
17:01 egonw cannot really match that with 88+20 versus 84+26 though...
17:01 egonw 16 fixes+12 new = 4 less...
17:01 egonw but overall stats indicate 2 more :(
17:08 CIA-16 bioclipse: jonalv * r4184 /trunk/playground/plugins/ (24 files in 5 dirs): More work on getting the databases view to work. Now there is a dialog for creating new structuredb. Also listener system for the database treeview. So far nothing very much fun to show...
17:11 hyalu ok, i've totally lost track of whats going on!
17:12 egonw about what?
17:12 hyalu junit failures and errors
17:13 egonw ok, so where did you loose track
17:13 hyalu what was the num errors before the last run?
17:14 egonw not sure...
17:14 egonw should be 88+20 now...
17:14 egonw was higher I think
17:14 hyalu yes it its
17:14 hyalu is
17:14 egonw [2007-11-07 17:19:05] <egonw> OK, nightly reports 99+22 JUnit fails
17:14 hyalu aah
17:14 hyalu ok, got it
17:15 egonw OK, the final results for my branch:
17:15 egonw 14 fixes, 16 new fails :)
17:15 hyalu so regarding the test annotations the question basically was: if we have @TestClass and @TestMethod, we need only have the method namein the latter annotation (rather than class+method as it is currently now)
17:15 hyalu where are the failures?
17:16 egonw http://cdk.pastebin.org/7331
17:18 hyalu hmm, it'd be good to fix the XlogP code - that's been failing for quite some time
17:18 hyalu smarts parser tests are all related N-pyridine oxide
17:20 egonw ack
17:20 egonw interesting...
17:28 egonw hyalu: http://chem-bla-ics.blogspot.com/2007/11/​comparing-junit-test-results-between.html
17:28 CIA-16 cdk: rajarshi * r9398 /trunk/cdk/src/org/openscience/cdk/qsar/descrip​tors/molecular/PetitjeanNumberDescriptor.java: Fixed a bug which wsa casting double to int and so loosing data.
17:28 hyalu nice
17:30 hyalu nightly just went to 89+19 :(
17:31 egonw one error became a fail?
17:31 hyalu yes, in RecursiveTest.java
17:32 egonw :)
17:32 hyalu one sec
17:32 egonw mmm... maybe you can use that code I just commited to show the JUnit changes between the last nightly run, and the one time earlier?
17:32 egonw got to run now...
17:32 egonw to catch the train home
17:32 hyalu ok, ttyl
17:45 CIA-16 cdk: rajarshi * r9400 /trunk/cdk/src/data/smiles/drugs.smi: Got rid of a bad smiles
17:46 CIA-16 cdk: rajarshi * r9401 /trunk/cdk/src/org/openscience/cdk/test/​smiles/smarts/parser/RecursiveTest.java: Added some more tests and separate a number of tests into individual test cases. I think most errros are due to aromaticity or atom typing, but not sure
18:52 CIA-16 cdk: shk3 * r9402 /branches/cdk-1.0.x/src/org/openscie​nce/cdk/test/io/MDLReaderTest.java: added a test case for bug #1826577
18:52 CIA-16 cdk: shk3 * r9403 /branches/cdk-1.0.x/src/data/mdl/hisotopes.mol: added a test case for bug #1826577
19:08 CIA-16 cdk: miguelrojasch * r9404 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (2 files in 2 dirs): new method to extract the exact mass and abundance of the molecularFormula
19:32 dleidert joined #cdk
19:51 CIA-16 cdk: rajarshi * r9405 /trunk/cdk/tools/nightly.py: Updated to report on the number of failures fixed and number of new failures since the last time the nightly was run
20:04 CIA-16 cdk: miguelrojasch * r9406 /branches/miguelrojasch/mf​/src/org/openscience/cdk/ (2 files in 2 dirs): added class Restriction into the method MassToFormulaTool. And tests.
20:05 edrin joined #cdk
20:15 dleidert_away joined #cdk
20:23 CIA-16 cdk: shk3 * r9407 /trunk/cdk/src/ (2 files in 2 dirs): added a test case for bug #1826577
20:47 rojasm joined #cdk
23:46 rojasm left #cdk

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