Camelia, the Perl 6 bug

IRC log for #bioclipse, 2007-12-18

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

All times shown according to UTC.

Time Nick Message
08:20 olas joined #bioclipse
08:21 olas egonw: how did the testing go?
08:30 steinbeck joined #bioclipse
08:45 _thomas_k joined #bioclipse
08:46 thomas_ku joined #bioclipse
09:02 egonw olas: bad...
09:03 egonw went to bad early...
09:03 egonw so no results at all
09:03 egonw will get some hot chocolate now... and test then
09:03 egonw brb
09:11 masak joined #bioclipse
09:19 egonw olas: you could have versioned the features 1.2.0.20071217
09:19 egonw like Eclipse does...
09:30 olas yes
09:30 olas I know
09:30 olas will do that in the future
09:30 masak ah, the pains of versioning...
09:32 egonw OK, the 'molecule from SMILES' tutorial is working lovely
09:32 egonw one down
09:35 egonw ok, the pubchem-search tutorial is working too
09:41 masak olas: the adress is eternal at http://bioclipse.blogspot.com/20​07/12/first-stab-at-logging.html
09:41 masak but as you see, "logging" is now "echoing"
09:44 egonw I'd like to remove 3D coord generation...
09:44 egonw not stable enough when combined with the JCP editor...
09:44 egonw and does not work for rings anyway...
09:45 egonw we should use the C++ smi23d code instead, I guess
09:45 egonw but not in 1.2.0
09:45 egonw for the rest: GO
09:45 egonw will remove that action right now... and update the tutorial
10:07 olas egonw: great
10:08 olas Will release 1.2.0 in teh afternoon
10:08 egonw thre is lots to improve... just like with 1.0
10:08 egonw but that's just the current state
10:08 olas as always
10:08 egonw but with the addition of the QSAR feature...
10:08 olas would be good to write a list
10:08 egonw oh wait... where did that tutorial go?
10:08 olas on the wiki
10:08 olas QSAR tutorial?
10:08 olas please check
10:09 olas and also: please update
10:09 egonw yes, will do... hang on
10:09 olas you can now visualize in chart
10:09 egonw I know
10:09 olas and select things in chart and matrix
10:09 egonw that I did not know
10:09 olas a late fix by EskilA
10:09 egonw mmm... the cheat sheet is not in...
10:09 olas reasonably stable IMO
10:11 egonw oh, and the joelib plugin is missing too...
10:11 egonw it seems
10:12 * egonw starts hacking...
10:13 olas joelib?
10:13 olas did not know we had that
10:13 egonw yeah, we do :)
10:13 egonw mom...
10:13 olas move from playground
10:14 egonw http://chem-bla-ics.blogspot.com/2007/10/mo​re-qsar-in-bioclipse-joelib-extension.html
10:14 olas Release is based on trunk, I omit everything in playground
10:14 egonw yeah, you should
10:14 egonw my mistake...
10:14 egonw forgot to move that...
10:14 egonw btw, we also need to move away the de.halle plugins from plugins/ into playground again
10:15 olas could you do that please?
10:15 olas I'm in the middle of something...
10:17 * olas is doing 10 things at the same time now
10:18 * masak hopes that's binary 10
10:23 olas masak is wrong
10:23 egonw it's actually hexadecimal
10:23 olas without chatting it's 9
10:23 olas :)
10:30 masak :)
10:40 ChanServ joined #bioclipse
10:40 masak joined #bioclipse
10:40 thomas_ku joined #bioclipse
10:40 steinbeck joined #bioclipse
10:40 olas joined #bioclipse
10:40 rojasm joined #bioclipse
10:40 egonw joined #bioclipse
10:40 ilogger2 joined #bioclipse
10:40 kaskelot joined #bioclipse
10:54 olas egonw: could you give me a small update on what new plugins should be included in bc2?
10:54 olas I assume no new features
10:54 egonw bc2?
10:54 EskilA joined #bioclipse
11:01 egonw olas: ?
11:04 olas oh sorry
11:04 egonw olas: moved around the plugins
11:04 olas in 1.2.0
11:04 olas so joelib
11:04 olas iand qsar tutorial
11:05 egonw and added the three plugins in the qsar feature.xml
11:05 egonw right
11:05 olas three?
11:05 olas tutorial + joelib + ?
11:05 egonw joelib and joelib.descriptor
11:05 egonw first, just the lib
11:05 olas ok
11:05 egonw second the wrapping
11:05 olas right
11:05 olas good
11:05 egonw will test after lunch...
11:05 olas and it works?
11:05 olas from eclipse?
11:06 egonw after lunch, please
11:06 olas ok, after lunch it is
12:30 olas joined #bioclipse
12:31 egonw hi olas
12:31 egonw I think I'm done
12:31 egonw things work from within Eclipse
12:31 egonw also updated the QSAR cheatsheet to add a step on the scatter plot
12:31 masak joined #bioclipse
12:36 jonalv joined #bioclipse
12:47 egonw olas: ping
12:55 steinbeck left #bioclipse
13:02 olas egonw: pong
13:02 egonw did you see my messages around 13:35 ?
13:03 olas yes
13:03 olas have updated now with joelib, joelib descriptor and tut.qsar
13:03 egonw ack, excellent
13:04 olas will build now
13:04 olas will let you test RC4 soon :-)
13:05 egonw still RC?
13:05 olas until you test it and it passes, yes
13:05 olas then it's release
13:05 egonw what further testing do you like me to do then?
13:06 olas test the binary
13:06 olas for the joelib
13:06 olas well, I guess I can do that as well
13:07 olas binary != eclipse-build
13:07 olas extremely true with eclipse
13:07 olas since completely different CP handling
13:08 olas and builds
13:08 olas that's what's kep messing things up so much for me
13:08 olas you can get it working in eclipse and the binary just dies
13:09 egonw ah, good point
13:10 egonw OK, will test RC5 for the qsar bits
13:12 olas egonw: hmm, I have a problem with qsar
13:12 egonw what is it?
13:12 olas can you verify that you can mix cdk descriptors with joelib descriptors?
13:12 egonw did that earlier...
13:13 egonw will try again...
13:13 olas I can calculate cdk descriptors and joelib descriptors individually
13:13 olas but taking 3 from cdk and 3 from joelib doesn't seem to give good results
13:13 olas at least it's not good in the resulting matrix
13:15 olas no, it's definitely not working for me
13:16 egonw ah...
13:16 egonw sort of working...
13:17 egonw but either set is lost...
13:31 egonw olas: debugging the problem...
13:39 olas egonw: ack
13:50 egonw olas: fixed
13:52 egonw rev 4539
13:52 egonw olas: thanx for that catch
14:04 olas great
14:04 egonw yeah, stupid bug...
14:04 olas all bugs are stupid, right?
14:05 egonw yeah, but some are more stupid than others...
14:05 egonw bit like pigs :)
14:10 olas egonw: I get errors now
14:10 olas java.lang.ArrayIndexOutOfBoundsException: Matrix does not have so many columns! at net.bioclipse.model.JamaMat​rix.set(JamaMatrix.java:32) at net.bioclipse.model.MatrixResource.​insertRow(MatrixResource.java:246) at net.bioclipse.model.MatrixResource.parseS​tringIntoMatrix(MatrixResource.java:214) at net.bioclipse.model.MatrixResource.pa​rseResource(MatrixResource.java:128)
14:10 egonw yeah, same here...
14:10 egonw not sure what that is...
14:10 egonw but does not effect the functionality...
14:10 egonw not that I could detect at least
14:11 olas I don't like to have stacktraces
14:11 egonw do you want me to obscure those exceptions?
14:11 olas or solve them, yes
14:11 olas either or
14:11 olas best is solve :-)
14:12 olas EskilA: I have found an error in plot selections
14:13 EskilA olas: ok, what happens?
14:16 egonw olas: can't seem to replicate the problem :(
14:22 egonw seriously...
14:33 olas hmm
14:34 olas EskilA will assist me
14:43 egonw olas: I just commited a fix for some other problem
14:43 egonw different problem from what you reported...
14:43 egonw but also gave a stacktrace, so obscured it, as you asked
14:43 egonw r4541
14:47 olas I have talked to EskilA, he's on the problem
14:47 olas we found a mtrix to reproduce error
14:47 egonw ok, now I get the problem too again...
14:47 egonw weird...
14:47 egonw hehehe :)
14:47 egonw works for me on the aldehyde/ dir
14:47 olas the problem is parsing the matrix
14:49 olas hmm, could it be that the descriptor calculation constructs a matrix which has diferent number of columns on different rows?
14:49 olas EskilA: is that the problem?
14:49 olas EskilA: are you there?
14:49 EskilA yeah
14:49 olas did you get the file?
14:49 EskilA yes, looking at it now
14:50 EskilA olas, I was thinking that maybe the file has extra entries on some row
14:52 EskilA the ArrayOutof... complains that an  entry is being inserted in a column greater than the number of columns in the matrix
15:01 olas is that so?
15:02 EskilA no, that wasn't the case
15:03 olas the file doesn't seem to have headers for all
15:03 olas and a lot of zeroes in the end
15:03 olas the last 5 columns are just 0's
15:03 EskilA hmm, perhaps that is what makes it wrong
15:03 olas and no headers
15:04 olas try removing them and test
15:04 olas I assume that the fix egonw committed has some sort of error in that case
15:04 olas but please catch these errors better
15:04 olas it should open a dialog and say what's wrong
15:04 olas not just throw exceptions
15:05 EskilA i just noticed that the file looks different in the text editor in bioclipse and mac's own text editor
15:06 EskilA when I removed the header it worked
15:07 egonw I'm with you...
15:08 egonw the zeros should not be there, I think...
15:08 egonw something wrong with the counting...
15:08 egonw let me try something...
15:09 egonw ok, just disregard the columns with higher number if labels are given...
15:09 egonw mom, will commit something for that...
15:15 egonw olas: OK, fix helps...
15:15 egonw no more stacktraces...
15:15 egonw I'll commit
15:16 egonw olas, EskilA: r4542
15:18 egonw http://hardware.slashdot.org/h​ardware/07/12/18/1255222.shtml
15:18 egonw cool... 16GB SSD as small as a penny
15:18 EskilA very nice :)
15:19 egonw please try that fix I just commited
15:22 olas EskilA: please do that
15:26 olas EskilA: can you confirm problem solved?
15:26 EskilA mom
15:27 EskilA I'm getting some very odd behavior
15:28 EskilA even though col > matrixImpl.getColumnCount(), the else clause is executed instead of the if clause
15:28 egonw sure
15:28 egonw if the found column is larger than the col count the matrix has...
15:29 egonw if: do nothing, else: ok, can hold that data
15:29 egonw matrixImpl.getColumnCount() is determined by the found number of labels
15:30 EskilA yes, but I'm checking both variables in the expressions editor, and the behavior is not the expected, the else clause is executed even though col > matrixImpl.getColumnCount
15:31 EskilA could someone else check this, this erratic behavior seems beyond the code
15:31 egonw that sounds rather impossible...
15:31 EskilA hey, blows my mind too
15:32 EskilA going to try restarting
15:32 EskilA It works fine for you?
15:35 egonw for me, yes
15:36 EskilA I think eclipse is executing some code other than the one I'm shown, i added a sysout to the code which was never executed
15:36 EskilA going to try to check out the project again
15:44 EskilA egonw,olas: tried it, works fine
15:46 thomas_ku joined #bioclipse
15:47 olas great
15:47 * olas will update
15:47 olas EskilA: Even the selection problem?
15:51 EskilA olas: not yet
15:52 olas let me know when it's fixed
15:53 EskilA olas: I have a clue at least
15:54 olas good
16:12 * egonw is going home
16:12 egonw bye
16:12 olas bye
17:25 CIA-19 joined #bioclipse
17:36 thomas_ku joined #bioclipse
18:27 egonw joined #bioclipse
18:51 egonw ruby script support??

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