Camelia, the Perl 6 bug

IRC log for #cdk, 2008-02-19

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

All times shown according to UTC.

Time Nick Message
04:10 CIA-48 joined #cdk
05:56 CIA-48 cdk: egonw * r10180 /trunk/cdk/src/org/openscience/cdk/test/ (2 files in 2 dirs): Converted into JUnit4 test
05:57 CIA-48 cdk: egonw * r10181 /trunk/cdk/src/org/openscience/cdk/test​/smiles/DeduceBondSystemToolTest.java: Added a few more unit tests, from DeduceBondOrderTestFromHybridization
05:57 CIA-48 cdk: egonw * r10182 /trunk/cdk/src/org/openscience/cdk​/smiles/DeduceBondSystemTool.java: Use the new atom typer...
05:58 CIA-48 cdk: egonw * r10183 /trunk/cdk/ (30 files in 6 dirs): Set up a proper qsarweka module
06:33 jbrefort joined #cdk
07:08 rojasm joined #cdk
07:21 rojasm moin
07:43 egonw joined #cdk
07:43 egonw moin
07:43 egonw will leave for a meeting in about 10 minutes
07:47 carsten joined #cdk
07:50 egonw bbl
08:21 Annzi joined #cdk
08:24 man-di moin
09:16 jpansanel moin
09:18 egonw man-di: fire your questions
09:18 steinbeck joined #cdk
09:20 man-di egonw: later, currently doing some real work
09:20 egonw ack
09:21 man-di but I have to say I dont like your massive log4j jar patch
09:21 egonw ack, noted
09:21 man-di its a maintenance burden
09:21 man-di IMO
09:22 steinbeck moin
09:22 egonw man-di: I'll think of a different way to handel module deps on third parties
09:23 egonw man-di: as an intermediate solution, I can use 1.2.15 in CDK SVN...
09:23 egonw then no patch is needed at all, nor any build system improvement...
09:23 egonw would you be happy with that?
09:23 man-di egonw: cant you just use the general link like log4j-1.2.jar and commons-cli.jar?
09:24 egonw yes, sure
09:24 man-di when you hardcode 1.2.15 and Debian updates to 1.2.16 we have the same problem
09:24 egonw right...
09:25 egonw man-di: how can I determine what the sym link to use is?
09:25 egonw does the java policy say something about this?
09:25 man-di look into the package
09:25 egonw ok, I'm fine with a best guess approach...
09:26 man-di normally its name.jar instead of name-version.jar
09:26 egonw but that does not garantee that a new deb for the lib will have a different sym link...
09:26 egonw that's my point...
09:26 man-di log4j is an exception, its uses name-apiversion.jar
09:26 egonw ok, that's good enough policy for me
09:27 man-di so name.jar/name-apiversion.jar always points to name-version.jar
09:27 man-di I will fix jgrapht soon. tonite perhaps
09:27 egonw good
09:56 masak joined #cdk
10:06 de_maas joined #cdk
10:14 CIA-48 bioclipse: j_viklund * r4903 /branches/bioclipse2/core/plugins/ (5 files in 4 dirs): Readded log4j to net.bioclipse.logging. Now it runs.
11:48 CIA-48 cdk: egonw * r10184 /branches/cdk-1.0.x/ (29 files in 3 dirs): Renamed file, to match Debian's scheme. Still at Log4j 1.2.13 in our SVN
11:51 CIA-48 cdk: egonw * r10185 /branches/cdk-1.0.x/ (6 files in 2 dirs): Renamed file, to match Debian's scheme. Still at version 1.0 in our SVN
11:53 CIA-48 cdk: egonw * r10186 /branches/cdk-1.0.x/ (26 files in 2 dirs): Renamed file, to match Debian's scheme. Still at XOM 1.1 in our SVN
11:55 CIA-48 cdk: egonw * r10187 /branches/cdk-1.0.x/ (56 files in 2 dirs): Renamed file, to match Debian's scheme. Still at 1.14 in our SVN
11:58 CIA-48 cdk: egonw * r10188 /branches/cdk-1.0.x/ (18 files in 2 dirs): Renamed file, to match Debian's scheme. Still at Jama 1.0.2 in our SVN
12:13 egonw man-di: I'
12:14 egonw man-di: I'm uploading 1.0.1.91 with no longer requires the debian/patches
12:14 egonw no new changes to any other debian/ files made
12:28 man-di egonw: thx, will look at it later
12:29 man-di egonw: there were some perl scripts, why are these not includeded anymore?
12:29 man-di dleidert wrote manpages for them
12:29 man-di are they still needed and not in the release tarball?
12:33 egonw not sure if they were part of the orig.tar.gz ...
12:33 egonw they might have... a long time ago...
12:33 egonw I vaguely remember having a tar.gz with that...
12:33 egonw but those have not been part of CDK library releases for ages...
12:33 egonw mom...
12:40 jpansanel egonw : I'd like to make debian package with Mychem.
12:41 jpansanel I'll ask to dleidert some hints
12:42 jpansanel (it's not so easy, as Mychem is an add-on or MySQL)
12:45 man-di jpansanel: you can also ask me
12:45 egonw rojasm: ping
12:45 jpansanel man-di : who are you ?
12:46 man-di jpansanel: a Debian developer
12:46 jpansanel ok :)
12:46 man-di jpansanel: currently packaging cdk
12:46 jpansanel man-di : I've the main developper of Mychem
12:47 jpansanel Mychem is a set of UDF for MySQL based on OpenBabel
12:47 jpansanel It's released under GPLv2+
12:47 jpansanel and I'm asking how to build the debian package
12:47 jpansanel (the makefile is based on CMake)
12:48 jpansanel s/makefile/make system
12:50 man-di jpansanel: can you mail me a link to the latest release tarball and I will prepare a debian package: konqueror@gmx.de
12:50 jpansanel sur
13:01 jpansanel man-di: sent
13:02 man-di thx
13:02 man-di I will mail you when I get to it
13:08 rojasm egonw: yes. Did you ping me?
13:08 egonw yeah, but never mind
13:08 egonw already solved
13:08 rojasm :)
13:10 egonw man-di: the scripts are available from: http://cdk.svn.sourceforge.net/viewvc​/cdk/trunk/cdk/packages/tar/cdk/bin/
13:10 egonw really old, as you can see
13:11 egonw that was when the cdk still had a tar.gz patch with a autotools set up
13:12 man-di egonw: are they still useful?
13:12 egonw no, don't think so...
13:12 egonw the 1.0.x release have really been library releases...
13:12 egonw those apps were more quick hacks... to test things...
13:12 man-di okay, then I will remove that manpages from the debian dir
13:13 egonw yeah, makes sense I htink
13:13 man-di okay
13:13 man-di food
13:13 man-di good
13:13 man-di damn typo
13:13 egonw possibly as dleidert about id he had something in mind
13:13 man-di egonw: he sent me to ask you
13:13 egonw :)
13:13 egonw ok, drop it
13:13 egonw I haven't had resources to support them...
13:13 egonw and no one else did either
13:42 thomas_ku_ joined #cdk
14:32 CIA-48 bioclipse: carl_masak * r4904 /branches/bioclipse2/core/plugins/n​et.bioclipse.ui/src/net/bioclipse/u​i/actions/DummyActionHandler.java:
14:32 CIA-48 bioclipse: [DummyActionHandler.java]
14:32 CIA-48 bioclipse: * circumcoded NullPointerException bug
14:32 CIA-48 bioclipse: * occurred when Bioclipse Explorer is empty
15:58 man-di egonw: I'm confused, what is the build.xml target to execute to get javadocs?
15:59 egonw that's the html target in javadoc.xml
15:59 egonw which I guess is not in the distrib?
15:59 egonw (guessing from your question...)
16:01 man-di just found javadoc.xml
16:01 man-di but it needs ant-doxygen
16:02 man-di and that is not in Debian
16:02 man-di either I package that too or I remove the -doc package...
16:04 carsten joined #cdk
16:06 egonw remove the -doc for now
16:06 egonw I would say
16:09 man-di egonw: the package will need to go through NEW queue again when we add a -doc package
16:10 man-di so the idea was to do it once
16:10 man-di packaging ant-doxygen shoulnt be that hard...
16:10 man-di egonw: why dont you use javadoc?
16:12 egonw oh wait...
16:12 egonw the html target is not using doxygen
16:13 egonw doxygen is only used to test JavaDoc quality...
16:13 egonw ummm... no, that's doccheck...
16:13 egonw arggh.... been a CDK developer for way too long...
16:14 egonw you are asking things I have not worked on for years...
16:14 egonw mom...
16:14 * egonw is reading the javadoc.xml
16:14 man-di I just did: ant -buildfile javadoc.xml as described in README
16:14 man-di egonw: hehe
16:14 egonw that called the html target, not?
16:14 man-di egonw: I get you back to all your worst dreams :-)
16:15 man-di egonw: taskdefs are loaded before targets are executed
16:15 egonw ack
16:16 egonw ok, so the conclusions are:
16:16 egonw 1. cdk-doc uses javadoc, so can be build
16:16 egonw 2. doxygen use in javadoc.xml has to be removed
16:17 egonw ad 2: a debian/patch, or do you want me to remove that upstream?
16:17 man-di patch is good enough for me, I can do that
16:18 egonw ok
16:20 man-di egonw: hmmm, what should be in the doc/javadoc/source dir?
16:20 egonw there were Doclets in there...
16:20 egonw which have been used in the past for website building
16:28 man-di its not in the dist tarball and its needed for html target
16:29 egonw oi...
16:29 egonw yes, I see it...
16:29 egonw I also see a hardcoded xom-1.1.jar in there...
16:29 egonw won't be able to fix that before this weekend...
16:30 egonw you could patch the javadoc.xml to exclude the use of those doclets...
16:30 man-di if you send me the needed files I can add them as patch
16:30 man-di or that
16:30 man-di I will do that
16:30 egonw that is, comment out thoe <taglet> lines
16:30 egonw ok, thanx
16:30 egonw will now file a bug report
16:31 egonw so that I'll still fix that for a next release
16:33 man-di thx
16:42 carsten jbrefort: done
16:52 jbrefort carsten: done what?
16:54 carsten s/jbrefort/jpansanel
16:54 carsten sorry
16:54 carsten jpansanel: done
16:54 jbrefort ;)
16:58 carsten egonw: Avogadro can now render surfaces
16:58 carsten still a bit buggy but rox already :)
16:58 Annzi left #cdk
16:58 egonw carsten: yes, saw the screenshots
16:58 carsten But you need the latest OpenBabel, latest in the meaning of "today" I think :-)
16:58 * egonw heads home
16:59 steinbeck_ joined #cdk
17:08 rojasm left #cdk
17:13 dleidert joined #cdk
17:45 steinbeck_ joined #cdk
17:55 CIA-48 cdk: rajarshi * r10189 /trunk/cdk/src/org/openscience/c​dk/charges/Polarizability.java: Updated code so that we can specify whether explicit H's should be added. This is useful since if we call the method multiple times for the same molecule, there's no point in repeating the addition of explicit H's
17:57 CIA-48 cdk: rajarshi * r10190 /trunk/cdk/src/org/openscience/cdk/qsar/d​escriptors/molecular/BCUTDescriptor.java: Updated call to polarizability calculation to indicate that there's no need to add explicit hydrogens, since we add it ourselves
18:25 CIA-48 cdk: rajarshi * r10191 /trunk/cdk/src/org/openscience/cdk/ (2 files in 2 dirs):
18:25 CIA-48 cdk: Updated code so that we only create a new molecule object (copy of the input
18:25 CIA-48 cdk: molecule) is we need to add explicit H's. If we don't need to add explicit H's
18:25 CIA-48 cdk: we can save a ton of time by not creating a new molecule. Also updated unit test
18:25 CIA-48 cdk: code
18:26 rojasm joined #cdk
18:33 thomas_ku_ joined #cdk
18:37 CIA-48 cdk: rajarshi * r10192 /trunk/cdk/src/org/opensci​ence/cdk/qsar/descriptors/ (2 files in 2 dirs): Updated code to conform with the modifications in the pol calculation to indicate that explicit H's should be added. Need to udpate autocorrelation descriptors to add explicit H's before calling pol calculation
19:06 man-di egonw: I dont know what the javadoc task is doing but it always fails with "argument list is too long"
19:07 man-di egonw: I think I will have to wait until that is somehow be fixed
19:10 carsten joined #cdk
19:25 egonw_ joined #cdk
19:29 egonw_ moin
19:34 man-di egonw_: wb
19:34 man-di egonw_: 20:06 < man-di> egonw: I dont know what the javadoc task is doing but it always fails with "argument list is too long"
19:34 man-di 20:07 < man-di> egonw: I think I will have to wait until that is somehow be fixed
19:34 man-di egonw_: you have too many source files, it seems ;-)
19:35 egonw_ ok, then it'll have to do without the -doc .deb
19:37 man-di okay, I will remove this later and upload the package to NEW
20:23 CIA-48 jmol: aherraez * r9007 /trunk/Jmol/src/org/jmol/tr​anslation/JmolApplet/es.po: Spanish
21:09 dleidert joined #cdk
21:25 egonw_ hi dleidert
21:25 dleidert hi
21:27 * egonw_ is going offline
21:27 egonw_ bye all
21:27 egonw_ man-di: if you email me the debian/ stuff you have now, I will have a look this weekend at packaging Jmol...
21:28 egonw_ which would be much easier
21:28 egonw_ really bye now
21:49 rojasm left #cdk

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