Camelia, the Perl 6 bug

IRC log for #cdk, 2009-03-09

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

All times shown according to UTC.

Time Nick Message
05:16 timvdm joined #cdk
05:40 tim__ joined #cdk
06:20 jbrefort joined #cdk
08:34 rojasm joined #cdk
09:15 shk3 joined #cdk
09:23 julio joined #cdk
09:32 Gpox joined #cdk
09:35 jonalv joined #cdk
11:27 egonw joined #cdk
11:47 CIA-62 cdk: gilleain * r14377 /cdk/branches/jchempaint-primary/src/main/org/​openscience/cdk/controller/ControllerHub.java: Better layout of more than two disconnected fragments in an atom container.
12:50 CIA-62 cdk: gilleain * r14378 /jchempaint/trunk/src/main/org/openscienc​e/jchempaint/action/CopyPasteAction.java: Using IAtom, IBond, IReaction instead of Atom, Bond, Reaction.
12:55 carsten joined #cdk
13:08 CIA-62 cdk: gilleain * r14379 /jchempaint/trunk/src/main/org/openscienc​e/jchempaint/action/CopyPasteAction.java: Narrower exception handling.
13:09 egonw rojasm: ping
13:09 rojasm hi egon
13:09 rojasm yes I book the registration
13:09 egonw hi rojasm
13:09 rojasm with Julio too
13:09 egonw when will you fly back?
13:09 julio hi rojasm and egonw
13:10 egonw forgot what we decided
13:10 egonw hi julio!
13:10 egonw julio: good to see that your IRC client doesn't log in/out all the time anymore :)
13:10 julio we fly back Wednesday 22nd in the afternoon-evening
13:10 rojasm we expected Wednesday afternoon to take the fly
13:10 egonw OK
13:11 julio I will  be discussing wednesday morning with Stephan about the structure generator
13:11 rojasm perfect!
13:12 julio if you guys want to join you are more than welcome
13:25 julio ping
13:25 julio egonw:ping
13:25 julio egonw: ping
13:25 julio finally
13:25 julio do you guys have any experience with concurrency in CDK?
13:26 julio making use of multi-core computing power?
13:26 egonw a bit
13:27 julio I have seen the post of Rajarshi Guha on Clojure
13:27 julio is Clojure the best way of making our code run on multicore?
13:27 egonw no idea
13:27 egonw it's a means
13:28 egonw it hides a lot of the stuff to do to make it run on multicores
13:28 egonw it looks promosing
13:28 egonw regarding structure generation...
13:28 egonw it's the way you search the connectivity space that gives the speed
13:28 egonw branch-and-bound
13:28 egonw where the bound is criticial
13:28 julio sure
13:29 egonw not the spreading of the branches
13:29 julio the issues are generation speed and storing of the generated structures
13:29 julio for instance
13:30 julio MOLGEN goes quite fast, but makes a huge ise of memory
13:30 julio on the other hand, Assemble2 (written in java) is slower but needs less memory
13:31 * egonw does not know assemble2
13:31 egonw btw, define speed
13:32 julio speed = CPU time
13:32 egonw but CPU time cannot be a cause
13:34 julio memory use I think is more sensitive
13:40 julio http://www.upstream.ch/products/assemble.html
13:40 zarah julio's link is also http://tinyurl.com/cypmqs
14:00 egonw joined #cdk
14:54 egonw joined #cdk
15:16 jbrefort joined #cdk
15:39 CIA-62 cdk: rajarshi * r14380 /cdk/branches/cdk-1.2.x/src/ (2 files in 2 dirs): Filled in the code for the Hill string generator. Updated test cases. Fixes bug 2672696
15:43 julio left #cdk
15:59 CIA-62 cdk: egonw * r14381 /cdk/branches/cdk-1.2.x/build.props: Bumped version, for 1.1.6 release
16:01 CIA-62 cdk: egonw * r14382 /cdk/tags/cdk-1.1.6/: Released 1.1.6
16:03 CIA-62 cdk: egonw * r14383 /cdk/branches/cdk-1.2.x/src/ (2 files in 2 dirs): Fixed adding entry end lines
16:08 rajarshi joined #cdk
16:09 rajarshi hi
16:09 zarah oh hai rajarshi
16:09 rajarshi egonw, i'm looking at the mol2 bug - how does one use a StringWriter to get the output?
16:09 egonw toString()
16:10 egonw got to catch my bus now
16:10 egonw will be online tonite
16:10 rajarshi ok
16:33 CIA-62 cdk: rajarshi * r14384 /cdk/branches/cdk-1.2.x/src/ (2 files in 2 dirs):
16:33 CIA-62 cdk: Updated Mol2Writer to address bug 2675188 so that atom symbols are included and
16:33 CIA-62 cdk: atom/bond numbering starts from 1. Also fixed the bond order, though all the
16:33 CIA-62 cdk: bond orders in the Tripos spec such as am are not yet supported. Also, the atom
16:33 CIA-62 cdk: types are currently wrong. Added a unit test
16:52 CIA-62 cdk: rajarshi * r14385 /cdk/branches/cdk-1.2.x/src/main/org/opensci​ence/cdk/atomtype/SybylAtomTypeMatcher.java:
16:52 CIA-62 cdk: Updated sybyl atom type matcher so that it correctly performs the mapping from
16:52 CIA-62 cdk: CDK atom type to sybyl atom type. The problem was that the mapping method took
16:52 CIA-62 cdk: an IAtomType rather than IAtom - as a result it did not have aromaticity
16:52 CIA-62 cdk: information and thus was unable to properly generate C.ar types.
16:56 CIA-62 cdk: rajarshi * r14386 /cdk/branches/cdk-1.2.x/src/test/org/o​penscience/cdk/io/Mol2WriterTest.java: Updated unit tests to check MOL2 output for an aromatic molecule. Works fine now
17:00 CIA-62 cdk: rajarshi * r14387 /cdk/branches/cdk-1.2.x/src/test/org/o​penscience/cdk/io/Mol2WriterTest.java: Fixed javadoc
17:13 CIA-62 cdk: rajarshi * r14388 /cdk/branches/cdk-1.2.x/src/ (2 files in 2 dirs): Added support for amide bond output in the MOL2 writer. Also added unit test
17:31 jbrefort joined #cdk
19:43 |newbie| joined #cdk
19:54 |newbie| hi rajarshi
19:54 egonw that's better :)
20:07 rajarshi aah :)
20:07 rajarshi i was wondering
20:09 egonw struglling with laptops
20:09 egonw this one has network
20:09 egonw but my wifes account does not work
20:09 egonw so, was trying to get it fixed
20:09 rajarshi aah
20:09 rajarshi btw, the mol2 bug from Noel is fixed, haven't closed though
20:09 egonw my other laptop which I had to reinstall this morning does not have the right password for the wireless yet...
20:09 egonw yes, saw the commits, thanx
20:10 egonw there wasn't even a unit test?
20:10 egonw weird...
20:10 rajarshi no!
20:10 egonw I though I had written that...
20:10 egonw anyway... there are some unfinished brain dumps here and there
20:10 egonw it's good we have this patch system now
20:11 rajarshi indeed
20:11 egonw so that unfinished stuff doesn't as easily enter main stream
20:11 egonw I am worried a bit on where people can host their git branches...
20:12 egonw need to figure out anyway how to put git branches on SF
20:12 egonw btw, nice clojure blogs...
20:12 egonw will need to look at this...
20:12 egonw I do have the feeling though...
20:12 egonw that it still requires the CDK to be nicely written to allow parallelization...
20:12 rajarshi i don't see that as a problem - there should be one person (you?) whose branches are the blessed ones
20:13 rajarshi periodically you can pull from other peoples to merge them in
20:13 egonw that's my point
20:13 egonw people need to be able to put their branch somewhere online for me to pull
20:13 rajarshi re clojure, thanks. Yes, can't make full use of parallelization - but it's a nice toy to play with :)
20:13 egonw or I need to figure out the whole mailbox thingy...
20:13 rajarshi sure, but I don't think that's a big problem
20:14 egonw which would require me to set up a local client again, I guess...
20:14 egonw which is actually not so bad an idea...
20:14 rajarshi especially since setting up a git server to pull from is not a big problem
20:15 egonw maybe something to blog about :)
20:17 rajarshi i think the move to git deserves a wiki page describing how our procedures would owrk
20:28 egonw yes, I think so too
20:29 egonw need to reboot
20:29 egonw brb
20:40 egonw joined #cdk
21:17 rajarshi off home, good night
21:20 egonw ok, bye!
21:23 jbrefort joined #cdk
22:02 dleidert joined #cdk
22:02 dleidert egonw: around?
22:02 egonw yes
22:02 dleidert nice.
22:02 egonw about to go to bed, but around for a few more minutes
22:02 dleidert Not much to say ;)
22:03 dleidert Can you check
22:03 egonw I have Lenny installed, but had no time yet for testing building CDK 1.1.x debs
22:03 dleidert http://bugs.debian.org/474908
22:03 egonw checking
22:03 dleidert The patch is at
22:04 dleidert mom
22:04 egonw mmm... interesing how a gcc upgrade triggered this :)
22:04 dleidert Maybe you just need a "*/" || "**/"
22:04 dleidert http://svn.debian.org/wsvn/pkg-java/tru​nk/cdk/debian/patches/10-MakeJavafilesF​iles.patch?op=file&rev=0&sc=0
22:04 zarah dleidert's link is also http://tinyurl.com/aqfumb
22:05 egonw it sounds like a change with the output of javacc
22:05 egonw had that version changed recentlt?
22:05 dleidert Uh, the report is about 1 year old
22:05 egonw really...
22:06 egonw did not see it 'upstream' :(
22:06 dleidert Michael is caring about cdk. But he seems MIA and I'm currentlyx checking cdk for an update and found this patch.
22:07 dleidert Sorry. Paul probably didn't know, that I have access to upstream VCS.
22:07 egonw yeah, that patch looks OK
22:08 egonw how to proceed?
22:08 egonw I can apply upstream...
22:08 egonw but to 1.1.x or 1.0.x?
22:08 dleidert Just wanted to show it to you. If you apply it, we can drop it.
22:08 dleidert That's up to you.
22:08 egonw this is for post-Lenny, right?
22:08 dleidert Yeah.
22:08 dleidert The packaging problem itself has been fixed.
22:09 egonw OK, then I'll apply it to 1.1.x
22:09 dleidert Ok. Good to know.
22:09 dleidert What would you prefer: Having 1.1.x in Debian or 1.0.x or both?
22:09 egonw but not tonite...
22:10 egonw will file it as bug report, and add comment to debug BTS
22:10 dleidert No problem. Just wanted to let you know.
22:10 egonw just 1.1.x
22:10 egonw it's much better
22:10 dleidert Ok.
22:10 egonw and since we do not have JChemPaint...
22:10 egonw or does debian?
22:10 egonw it doesn't, right?
22:11 dleidert Not yet. There is a report requesting it.
22:11 dleidert But I'm not happy with the I18N system. gettext would have been fine.
22:11 dleidert ... for jchempaint.
22:12 egonw but you introduced the jmol system, not??
22:12 egonw or are you refering to the old JCP?
22:12 dleidert we did, IIRC
22:12 dleidert Nope, I saw some I18N work recently on JCP
22:12 egonw yes, that is using the same set up as Jmol, AFAIK
22:12 dleidert The first try was with gettext-commons (not with gettext for some reason?)
22:13 egonw dunno... stefan kuhn ported the Jmol system to JCP
22:14 dleidert Yeh. It was not that long ago. I'm currently not following the development closesly.
22:14 egonw https://sourceforge.net/tracker2/?func=detail&am​p;aid=2676510&group_id=20024&atid=120024
22:15 egonw mmm... the bug report actually mentions that I planned to apply to patch... but will verify anyway...
22:16 dleidert If you apply the patch, we will get the information anayway - because our patch will fail to apply then ;)
22:17 egonw indeed
22:17 egonw OK, the 1.1.x branch has the same pure dist .tar.gz, and I have Lenny set up in VirtualBox, but had not time for testing to build it yet...
22:17 egonw my pbuilder environment got lost...
22:18 egonw but just in case you want to try to build a deb for 1.1.6 or so...
22:18 egonw just let me know all trouble you run into...
22:18 egonw otherwise, I should have time for it in a few weeks... (post grant writing)
22:18 egonw OK, going to bad now...
22:18 egonw ok, my reply to the BTS was rejected...
22:19 egonw OK, going to bed now
22:19 egonw bye!

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