Camelia, the Perl 6 bug

IRC log for #cdk, 2010-07-28

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

All times shown according to UTC.

Time Nick Message
03:24 s9asad joined #cdk
03:59 azeem_ joined #cdk
04:44 egonw joined #cdk
05:26 sneumann__ joined #cdk
05:49 bag_ joined #cdk
05:55 jbrefort joined #cdk
06:22 s9asad joined #cdk
06:48 sneumann__ joined #cdk
08:12 CIA-52 cdk: Egon Willighagen cdk-1.2.x * r6ef1fb1 / src/test/org/openscience/cd​k/debug/DebugBondTest.java : Updated the DebugBond unit test too now: new DebugBond() has zero atoms - http://bit.ly/bZAYpg
08:46 s9asad joined #cdk
09:17 maclean joined #cdk
09:28 s9asad left #cdk
09:53 egonw ok, every one ... CDK 1.2.6 is released
09:53 egonw see: http://chem-bla-ics.blogspot.com/2010​/07/cdk-126-changes-and-authors.html
10:02 maclean \o/
10:02 egonw I'm hoping 1.3.6 will follow next week or so
10:02 maclean (Sorry, I didn't review the bond thing - I got distracted with permutations)
10:03 egonw no worries
10:06 maclean Did you see Mark Southern's rendering options format string?
10:06 maclean It looked good.
10:06 egonw yeah, those were for the servlet, right?
10:07 maclean Think so, yes.
10:08 maclean Obviously, the first thought I had was "json"! (or possibly XML!) :)
10:08 maclean But a one-line format has the advantage of being stuffable into a URL.
10:09 egonw indeed
10:15 s9asad joined #cdk
10:53 maclean Hmmm. http://java.freehep.org - java physics libs.
10:54 maclean :) eg : Class AbstractJetFinder : "Base class for jet finders" - well, indeed.
10:54 * maclean has always wanted a jet finder.
10:54 egonw :)
11:21 s9asad Hi egonw, I have a release version of SMSD and I was wondering if I could submit it
11:22 s9asad Now we have two option, either I can submit it a lib with CDK interface i.e. like signature module or as a patch
11:23 egonw please define SMSD...
11:23 egonw are you talking about a program here, or about the library...
11:23 s9asad http://www.ebi.ac.uk/thornton-srv/software/SMSD/
11:23 egonw whether as lib or as patch...
11:24 egonw both have their value
11:24 egonw depends on what you want
11:24 s9asad The code is already in the CDK but I have made some major updates to it
11:24 s9asad so writing all the updates would be ....
11:25 s9asad Lib idea came from our previous conversation where user gets confused by seeing so many modules
11:25 egonw well, that's why I ask about program versus lib
11:25 s9asad and where as the core of the code is SMSD.java
11:26 egonw what I would suggest is to factor out the application stuff from the library stuff
11:26 egonw the CDK library does not have application code
11:26 egonw it's a library
11:26 s9asad So If I just provide SMSD.java in the CDK and the rest as lib then this will cut down everyone's confusion
11:26 egonw and people develop applications around it
11:26 egonw I don't think that will benefit the CDK library, nor the SMSD code
11:27 maclean so we are talking factoring out in two directions? Up to the application, and Down to a dependent lib?
11:27 s9asad So if you think the whole code is a good idea I am glad to submit the patches
11:27 egonw s9asad: that's how it is now...
11:28 s9asad yes
11:29 s9asad be right back...lunch :-)
11:29 Topic for #cdkis now The Chemistry Development Kit (http://cdk.sf.net) | JUnit CDK 1.2.6: 15673, 76 fails + 15 errors (0.57%) | Logs at http://moritz.faui2k3.org/​irclog/out.pl?channel=cdk | CDK News Feed: http://beta.friendfeed.com/ro​oms/chemistry-development-kit
11:30 Topic for #cdkis now The Chemistry Development Kit (http://cdk.sf.net) | JUnit CDK 1.2.6: 15673, 76 fails + 15 errors (0.57%) | Logs at http://moritz.faui2k3.org/​irclog/out.pl?channel=cdk | CDK News Feed: http://friendfeed.com/rooms​/chemistry-development-kit
11:30 * egonw is wondering what s9asad tries to solve...
11:33 maclean Well, there is the architectural problem of choosing between putting javafiles at the CDK level, or classes in dependant jars.
11:33 maclean Or do you mean scientific problem>
11:34 egonw maclean: the first... but why is that a problem? that's what he wanted in the first place... that's why I have spend hours getting his code in shape
11:35 maclean Well, a) it bulks out the CDK and b) it means you can't easily use the same code in other (non-CDK!) projects
11:36 egonw what do you mean with a) ?
11:36 maclean I know that a) may not seem like a problem, but there is a lot of stuff in the CDK...
11:36 egonw b) -> but isn't it based on the CDK anyway?
11:37 maclean Well, some of it is surely just graph matching, which is more universal.
11:37 egonw not SMSD...
11:37 egonw that's very much oriented at molecules
11:37 maclean Ok, but you could still use it with some other java chemistry lib.
11:38 egonw yes, but you would have to port the code anyway
11:39 maclean Well, that's my point - if the architecture of his code is such that 'porting' just means writing a few implementation classes, it's not really necessary to have the whole thing in the CDK.
11:39 maclean Perhaps I am imagining an architecture that doesn't actually exist...
11:43 egonw well, I guess after lunch we'll learn more about asad's ideas...
11:44 maclean :) agreed.
12:26 s9asad Hi filks
12:26 s9asad folks
12:26 s9asad I am back
12:26 s9asad was reading the conversation
12:28 s9asad a) My point is people who want to use SMSD will have to use the SMSD.java and rest of the code is not interesting to them
12:29 s9asad now if they have n folders which are of least importance to them, and every time I make a change in the core class, the interface might remain the same.
12:30 s9asad and I have to submit large DOCs of my changes which is...I can submit patches and highlight the main changes.
12:30 egonw right
12:30 egonw that sounds like factoring out the application code from the library code
12:30 egonw and that I find a really good idea
12:30 s9asad yes
12:31 egonw I had noted before that to be present in your patch, but I complained about so much already...
12:31 egonw in general, the CDK library indeed does not have any code to interact with users...
12:31 s9asad yes "sounds like factoring out the application code from the library code"
12:31 egonw and it makes perfect sense to make a SMSD application (command line, or GUI) separate from the CDK library
12:31 egonw that's how JChemPaint and Bioclipse work too
12:32 egonw and also solves the problem, that you get a free hand at whatever you like to do with the wrapping application
12:32 egonw and keep the patches for review just about the core functionality, and not about the user interaction
12:32 s9asad so I will compile a .jar of the i.e. SMSD.jar and SMSD.java will remain in the CDK
12:33 s9asad this will be lot easier to review and will come with far lesser complications/confusions :-)
12:33 egonw maclean: thanx for the patch... I hope to look at it shortly...
12:33 maclean ok
12:33 s9asad agree?
12:34 egonw the SMSD.java is the GUI wrapping code, right?
12:34 * egonw 's eclipse is stilll in 1.2.x mode
12:34 egonw if so, you can even put that in a separate project
12:34 egonw mom, let me fire up Eclipse for master
12:35 s9asad The new class will look something like http://github.com/asad/SMSD-CMD/blob/ma​ster/src/org/openscience/smsd/SMSD.java
12:36 egonw ack
12:37 egonw yeah, I guess this one is the API that people would use, so should stay in the CDK lib, I guess
12:37 s9asad yes indeed
12:37 egonw also, I do not spot a main() method, so it is not an application, correct?
12:37 s9asad no
12:37 s9asad its not an application just a api for programmers :-)
12:38 s9asad application is here http://github.com/asad/SMSD-CMD/b​lob/master/src/gui/SMSDFrame.java
12:39 s9asad command line interface is http://github.com/asad/SMSD-CMD/bl​ob/master/src/smsdcmd/SMSDcmd.java
12:39 egonw ack
12:39 egonw and those are nicely in a separate project
12:40 egonw OK, I think I understand now what you mean :)
12:46 egonw s9asad: you might want to make the DOI in the README clickable
12:47 s9asad oh great thanls
12:47 s9asad thanks!
13:38 s9asad Now...
13:39 s9asad experts any idea what is causing this error
13:39 s9asad Exception in thread "main" java.lang.AbstractMethodError:
13:39 s9asad gnu.xml.dom.DomElement.getTex​tContent()Ljava/lang/String;
13:39 s9asad at
13:39 s9asad net.sf.jnati.deploy.artefact.Manifes​tReader.read(ManifestReader.java:76)
13:39 s9asad at
13:39 s9asad net.sf.jnati.deploy.source.ArtefactSour​ce.loadManifest(ArtefactSource.java:42)
13:39 s9asad at
13:39 s9asad net.sf.jnati.deploy.NativeArtefactLocator.​getArtefact(NativeArtefactLocator.java:83)
13:39 s9asad at
13:39 s9asad net.sf.jnati.deploy.NativeLibraryLoader.​loadLibrary(NativeLibraryLoader.java:37)
13:39 s9asad at
13:39 s9asad net.sf.jniinchi.JniInchiWrapper.lo​adLibrary(JniInchiWrapper.java:84)
13:39 s9asad at
13:39 s9asad org.openscience.cdk.inchi.InChIGeneratorFac​tory.<init>(InChIGeneratorFactory.java:79)
13:41 egonw sounds like a library version incompatibililty
13:43 s9asad library version on the cdk?
13:44 s9asad in*
13:44 egonw no, of the gnu.xml.dom lib
13:44 s9asad Oh I thought I am using the lib which comes with jcp n cdk
13:45 s9asad so which lib is it?
13:47 maclean joined #cdk
15:18 sneumann__ joined #cdk
15:35 egonw joined #cdk
16:19 maclean joined #cdk
17:35 s9asad joined #cdk
17:45 s9asad left #cdk
17:57 egonw joined #cdk
20:40 _ilbot2 joined #cdk
20:40 Topic for #cdkis now The Chemistry Development Kit (http://cdk.sf.net) | JUnit CDK 1.2.6: 15673, 76 fails + 15 errors (0.57%) | Logs at http://moritz.faui2k3.org/​irclog/out.pl?channel=cdk | CDK News Feed: http://friendfeed.com/rooms​/chemistry-development-kit
20:50 bag_ joined #cdk
22:03 maclean joined #cdk

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