Camelia, the Perl 6 bug

IRC log for #bioclipse, 2011-02-16

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

All times shown according to UTC.

Time Nick Message
06:14 egonw joined #bioclipse
07:00 sneumann_ joined #bioclipse
08:08 Gpox joined #bioclipse
08:33 olass joined #bioclipse
08:38 samuell joined #bioclipse
08:45 egonw hi sneumann_, all
09:11 jonalv joined #bioclipse
10:19 olass_ joined #bioclipse
10:23 olass left #bioclipse
10:23 olass_ is now known as olass
12:41 egonw left #bioclipse
13:25 egonw joined #bioclipse
16:15 egonw_ joined #bioclipse
16:18 olass left #bioclipse
16:23 olass joined #bioclipse
16:31 sneumann_ left #bioclipse
16:33 samuell Only one bioclipse question on stackoverflow ... that's not good PR :) http://stackoverflow.com/q​uestions/tagged/bioclipse
16:33 samuell (Or maybe it is ... everything is crystal clear, no questions needed :) )
16:37 egonw_ I see no questions...
16:54 samuell egonw_: And I see one ... strange ...
16:55 egonw_ ah, now I see one too...
16:55 egonw_ but not with your link... which is indeed weird
18:14 jonalv left #bioclipse
18:35 Sonya joined #bioclipse
18:49 egonw_ hi Sonya!
18:49 egonw_ olass: http://gilleain.tumblr.com/post/3329324​344/now-with-transparent-filled-circles
18:49 zarah egonw_'s link is also http://tinyurl.com/5r8xtns
18:49 egonw_ Sonya: welcome
18:54 Sonya hi :)
18:54 zarah Sonya: You have new messages. Write '/msg zarah @messages' to read them.
18:55 Sonya I asked yesterday about cdk and bioclipse IMolecules, thank you for the answers, but my problem is still unresolved)
18:55 egonw_ what was the question?
18:56 Sonya Thing is, I have an IMolecule (from bioclipse), and I want to get a ChemFile from it (which is from cdk). I know that if I have a CDKMolecule I can use getAtomContainer() on it and get a cdk IAtomContainer, and it's easy from there - but what if I have just a general IMolecule?
18:58 olass Sonya: you can do if instanceof(ICDKMolecule)
18:58 Sonya and if it's not, there's nothing I can do?
18:58 olass else you can use the CDKManager like:  cdk.toCDKMolecule(IMolecule mol)
18:58 olass hmm, asCDKMolecule probably..
18:58 egonw_ Sonya: the CDKMolecule has internally a IAtomContainer, not a IChemFile
18:59 Sonya oh, that must be it
18:59 egonw_ Sonya: what do you need the IChemFile for
18:59 egonw_ ?
18:59 * egonw_ has a feeling where this is going ...
18:59 Sonya egonw_: yes, I know, but if I have an IAtomContainer, I think I can convert it from there
18:59 egonw_ right
18:59 olass Sonya: if it is not instanceof then it uses the CML as serialized format
19:00 Sonya egonw_: :) It's actually related to showMolecule in JmolEditor, which takes an IMolecule
19:00 egonw_ Sonya: I would not mind a ChemFileManipulator.createChemFile(IAtomContainer) patch
19:00 egonw_ Sonya: yes, that we need to fix
19:00 egonw_ we need a PDBJmolEditor and a MoleculeJmolEditor
19:01 Sonya egonw_: well, I though that a good way to fix it would be to get a ChemFile from the molecule, which we can then show using openClientFile (which is deprecated for some reason)
19:01 egonw_ because the CDK-Jmol bridge is broken
19:02 Sonya well, there's the CDKJmolAdapter...
19:02 egonw_ the CDK implementation of the Jmol IAtomFoo API need reimplementing
19:02 egonw_ yeah, I know... I wrote it :)
19:02 Sonya I know :)
19:02 egonw_ it doesn't work with Jmol 12
19:02 Sonya why?
19:02 egonw_ Bob tweaked the model
19:02 egonw_ in Jmol
19:03 egonw_ but it can be fixed...
19:03 egonw_ would be great, in fact!
19:03 Sonya I would be glad to try, if only I knew what must be fixed
19:03 egonw_ but we first need to split IProtein and IMolecule at the Jmol leve
19:03 egonw_ l
19:04 egonw_ that I do not know at this moment in detail either :)
19:04 egonw_ but let's simplify things first...
19:04 egonw_ the CDK manager doesn't support protein anyway
19:04 egonw_ only small molecules...
19:04 egonw_ Sonya: in fact, the simplest approach would just be to convert an ICDKMolecule to a IJmolMolecule
19:05 egonw_ via CML, as olass suggested earlier
19:05 egonw_ Jmol has a decent CML reader
19:05 egonw_ and CDK has no trouble with that either
19:05 Sonya ok, I'll look into that, thank you
19:06 egonw_ and the JmolEditor needs a matching split up into ProteinJmolEditor
19:06 egonw_ and MoleculeJmolEditor
19:06 egonw_ that will also simplify the Outline provider
19:07 egonw left #bioclipse
19:07 egonw_ is now known as egonw
19:23 Eugene joined #bioclipse
19:24 Eugene is now known as Guest52952
19:25 egonw hi Guest52952
19:28 Guest52952 is now known as Eugeny
19:28 Eugeny Hi, Egon
19:30 Eugeny is now known as eugenek
19:30 eugenek is now known as Guest16013
19:30 Guest16013 is now known as dfghjkhg
19:31 dfghjkhg left #bioclipse
19:33 samuell ^^
19:50 olass ~o~
19:55 Eugene joined #bioclipse
19:55 Eugene is now known as Guest29344
20:13 Guest29344 is now known as Eugeny
20:13 Eugeny I came back )
20:13 egonw :)
20:16 Eugeny To make our infusion in bioclipse development we (+Sonya) decided to make some simple tasks at first
20:17 egonw sounds good...
20:17 egonw what did you have in mind?
20:19 Eugeny Sonya told you about CDK, Jmol etc... So, at first the simpliest task can be thougth up is to eliminate using of temporary files in JmolEditor
20:19 egonw temp files?
20:19 Eugeny Then to split interfaces as you offer
20:19 egonw or passing data around as 'files'?
20:20 egonw it's actually using temp files on disk?
20:20 Eugeny You are right, temp files. I think now using CML or something like that
20:20 Eugeny But in memory, not on a disk
20:20 egonw ah, ok...
20:21 egonw then I would suggest to split the editors and interfaces first
20:21 egonw then you can make two different mappings, which should be easier
20:21 Eugeny Ok, it would be our next task )
20:23 Eugeny Oh, and what about task managing in bioclipse development? I mean is there anything except bugzilla? Redmine, JIRa for instance
20:26 egonw no, there currently is just bugzilla
20:26 egonw doesn't mean there cannot be anything else...
20:26 egonw dunno redmine
20:27 Eugeny Ok. And coming back to my very first question on that IRC channel...
20:28 Eugeny Within our plugin there is a task to show something near molecule in Jmol viewer
20:28 Eugeny I suppose to make a pseudo atom (or Molecule) type to use in Jmol and analyse selections etc...
20:30 egonw jmol can also draw arbitrary things...
20:30 Eugeny Is it the right way and where I can see some examples of new types of atoms for Jmol?
20:30 Eugeny If it'll be less arbitrary (line, sphere...) it would be easier to act with it :)
20:31 egonw yeah, I need some more context
20:32 Eugeny For example I have a molecule (protein) showing in Jmol
20:32 Eugeny I need to draw a figure (consisting of spheres and lines) and make it selectable and moving
20:33 Eugeny The main idea is to make pseudomolecule (now we use hydrogen atoms :) ) and draw it like the second model on the same Jmol panel
20:40 egonw I suggest you ask on the jmol developers mailing list how that is reflected in the data model
20:41 egonw you can always do it via jmol scripting, of course
20:43 Sonya I tried to do it via jmol scripting, but failed. it seems that there is no way to combine arbitrary drawn objects (lines and points) into one new object (or I just haven't found it)
20:45 egonw best to ask that with the master: Bob (on the jmol ML)
20:52 olass If you have a good use case, Bob might implement it for you if missing
20:52 * olass is going to bed
20:52 olass good night!
20:52 olass left #bioclipse
20:57 olass joined #bioclipse
20:58 egonw_ joined #bioclipse
21:01 olass left #bioclipse
21:01 egonw left #bioclipse
21:06 egonw_ Sonya, Eugeny: I got to go too...
21:06 egonw_ cu later
21:07 Sonya bye
21:07 Sonya left #bioclipse
21:11 egonw_ left #bioclipse
21:11 Eugeny left #bioclipse
21:17 samuell left #bioclipse
21:52 jonalv joined #bioclipse
22:24 Gpox left #bioclipse
22:38 jonalv left #bioclipse

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