Camelia, the Perl 6 bug

IRC log for #cdk, 2010-07-26

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

All times shown according to UTC.

Time Nick Message
03:02 azeem_ joined #cdk
05:00 sneumann_ joined #cdk
05:21 jbrefort joined #cdk
05:40 bag_ joined #cdk
06:26 sneumann_ joined #cdk
07:19 mgerlich joined #cdk
07:53 s_wolf joined #cdk
08:37 egonw_ joined #cdk
08:38 egonw joined #cdk
08:38 egonw moin
10:12 egonw_ joined #cdk
12:58 maclean joined #cdk
13:15 maclean joined #cdk
13:41 egonw_ joined #cdk
14:10 bag joined #cdk
14:20 maclean hi egonw_ : did you get an invite to a google wave? I ask mainly out of interest - it's not a pressing issue...
14:21 egonw_ the one about 3D coords?
14:22 maclean yup.
14:23 egonw_ want me to comment on the commits?
14:24 maclean No, not especially. I will put in a copyright header, as you mention.
14:25 egonw_ have been busy reviewing papers and documentation
14:25 maclean Absolutely - sorry, I don't mean to distract you :)
14:26 egonw_ oh, no worries...
14:26 egonw_ was more an update on my CIP state :)
14:26 egonw_ e.g. still need to finish Java code to calculate signatures
14:28 maclean ? What a command-line tool or somesuch...
14:28 egonw_ right
14:31 maclean Well, the 3D stuff (StereoTool) is at least now there to be used, if it becomes necessary.
14:31 egonw_ well, I think it is pretty cool to have this in!
14:31 maclean Although it would be nice to be able to throw a bunch of atoms at it, and get "Tetrahedral" or "Square Pyramidal" back, I'm not actually sure if that is the use-case.
14:31 egonw_ I hope I can soon find time to contribute
14:31 egonw_ e.g. with unit tests
14:32 maclean I've actually made quite a few unit tests, but it's hard to make test cases for geometric situations.
14:32 egonw_ I'd be happy it if I could throw 3D coords of 5 atoms, and get back a ITetrahedralChirality :)
14:32 egonw_ maclean: how so?
14:33 maclean Well, say you want to test the trigonal bipyramidal. It's easy enough to make a 'perfect' configuration, and test that.
14:33 maclean But what about various situations where the atoms are not in some platonically ideal positions.
14:34 maclean In other words: what degree of tolerance should there be, and how to make all these situations.
14:34 egonw_ ah, yes...
14:34 egonw_ well, I guess the tolerance would be configurable, not?
14:35 maclean At the moment, not. It's a class with only static methods. Making it a singleton, with configurable tolerances would be good.
14:37 slyrus joined #cdk
14:39 maclean Also, testing with permutations of the atoms, which should be okay to brute-force test for 4! to 7! - but this can probably all be done at some other point :)
14:40 * maclean should stop inventing new excuses not to do what he's meant to be doing...
14:43 * egonw_ can do brute force testing, if you gave him a Java class with a main()
14:44 maclean well, the best way would be to use the AtomContainerAtomPermutor, I think.
14:45 maclean I should really, really, submit that patch.
14:45 maclean And the group handling stuff, too...
14:48 * egonw sets up to write the CLI to signature calculation
16:06 maclean joined #cdk
17:58 sneumann_ joined #cdk
20:17 sneumann_ joined #cdk
22:39 bag_ joined #cdk
23:36 s9asad joined #cdk

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