Camelia, the Perl 6 bug

IRC log for #cdk, 2010-09-16

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

All times shown according to UTC.

Time Nick Message
03:15 azeem_ joined #cdk
03:19 azeem left #cdk
03:55 sneumann joined #cdk
04:03 egonw joined #cdk
04:06 sneumann left #cdk
04:40 egonw left #cdk
04:40 jerome_ left #cdk
04:44 jerome_ joined #cdk
04:55 jerome___ joined #cdk
04:58 jerome_ left #cdk
05:23 sneumann joined #cdk
06:24 egonw joined #cdk
06:43 sneumann left #cdk
07:01 Gpox joined #cdk
07:23 zarah left #cdk
08:07 sneumann joined #cdk
08:29 jbrefort joined #cdk
08:41 maclean joined #cdk
10:41 maclean egonw : what are the requirements for data files to be compatible with a software licence?
10:42 azeem_ ugh, that's a mine field
10:42 egonw nothing I am aware of
10:42 egonw it's part of the cdk distrib, so LGPL?
10:42 egonw maclean: you know of better solutions?
10:42 egonw azeem_: aren't data files not always with the same license as the software they come with?
10:42 maclean I know nothing about nothing in this area :)
10:43 azeem_ egonw: yeah well
10:43 azeem_ but I doubt lots of them got donated under that impression
10:43 azeem_ e.g. some PDB downloaded fromt the PDF database
10:43 maclean For example, if I type this smiles : "CCCC(C)CCC1CCC1(O)CC" is it GPL?
10:43 azeem_ eh, PDB database*
10:43 azeem_ maclean: your question is whether it is copyrightable at all?
10:44 maclean indeed.
10:44 azeem_ 12:41 < azeem_> ugh, that's a mine field
10:44 egonw well, if you include it in a CDK unit test, I do hope it is not some imcompatible copyright/license...
10:44 egonw I really don't know how to rephrase it... or what to ask...
10:44 azeem_ what if somebody gets the idea that the "source code" of the PDB really is the raw x-ray data`?
10:44 egonw just want to make sure we do not start distributing stuff we are not supposed to
10:45 maclean well, sure. I'm just not convinced that it's something to worry about...
10:45 egonw I don't
10:45 egonw I put any test data in the repos
10:45 egonw if I created it...
10:46 egonw or 'found' it in a unit test
10:46 egonw and when I commit that data, at make it LGPL compatible... (aka "it can be released with the CDK")
10:46 maclean Ok, but you are still concerned that there might be data 'out there' that is under some more restrictive licence?
10:46 azeem_ of course
10:46 egonw oh, there sure is
10:46 maclean Say, from a GPL project.
10:47 azeem_ eh
10:47 egonw in chemistry most data is some obscure, restrictive license
10:47 egonw maclean: yes, indeed...
10:47 egonw I am not eager in putting in test data from OpenBabel
10:47 azeem_ in the end, assigning source code licenses to data is not so great
10:47 egonw (Geoff and I discussed this... no worries)
10:47 azeem_ CC makes more sense; but really, I wouldn't worry too much
10:47 maclean But what are they licencing? The connectivity? The coordinates?
10:48 egonw no, the file formatting
10:48 maclean Benzene is benzene is benzene, I would say.
10:48 egonw which is copyrightable in many countries
10:48 azeem_ egonw: I would expect OpenBabel to happily relicense any test data to LGPL, I don't think any OpenEye data files are in there
10:48 egonw not sure about that
10:48 egonw geoff and I did speak about more liberal test data
10:48 egonw but we never found enough time to bootstrap that yet
10:49 egonw i'm not worried about the data that is in the repos
10:49 azeem_ there was talks about a chemical data repository
10:49 egonw someone with commit rights added that there
10:49 egonw I only want to make sure what people do when they submit test data...
10:49 egonw and since we deal with pharma customers too ...
10:49 maclean True.
10:50 egonw azeem_: yeah, that could be jerome___'s project
10:53 azeem_ I can't find the mail right now
10:54 maclean well, thankyou both. I will now know to be more careful about this in the future. Although many of my test cases don't even exist as chemicals...
11:54 maclean egonw : sorry, another quick question - do you have any objections to IRenderer<T extends IChemObject> becoming just IRenderer<T>?
11:57 egonw ah, to render non-chemical stuff?
11:58 maclean well, to render things that might not necessarily be chemobjects.
11:59 maclean like, wrappers around chemobjects, for example.
11:59 egonw wrappers?
12:00 maclean I just mean classes that contain references to instances of other classes.
12:00 maclean I guess even List<IMolecule>, for example.
12:00 egonw ah, like that...
12:01 maclean Map<IMolecule, IMolecule>, etc.
12:57 egonw left #cdk
13:08 s9asad left #cdk
13:10 jbrefort left #cdk
13:44 sneumann left #cdk
13:56 s_wolf left #cdk
14:31 egonw joined #cdk
14:38 egonw left #cdk
14:44 egonw joined #cdk
14:59 s9asad joined #cdk
15:11 s9asad left #cdk
15:33 jbrefort joined #cdk
15:42 egonw left #cdk
16:07 maclean left #cdk
16:22 egonw joined #cdk
17:13 s9asad joined #cdk
17:32 slyrus left #cdk
17:37 slyrus joined #cdk
18:03 Gpox left #cdk
18:52 egonw left #cdk
19:16 maclean joined #cdk
19:27 maclean left #cdk
20:23 jbrefort left #cdk

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