Camelia, the Perl 6 bug

IRC log for #cdk, 2009-10-02

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

All times shown according to UTC.

Time Nick Message
03:07 slyrus_ joined #cdk
03:13 mgerlich joined #cdk
03:13 Conrad joined #cdk
03:44 slyrus_ hmm.. calling SmilesParser.parseSmiles doesn't add implicit H's?
04:29 jbrefort joined #cdk
04:30 slyrus_ excuse the clojure syntax, but... what should this return? (MolecularFormulaManipulator/getNaturalExactMass (MolecularFormulaManipulator/getMolecularFormula (parse-smiles "CC")))
04:32 slyrus_ err... let me try that again: (MolecularFormulaManipulator/getNaturalExactMass (MolecularFormulaManipulator/getMolecularFormula (.parseSmiles (SmilesParser. (DefaultChemObjectBuilder/getInstance)) "CC")))
04:48 slyrus_ "The hydrogens must be implicit." <-- this comment is rather uninformative... what happens if they are or are not implicit?
04:48 slyrus_ (from MolecularFormulaManipulator.java)
04:49 slyrus_ cosmetic change request: remove CDK from CDK*, esp. CDKHydrogenAdder...
05:45 egonw joined #cdk
05:46 slyrus_ hey egonw
05:46 egonw good morning
05:54 slyrus_ wow, this is a lot of work to get the mass of ethane: (let [mol (.parseSmiles (SmilesParser. (DefaultChemObjectBuilder/getInstance)) "CC")]
05:54 slyrus_ (AtomContainerManipulator/conv​ertImplicitToExplicitHydrogens mol)
05:54 slyrus_ (MolecularFormulaManipulator/getNaturalExactMass
05:54 slyrus_ (MolecularFormulaManipulator/getMolecularFormula
05:54 slyrus_ mol)))
05:55 slyrus_ but at least it works...
05:55 egonw :)
06:02 egonw it's not that bad
06:03 egonw but I think we used to have a method too to calculate the exact mass for a IAtomContainer directly
06:03 egonw removing one step
06:06 slyrus_ I couldn't get that to work with the implict hydrogens
06:06 slyrus_ or, should I say, it wasn't counting the mass of the hydrogens when I tried that
06:06 egonw that's a step you need to do anyway...
06:06 egonw or, alternatively...
06:06 egonw just calculate the number of 'missing' hydrogens
06:06 egonw and add that multiplied by the exact mass of hydrogen
06:07 egonw right
06:07 egonw it's not making assumptions...
06:07 slyrus_ easier to do it my way, I suppose...
06:07 egonw "oh, perhaps he forgot to add hydrogens"
06:07 slyrus_ easy enough to write a little wrapper function to hide the molecularformula stuff...
06:07 egonw yes
06:10 bag joined #cdk
06:15 slyrus_ oh noes... the smiles parser ignores configuration around double bonds :(
06:17 egonw it's on our schedule...
06:17 egonw sorry about that...
06:17 egonw we have the data model
06:17 egonw but not the parsing from SMILES
06:17 egonw IIRC
06:17 slyrus_ ok, thanks
06:18 slyrus_ not having the data model for this in my lisp-based chemistry stuff was one of the things that motivated me to abandon it for cdk... that and the lousy homegrown algorithm for trying to layout 2d structures...
06:19 slyrus_ I'm liking clojure though and i'm starting to figure my way around the nooks and crannies of cdk
06:19 egonw please do feel free to communicate your trouble...
06:20 egonw we are very open to that
06:20 egonw we know it's not perfect
06:20 egonw but we only have very limited resources
06:20 egonw which why not everything is fixed or implemented yet
06:23 slyrus_ my frustrations seem to come from three main sources: 1) the initial wrapping of my head around the various cdk/jcp-primary source repos, 2) a general distaste for the style of code found in large java code bases vs. what i'm used to in lisp and 3) a lack of documentation for cdk
06:23 slyrus_ not to complain too much -- there's certainly a lack of documentation for my own code :)
06:24 slyrus_ but, after a couple days, i feel like i'm starting to get the hang of things
06:37 egonw these are fair comments
06:38 egonw we have submitted last week a grant application to solve these problems
06:38 slyrus_ cool! good luck!
06:39 egonw it will take until April next year before we know the outcome
06:39 egonw but we have reasonable chances
06:39 egonw anyways...
06:39 egonw please do file bug reports
06:39 egonw for anything
06:39 egonw those get priority in getting fixed
06:39 egonw and fixing a bug makes people feel good
06:39 slyrus_ i should probably be able to figure this out by looking at the source, but are there any facilities for IUPAC naming of molcules (and vice versa)?
06:40 slyrus_ true...
06:40 egonw there is a framework developed originally in Manchester
06:40 egonw but, it implements only a very minor subset of rules
06:42 slyrus_ speaking of bugs... where is the bug database? the link from the mediawiki page doesn't seem to work
06:43 egonw oh, what page?
06:43 egonw will fix it...
06:43 egonw the new SF project front page is crap
06:43 slyrus_ search for bug here: http://sourceforge.net/apps/mediaw​iki/cdk/index.php?title=Main_Page
06:43 egonw hides all the good stuff, like the BTS
06:43 zarah slyrus_'s link is also http://tinyurl.com/yce4l93
06:43 egonw ok, that works
06:44 egonw mom, I'll look up the URL
06:44 egonw http://sourceforge.net/tracker/?group_id=20024
06:44 zarah egonw's link is also http://tinyurl.com/y9pr9ts
06:44 slyrus_ ah thanks...
06:45 slyrus_ I forgot to add gripe 4: i generally hate sourceforge :)
06:45 egonw wikis are bad regarding keeping them correct
06:45 egonw it's not perfect, but free, and offer good stability for many kinds of services
06:45 slyrus_ wikis are nice, but there no replacement for a proper website/hoomepage thing
06:45 slyrus_ s/there/they're/
06:46 slyrus_ that's true, about sf.net... but, still, drives me crazy
06:46 slyrus_ oh, and 5) all the tabs in the .java files drive me crazy too :)
06:46 egonw ah, yes, those are evil, agreed... :)
06:47 slyrus_ well, I'm sure i could stay up and think of more things to complain about, but I should get some sleep. looking forward to some more cdk hacking time this weekend.
06:48 egonw ack
06:48 egonw when you file bug reports, wish lists, patches, whatever, please use a SF account, so that I can see it was you who filed it...
06:48 slyrus_ will do
07:02 Gpox joined #cdk
07:31 s_wolf joined #cdk
08:52 egonw joined #cdk
12:16 jpansanel moin
12:16 zarah privet jpansanel
12:20 shk3 joined #cdk
16:38 slyrus_ joined #cdk
19:20 egonw joined #cdk
19:20 jbrefort joined #cdk
19:20 Conrad joined #cdk
21:28 egonw joined #cdk
21:28 Conrad joined #cdk

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