Camelia, the Perl 6 bug

IRC log for #cdk, 2008-02-11

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

All times shown according to UTC.

Time Nick Message
00:09 CIA-19 jmol: aherraez * r8987 /trunk/Jmol-web/source/doc/jmol/ (18 files): update of applet in website (from 11.3.51 to 11.4RC4)
05:22 CIA-19 cdk: egonw * r10050 /trunk/cdk/src/org/openscience/cdk/test/atomtype/ (AbstractAtomTypeTest.java CDKAtomTypeMatcherTest.java): Test that after assigning atom types, perception is identical as befor those assignments
06:19 jbrefort joined #cdk
06:58 egonw joined #cdk
07:24 CIA-19 cdk: egonw * r10051 /trunk/cdk/src/org/openscience/cdk/test​/atomtype/CDKAtomTypeMatcherTest.java: More testing that perception before and after atom type property assignments are the same
07:38 rojasm joined #cdk
07:38 rojasm moin
07:38 egonw hi rojasm
07:38 egonw I got some updates...
07:39 egonw worked on the atom typing this weekend
07:39 rojasm hi egonw. I am reading still the mail. perfect
07:39 rojasm sorry this weekend I was in Cologne
07:44 egonw ah, nice
07:44 rojasm how deep do yo want to go describing atomtypes?
07:45 egonw not much deeper than currently...
07:45 egonw what do you have in mind?
07:45 rojasm because It is not the same C(H)(H)(H)(H) than C(F)(H)(H)(H)
07:45 rojasm according the oxidation state
07:45 rojasm -4 and -3
07:46 egonw right...
07:46 egonw that's a level I do not want to go into...
07:46 rojasm ok
07:46 egonw oxidation numbers are rather fuzzy...
07:46 rojasm :)
07:46 egonw I think I can do better
07:46 egonw seriously...
07:46 rojasm I supose too
07:46 egonw I've been reading my inorganic chem book :)
07:47 egonw and it has a table with rules on how to determine the ox number
07:47 egonw DT-like
07:47 egonw that's not good
07:47 egonw anyway...
07:47 egonw it's implicit in the current design, I think
07:47 egonw I'm thinking of an improvement... but will talk with you about that later...
07:48 egonw the current system seems to hold quite well...
07:48 egonw the only thing that seems to be missing, is a way to identify the difference between .sp2 and .planar
07:48 egonw an example will have to do...
07:49 egonw the concept of the difference is quite well defined, so I don't worry about that
07:50 rojasm perject I am stil checking the new updates :)
07:50 rojasm cafe first
07:50 egonw :)
07:51 egonw busy today... but will be around for chats
08:34 rojasm egonw: the atomtype C.plus.sp1 is not recognized by percieveAtomTypesAndConfigureAtoms as a valid atomtype
08:34 egonw please file a bug report
08:34 rojasm ok
08:34 egonw won't have time to look at those kind of details before this evening
08:34 rojasm ok, not problem
08:35 egonw btw, did you notice all test examples you added are no longer failing?
08:35 rojasm yes
08:35 egonw also, note the severe bug...
08:35 egonw that after calling percieveAtomTypesAndConfigureAtoms(), a subsequent atom typing fails in certain cases...
08:36 egonw filed as bug report...
08:36 egonw will fix that with high priority...
08:36 egonw meanwhile, keep adding tests cases to cdk.test.atomtype
08:39 Annzi joined #cdk
08:45 rojasm egonw: I am thinking about! is it possible to have a C#[C-]?
08:45 egonw yes, should be possible I think
08:46 egonw two sp1 orbitals... one with a sigma bond, one with a LP
08:46 egonw and py and pz free for the triple bonding
08:46 egonw not very stable, at least not in water...
08:46 egonw I think the species will be quite basic
08:46 rojasm yes. But the there is not hybridization
08:47 egonw no resonance, you mean?
08:47 rojasm sorry.
08:47 rojasm hybridization sp1
08:48 de_maas joined #cdk
08:49 de_maas left #cdk
08:49 rojasm sorry. all false! I wante to mean the C#[C+]?
08:50 egonw ok, same story... sp1
08:50 egonw but now an empty orbital instead of the LP
08:50 rojasm not there is not hybridization
08:50 egonw why not?
08:50 rojasm because it has only one electron
08:51 egonw btw, I'd say the resonances into [C+]=C|
08:51 rojasm which can take part
08:51 egonw particularly if the left C has other carbons attached... or oxygen
08:51 rojasm but ressonance is not hybridization.
08:52 rojasm for hybridization you need at less two electrons which are not in pibonds
08:52 rojasm in this case C#[C] only has one
08:53 rojasm you can not create a sp
08:53 rojasm is it correct?
08:54 rojasm (09:52:55) rojasm: in this case C#[C+] only has one
08:56 rojasm in this case I would write as C.plus.s
08:56 rojasm but I don't know
09:02 egonw I don't understand your argument...
09:03 jonalv joined #cdk
09:03 egonw you mean that an empty bond is not very favorable?
09:03 egonw agreed.... I'm sure the charge will relocate to the other carbon, by dropping one bond order
09:03 egonw C#[C+] -> [C+]=C|
09:05 rojasm No I see what you say. That the C#[C+] will not exist as this but as [C+]=C| which it is a sp1
09:05 rojasm No = know
09:06 jpansanel moin
09:07 jpansanel egonw : ok for the patch
09:07 rojasm But so we are defining the atomtype according what structure is more stabil!
09:07 rojasm and not what we have
09:09 egonw rojasm: keep in mind that atom typing is to some extend about what is reasonable...
09:10 egonw but clearly in this case also needed for represention reaction mechanisms...
09:10 egonw about stability... try MM or QM
09:12 jpansanel egonw : in Ruby on Rails, if you have a database class named "compound", does the database needs to be called "compounds" ?
09:12 egonw jpansanel: no idea
09:12 egonw I'm not a RoR guy
09:13 * egonw has to go into serious meeting preparation mode now
09:13 egonw will be offline until late this afternoon
09:13 rojasm I am thinking that the good think about describing is the atomtype will be the geometry! in this case it will be not the same when we have C#[C+] sp1 or s hybridization according a nucleophilic reaction
09:14 rojasm cu later
09:15 jpansanel bye
09:59 CIA-19 cdk: miguelrojasch * r10052 /branches/miguelrojasch/reaction/src/org/openscien​ce/cdk/test/atomtype/ReactionStructuresTest.java: added new test which fails for O.sp2.radical. The radical is not coming from the lone pair electron but for the pi bonding.
10:03 egonw rojasm: that would then be a O.planar.radical, not O.sp2.radical
10:04 rojasm ups, ok. sorry.
10:04 egonw yeah, no worries...
10:04 egonw those names are not meant to be self-explanatory
10:04 rojasm ist there some difference beetween C[O*] and C[*+?]
10:04 rojasm C[O*] and C[O*+]
10:05 egonw yeah, one charged, the other not
10:05 rojasm good
10:05 egonw one resulting from electron withdrawn from pi bond, other from LP
10:05 rojasm and the atomtype should be O.planar.radical and the other ?
10:05 egonw O.sp2.radical
10:06 egonw ummm
10:06 egonw O.plus.sp2.radical, I guess
10:14 CIA-19 cdk: miguelrojasch * r10053 /branches/miguelrojasch/reaction/src/org/openscien​ce/cdk/test/atomtype/ReactionStructuresTest.java: modified atomtypeName. Now not errors
10:15 CIA-19 cdk: miguelrojasch * r10054 /branches/miguelrojasch/reaction/src/org/openscien​ce/cdk/test/atomtype/ReactionStructuresTest.java: removed prints not necessaries
10:23 rojasm egonw for the N with the structure as [N*]=[C*] the atomtype assigned N.sp1 but don't make reference about the radical
10:23 rojasm is it correct?
10:23 egonw no, does not sound correct
10:23 egonw file a bug report, and I'll look at it
10:23 rojasm ok
10:39 CIA-19 cdk: miguelrojasch * r10055 /branches/miguelrojasch/reaction/src/org/openscien​ce/cdk/test/atomtype/ReactionStructuresTest.java: added new Test for N.
10:51 carsten joined #cdk
10:57 rojasm egonw: after to be membership of xcms in google groups, is it not possible to receive the discussion as emails like sourcefourge list?
10:57 rojasm do I have to read each time in web?
11:10 egonw you can have them emailed
11:10 egonw look up the group settings...
11:10 egonw but maybe you need a google email account for that...
11:11 rojasm I have a google mail account but I read that you have to sent a mail
11:11 rojasm Welcome to the xcms list
11:11 rojasm This list is for discussion and questions about the use of xcms.
11:11 rojasm This is just an archive of messages from the Scripps xcms e-mail list. To post a message, you must first subscribe using the Scripps e-mail server. To do so, send a message to majordomo@scripps.edu with the words "subscribe xcms" in the body of your message.
11:12 egonw sorry... going into stress mode now...
11:12 egonw will be back later
11:12 rojasm not problem
13:18 alexandros_p joined #cdk
13:29 alexandros_p left #cdk
13:47 rojasm egonw: around?
13:55 egonw sort of...
13:56 rojasm Not important. I was only looking for some paper that you know as reference to predict protonation. But I found some thing
13:57 rojasm other thing. did you read the bug that I summited : [cdk-bugs] [ cdk-Bugs-1891065 ] construction of mechanism
13:57 egonw no
14:00 rojasm Because now a lot of reaction are being overlaping
14:06 jonalv joined #cdk
14:06 rojasm egonw: when do you have exactly your defense?
14:08 egonw April 2, 13:30
14:08 rojasm perfect
14:20 masak joined #cdk
15:05 CIA-19 cdk: miguelrojasch * r10056 /branches/miguelrojasch/reaction/src/org/openscien​ce/cdk/test/atomtype/ReactionStructuresTest.java: new test for N.plus.sp3.radical which is not found (C[N*+])
15:07 rojasm egonw: skype!
15:08 jpansane1 joined #cdk
15:53 ilbot2 joined #cdk
15:53 Topic for #cdkis now The Chemistry Development Kit (http://cdk.sf.net) | JUnit: 5214, 77 fails + 44 errors (2.3%) | CDK 1.0.x TODO: see http://cdk.sf.net/wiki/ | Logs at http://moritz.faui2k3.org/​irclog/out.pl?channel=cdk
16:14 jonalv joined #cdk
16:16 CIA-19 cdk: miguelrojasch * r10058 /branches/miguelrojasch/reaction/src/o​rg/openscience/cdk/test/reaction/type/ (2 files): added more test and specified inchi ID
16:18 CIA-19 cdk: miguelrojasch * r10059 /branches/miguelrojasch/reaction/s​rc/org/openscience/cdk/reaction/ty​pe/ElectronImpactPDBReaction.java: perception of atomtype to verify if the product is reasonable. Removing possibility to handle single bonds. It is now processed by ElectronImpactSDBReaction.
18:34 rojasm joined #cdk
18:38 egonw_ joined #cdk
18:40 egonw_ hi rojasm
18:50 Annzi left #cdk
18:54 CIA-19 cdk: egonw * r10060 /trunk/cdk/src/org/openscience/cdk/test​/atomtype/CDKAtomTypeMatcherTest.java: All tests now do atom type perception before and after assigning atom type props
19:16 rojasm joined #cdk
19:23 jpansane1 egonw : can I pursue the junior jobs ? Or is it better to reserve it to other people ?
19:51 egonw_ jpansanel: no, please go ahead...
19:51 egonw_ I'll make sure to write up enough JJs to keep you and others busy...
19:51 egonw_ seriously, there is plenty to do...
19:53 jpansanel cool :)
19:54 rojas1 joined #cdk
19:57 * egonw_ is going offline
20:02 CIA-19 jmol: aherraez * r8988 /trunk/Jmol-web/source/doc/ (22 files in 2 dirs): updated promotion script (demo in home page)
20:36 jbrefort joined #cdk
21:42 CIA-19 jmol: nicove * r8989 /trunk/Jmol-web/ (cache.properties source/doc/fah/fah-projects.xml): Folding@Home
21:48 rojasm joined #cdk
22:48 CIA-19 cdk: miguelrojasch * r10061 /branches/miguelrojasch/react​ion/src/org/openscience/cdk/ (2 files in 2 dirs): added new method getHeavyElements_IMolecularFormula to supply the MFTools class. And corresponding test.

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