Camelia, the Perl 6 bug

IRC log for #cdk, 2011-10-04

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

All times shown according to UTC.

Time Nick Message
01:13 Alchimiste joined #cdk
01:13 zarah joined #cdk
01:13 CIA-38 joined #cdk
04:00 egonw__ joined #cdk
04:38 sneumann joined #cdk
04:58 egonw__ joined #cdk
05:24 sneumann joined #cdk
05:52 jbrefort joined #cdk
06:11 egonw joined #cdk
06:36 egonw joined #cdk
06:40 Gpox joined #cdk
07:11 mgerlich joined #cdk
07:22 s_wolf joined #cdk
07:42 jonalv joined #cdk
08:15 jonalv joined #cdk
08:58 maclean joined #cdk
08:58 maclean monning
08:58 maclean
08:59 Gpox joined #cdk
10:03 egonw moin
10:03 zarah oh hai egonw
10:11 maclean joined #cdk
10:43 Alchimiste moin
10:43 zarah hello Alchimiste, you fantastic person you
11:16 maclean joined #cdk
11:25 egonw maclean: thanx for your answers on the MLs
11:25 egonw maclean++
11:25 maclean n/p
11:25 maclean currently looking at the UIT/aromaticity problem.
11:26 egonw I think it's just a difference in aromaticity perception
11:27 maclean but there's no call to aromaticity detection in the code (that I can see)
11:27 maclean weird.
11:28 egonw ah, but, but...
11:28 egonw the UIT uses it still
11:28 egonw here's the deal
11:28 egonw C=C-C will match toluene in two ways
11:28 egonw that is overcome with capturing delocalization
11:28 egonw so, there's your aromaticity
11:28 egonw brb
11:29 maclean ok
11:56 maclean right. was the smiles parser's fault.
11:57 egonw oi, another one...
11:57 egonw that code has seen more bug fixes than LOCs
11:57 egonw tell me...
11:57 maclean Not necessarily a bug.
11:58 maclean There is a setPreserveAromaticity method
11:58 maclean That is false by default, so if you use the SMILESReader to read smiles, you get aromaticity detection.
11:59 maclean Oh... Bother.
11:59 maclean But then if it's set to true, you don't get atom typing.
12:00 egonw oh...
12:00 egonw indeed
12:00 egonw line 257-270
12:00 egonw but that should happen indepently of that value set
12:01 egonw mmm... what happened to the git blame in my eclipse...
12:03 maclean try "eclipse blame 'git blame'" to find out.
12:04 maclean Heh. "eclipse blame" would be awesome. "Eclipse : why are the tooltips not working for the java standard library methods?"
12:04 maclean "Eclipse :  why do you no longer cycle between occurrences when I press cmd-dot?"
12:05 egonw hahah :)
12:05 egonw for the latter there is a 'reset' optino
12:05 egonw some plugins mess it up
12:06 maclean A reset menu option?
12:06 egonw no, somwhere in the prefs
12:06 egonw but umm...
12:06 egonw atom typing should always be done, agreed?
12:06 egonw does that fix the reported problem?
12:11 maclean Oh, sorry. Yes it does.
12:11 egonw well analyzed!
12:12 egonw are you making patches? unit test + fix?
12:12 egonw (pref separate)
12:12 maclean http://pastebin.com/HC5cdhAW
12:13 maclean Patches? What is patches? (that won't work, will it...)
12:13 egonw oh, it's not the SmilesParser that is wrong?
12:14 maclean Well, it might be wrong, I don't know really. I know that this workaround works...
12:15 maclean Cl-C=N should match  C1=CC=C2C(=C1)C(=O)C3=C(C2=O)C(=C(C=C3NC4=C​C(=C(C=C4)NC5=NC(=NC(=N5)NC6=CC=C(C=C6)S(=O​)(=O)[O-])Cl)S(=O)(=O)[O-])S(=O)(=O)[O-])N
12:15 maclean really.
12:19 egonw crap, and I signed off that patch
12:19 jbrefort joined #cdk
12:19 maclean which patch?
12:20 egonw the patch that moved atom type perception to only be applied when arom from SMILES is not taken literally
12:20 maclean ae21ee2?
12:20 egonw mmm.... 5 papers published this year so far :)
12:20 egonw 2 first author
12:20 maclean "Smiles parser setting to preserve aromaticity as provided in the Smiles String itself."
12:21 maclean Congrats.
12:21 egonw right
12:21 egonw that option should not mean that atom types should not be perceived when set to false
12:21 maclean Achivement unlocked : first author hat!
12:21 egonw that option should not mean that atom types should not be perceived when set to true
12:21 egonw no, false
12:21 egonw crap those double negatives
12:22 egonw yeah, that's a cool idea... stackoverflow-style badges for scientists! /me wants that
12:22 egonw with bronze, silver, gold
12:23 maclean Hmmm. Could be good. Ohlo.com kept some kind of commit score, I seem to remember.
12:23 egonw "first author paper": 1=bronze, 10=silver, 100=gold
12:23 maclean hmm. not ohlo.com, seemingly.
12:23 egonw "first author in nature, science, or cell": 1=bronze, 5=silver, 10=gold
12:23 egonw we only need catchy names
12:24 egonw ohloH.something
12:24 maclean dot net
12:24 maclean http://www.ohloh.net/p/egonws_cdk
12:24 egonw "writer" badge: 1=bronze, 10=silver, 100=gold
12:24 egonw "cool writer": 1=bronze, 5=silver, 10=gold (cool == nature, cell, science)
12:25 egonw what, I have not claimed this one???
12:25 egonw where's my ohloh pwd...
12:25 maclean I thought you used to have.
12:25 maclean Er...used to have done so.
12:25 maclean nvm
12:26 egonw but I don't get this project anyway... "egonw's cdk"??
12:26 egonw instead, we just have: https://www.ohloh.net/p/cdk
12:26 maclean Hmmm. Weird.
12:26 maclean http://www.ohloh.net/p/egonws_cdk/estimated_cost
12:27 maclean Gah. Paste text, dammit!
12:27 maclean Estimated project cost: $5,615,089
12:27 egonw some hamster created it
12:27 egonw (seriously)
12:28 maclean Some...hamster?
12:28 maclean CDK : "No managers have claimed this project yet..."
12:29 egonw "The Ohloh Hamster"
12:29 egonw that account created the project
12:29 egonw https://www.ohloh.net/accounts/ohloh_slave
12:30 maclean Oh. A bot, in other words.
12:30 maclean I think the /cdk vs egonw's_cdk distinction might be due to github cdk/cdk egonw/cdk ?
12:31 egonw ok, "egonw's cdk" is now marked as duplciate
12:31 egonw yes, I think so too
12:31 maclean The cdk/cdk project cost $6 million, so egonw/cdk is undercutting by a million dollars!
12:32 maclean Anyway, enough distractions. Back to result paging and div accordions.
12:37 egonw btw, impressed and pleased with the amount of traffik we are seeing on the CDK MLs, but happy that others helps out with answering too!
12:45 maclean "This new version of Jmol has been in develoment since July of 2010 and contains over 200 new features..." Wow. Just wow.
14:21 egonw joined #cdk
14:42 egonw joined #cdk
14:48 egonw_ joined #cdk
15:17 egonw__ joined #cdk
16:56 egonw__ joined #cdk
17:01 sneumann joined #cdk
17:16 sneumann_ joined #cdk
18:02 jbrefort joined #cdk

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