Camelia, the Perl 6 bug

IRC log for #cdk, 2009-05-05

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

All times shown according to UTC.

Time Nick Message
05:13 jbrefort joined #cdk
05:38 egonw joined #cdk
05:56 bag joined #cdk
06:32 Gpox joined #cdk
06:38 Gpoks joined #cdk
06:56 edrin joined #cdk
07:00 egonw joined #cdk
07:20 mgerlich joined #cdk
07:20 s_wolf joined #cdk
07:27 * egonw will release CDK 1.2.2 today
07:27 egonw and possibly 1.3.0
08:09 shk3 joined #cdk
08:35 jpansanel hi
08:35 jpansanel moin
08:35 zarah saluton jpansanel
08:35 zarah ni hao jpansanel
08:41 egonw hi jpansanel
09:56 maclean joined #cdk
10:05 egonw CDK 1.2.2 released
10:20 maclean yay
10:20 maclean \o/
12:22 maclean joined #cdk
13:54 rajarshi joined #cdk
13:55 rajarshi good morning
13:55 egonw hi
13:55 zarah ni hao egonw
13:55 rajarshi what was that mail re module info pages about?
13:59 egonw mom
13:59 egonw nightly deleted those I had uploaded to pele
13:59 egonw it are these pages providing an overview of a module
13:59 egonw dependencies
13:59 egonw which classes
13:59 egonw and had some info on unit tests
13:59 rajarshi hmm, i don't think they went into sf cdk-1.2.x or master
14:00 rajarshi can you confirm?
14:00 egonw I think into master...
14:00 egonw mom
14:00 egonw yes, master 11 days ago
14:00 egonw http://cdk.git.sourceforge.net/git/​gitweb.cgi?p=cdk;a=commit;h=71492aa​ba87c58eba2fa417a156e30f1e318cafa
14:00 zarah egonw's link is also http://tinyurl.com/dkkahv
14:00 egonw and four earlier patches
14:00 rajarshi aah, ok
14:01 rajarshi these need to go into 1.2.x, right?
14:01 egonw yes
14:01 rajarshi ok, let me see about extracting these patches by hand
14:01 egonw so, that we can properly fix Nightly's JavaDoc
14:01 egonw which points to module pages
14:01 egonw rajarshi:
14:01 egonw cherry-pcick
14:01 egonw cherry-pick them
14:01 egonw did that this morning, worked fine
14:02 rajarshi cherrypick will generate patches?
14:02 egonw no:
14:02 egonw git cherry-pick <HASH>
14:02 egonw that will apply the patch from your current repos
14:02 egonw no matter what branch
14:16 rajarshi ok, done and applied to 1.2.x
14:19 rojasm joined #cdk
14:26 egonw rajarshi++
14:26 egonw thanx
14:26 rajarshi rojasm, did you get a chance ot look at the source links patch I had sent?
14:27 maclean egonw: I'm trying to help out someone (by mail) that wants to use the Java2DRenderer, unfortunately I never quite realised how crap it is
14:27 egonw :(
14:27 maclean so, I wonder : what is it I can do to get the new rendering code into people's hands?
14:28 egonw maclean: can you please formulat the design problems in a mail to cdk-jchempaint?
14:28 maclean er. why?
14:28 egonw maclean: first thing would be to make a proper clean patch series against cdk-1.2.x
14:28 egonw ummm
14:28 egonw I understood from your comment that the new jcp rendering code is crap...
14:28 egonw but that's not what you meant?
14:28 maclean no
14:29 maclean the OLD code (not the old old code)
14:29 maclean niels out's stuff
14:29 egonw que?
14:30 maclean well, that's what is available to people, isn't it?
14:30 rojasm not yet. I came now I check it now. Today is Befreitag day. Also alone in the office
14:30 egonw what is there still left that is not part of jcp-prim?
14:30 rojasm :(
14:30 egonw ha, rojasm is having a party in the lab :)
14:30 rajarshi :)
14:30 rajarshi what is befreitag day? beer? food?
14:32 egonw may 5
14:32 egonw liberation of NL from the nazis
14:32 egonw bevrijdingsdag
14:33 rajarshi aah
14:37 rojasm right bad spelling. I though naive of my part it was the same
14:38 rojasm At least the music is loud.
14:38 rojasm nobody is disturbed
14:40 egonw :)
14:40 egonw rojasm: that is not May 4
14:41 egonw when we remember the falling soldiers and others
14:41 egonw of WOII
14:41 egonw May 5 we now celebrate freedom
14:41 egonw actually...
14:41 egonw we should celebrate the Blue Obelisk next year :)
14:42 rajarshi indeed - 5 years next year (spring to be specific)
14:42 rojasm ei that sound much better.
14:43 egonw rajarshi: comments on patch emailed
14:44 rajarshi thanks
14:46 maclean anyway, egonw : ?
14:46 egonw maclean: huh?
14:46 egonw on the crap stuff?
14:46 maclean yup
14:47 maclean I'm concerned that people are trying to get the old code working, when it clearly doesn't
14:47 maclean you can't color atoms by type, it seems
14:47 maclean (for example)
14:47 egonw they should be trying to use rendering code from cdk-1.2.x
14:47 egonw I though I removed all
14:47 maclean hmmm.
14:47 egonw but I guess I have not?
14:47 egonw oops
14:47 egonw they should NOT be trying to use rendering code from cdk-1.2.x
14:47 maclean well, I'm confused. I will check what people actually have.
14:48 egonw either cdk-1.0.x rendering code
14:48 egonw or jcp-prim
14:48 maclean ah.
14:48 egonw those are the options they have
14:48 maclean well I don't know how they would know that, but okay.
14:49 * egonw is checking if there still are renderers in cdk-1.2.x
14:49 egonw and if there are, will remove them immediately
14:50 egonw mmm... there is
14:50 egonw OK, will remove that
14:50 egonw I guess I forgot to remove that earlier
14:52 maclean okay, that will be great, but what can I do to get the NEW stuff to the average user? As in the fabled merging of jcp-primary? I know you linked me to that wiki page, but...
14:52 maclean Should I just be making a patch of the whole renderer package?
14:53 maclean Or diffing to find the changes outside the control and render, or...
14:53 maclean re-writing the ControllerHub?
14:54 shk3 the renderer stuff (and controller) would need quite a few changes (comments etc.) for inclusion in 1.2/head
14:54 shk3 i think
14:55 maclean it's mixed. The Renderer class is well doc-ed. The others may not be.
14:55 shk3 egonw said he started with the diffs outside renderer/controller
14:55 shk3 how far did you proceed?
14:56 * maclean will write a short tutorial (with pictures!) on how to use the new rendering architecture, but it really needs to coincide with a tagged release.
14:57 maclean I am aware that 1.2.2 was only released at 11.05 am today.
14:57 maclean :)
14:59 CIA-42 cdk: shk3 * r14581 /jchempaint/trunk/ (4 files in 2 dirs): changed to work with updates in jcp-primary
15:03 egonw no, have not started with the patches yet
15:03 egonw what I said is that we first need to define a series of patches
15:03 maclean right, yes
15:03 egonw and that is something arvid and I started talking about
15:04 maclean anything I can do to help this process?
15:04 egonw and patches outside cdk.renderer and cdk.controller need to be defined first
15:04 egonw maclean: write that tutorial would very much help
15:04 maclean well no
15:04 egonw maclean: it would also help very much
15:04 egonw if you and arvid can work out and describe in the wiki
15:04 maclean that doesn't help get the patches into master
15:05 egonw how scaling and zooming is (supposed to) work(ing)
15:05 egonw it would
15:05 egonw as it helps figure out what patch sets we need to have
15:05 egonw so far, we have:
15:05 maclean well, okay, if you say so.
15:05 egonw serie 1: patches outside cdk.controller and cdk.renderer
15:06 egonw series 2: patches for interfaces of cdk.renderer
15:06 egonw series 3: implementation of cdk.renderer
15:06 egonw series 4: patches for interfaces of cdk.controller
15:06 egonw series 5: implementation of cdk.controller
15:06 maclean I would prefer it if the audience was mostly developers, so that users don't read it and say "now where is the code"...
15:07 maclean I mean, obviously I could provide instructions for chekcing out jcp-primary, but that is less than ideal
15:08 egonw yes, but we are not that far yet
15:08 egonw and there is still an API change to happen
15:09 egonw clean up and fixing of the controller hub
15:09 egonw which needs to be modularized
15:10 maclean what API change is this?
15:10 egonw right now the ControlerHub is one big blob
15:10 egonw this will need to be modules, just like the renderer
15:10 maclean yes, I know.
15:10 egonw with the current monolithic hub, a small editor is impossible
15:11 egonw which is one of the design goals which we failed to implement yet
15:11 egonw s/we/I/ if you like
15:11 maclean well, whatever. It hasn't happened. Could I do it?
15:12 * maclean is really quite keen to kick the new system out the door.
15:12 egonw getting the zoom/scroll/stuff working is more important at this moment
15:13 egonw rajarshi: ping
15:13 CIA-42 cdk: shk3 * r14582 /cdk/branches/jchempaint-primary/src/m​ain/org/openscience/cdk/applications/: this is an empty package
15:14 rajarshi yes
15:15 egonw apparently, I had forgotten to remove the final bits of renderer code from cdk-1.2.x
15:15 egonw any objections against removing it now?
15:15 shk3 don't think so
15:16 egonw rajarshi?
15:17 rajarshi egonw, yes?
15:17 egonw [17:15:15] <egonw> apparently, I had forgotten to remove the final bits of renderer code from cdk-1.2.x
15:17 egonw [17:15:22] <egonw> any objections against removing it now?
15:17 rajarshi fine with me
15:17 rajarshi anything to get the new renderer code in 1.2.x or master :)
15:18 egonw :)
15:18 egonw well, this is just to prevent people from being tempted to use those bits of renderer code
15:18 egonw instead of using jcp-prim or cdk-1.0.x
15:18 rajarshi yes, makes sense
15:22 egonw ok, cdk-1.2.x updated
15:22 egonw I should make a 1.2.3 release
15:22 shk3 I just had a look at a diff of jcp and 1.2.x
15:22 egonw will do that tomorrow
15:22 shk3 unfortunately, it is quite a lot
15:22 shk3 shall we device it between us?
15:22 shk3 packagewise or so?
15:24 egonw we need a git repos for this
15:24 egonw where we can have branches for the different series of patches
15:24 egonw which can be pulled from one by one to reproduce jchempaint-primary
15:24 maclean wow, git sure does make things easier!
15:24 maclean :P
15:24 shk3 sounds like making a worse mess out of a mess
15:24 egonw but we need jcp-prim in a git repos in the first place so that we can cherry-pick every each and single patch
15:25 egonw shk3: yes, but you are not yet in the git world
15:25 maclean oo.kaa.y
15:25 egonw shk3: so, I suggest then, leave it to me and arvid
15:25 maclean no, no that's fine.
15:25 maclean I will try and help, any which way.
15:26 shk3 I just don't see the point, I must admit
15:26 shk3 shouldn't we transfer stuff either
15:26 shk3 from primary to 1.2.x
15:26 egonw maclean: right now, you can best help by helping arvid work out scrollbar/zoom/etc for jcp in bc2
15:26 shk3 or the other way round
15:26 shk3 till diff gives no difference?
15:26 shk3 or is that too simplistic?
15:26 egonw shk3: yes, but in such a way that rajarshi and miguel can review it in reasonable manner
15:27 shk3 but single patches for 1.2.x could be reviewed
15:27 egonw as they are in jcp-prim now?
15:27 egonw sure not
15:27 shk3 and what's ported from 1.2.x to primary is reviewed
15:27 shk3 no, via submitting patches
15:27 egonw because patches are reset, others changed, ...
15:28 egonw shk3: what do you want to submit then?
15:28 shk3 I would think if X finds a change in primary
15:28 shk3 which is not yet in 1.2.x
15:28 egonw what change?
15:28 egonw a commit in jcp-prim?
15:28 shk3 yes
15:28 shk3 any change in there
15:28 egonw see above
15:29 shk3 let me explain
15:29 egonw yes, please
15:29 shk3 he ports this to his personal 1.2.x repository
15:29 shk3 and makes a patch
15:29 egonw hahaha
15:29 shk3 which is treated like any other patch
15:29 shk3 ok, I am glad I made you laugh
15:29 egonw porting current jcp-prim to cdk-1.2.x is rather non-trivial and not the reviewers task
15:30 shk3 I am sure you will do it
15:30 shk3 not the reviewer
15:30 egonw he == us/me
15:30 egonw yes
15:30 shk3 the person who checks differences
15:30 egonw I don't often speak 3rd person about myself... was confused
15:30 egonw yes, but that is what this wiki page is about
15:31 egonw sorry, for interrupting
15:31 egonw that was not appropriate
15:31 shk3 go ahead with git and tell me if I can help
15:32 shk3 I will look at jcp bugs in the meantime
15:32 s_wolf joined #cdk
15:34 rojasm joined #cdk
15:34 maclean egonw : why does this jcp-primary merge depend on scrolling working in bc2?
15:35 maclean I don't mean to be argumentative about this, I'm just trying to understand, you see.
15:35 egonw you mean, scrolling is supposed to work in jcp-prim?
15:35 CIA-42 cdk: shk3 * r14583 /jchempaint/trunk/src/main/org/openscienc​e/jchempaint/application/JChemPaint.java: made the workaround for readers to work for CMLReader as well
15:36 maclean sigh. what?
15:36 maclean I am aware that scrolling and zoming doesn't work right.
15:36 maclean it is better in swing-jcp, than in bc-jcp.
15:36 maclean it may not be fixed to everyone's satisfaction in my lifetime.
15:38 egonw ok, so here we have a patch series that people may want to apply or not:
15:38 egonw partial scrolling/zooming support
15:39 maclean the real hard problem may be that scrolling in java assumes a certain paint cycle.
15:40 maclean and it may be that our 'clever' painting is fucking with that system.
15:40 maclean for bioclipse, I would even suggest rendering to an image, then painting that image.
15:40 maclean custom double-buffering, in other words.
15:41 maclean in the hope that this will scroll properly.
15:42 * maclean will do as he is told, and write the damn tutorial.
15:43 egonw since I don't want you or me to run into trouble... I am *never* telling anyone to do anything
15:43 egonw I am boss to no one
15:43 egonw (just the regular disclaimer)
15:43 maclean :) I didn't take it that way.
15:43 egonw good :)
15:43 maclean but you are benign dictator
15:43 maclean (which is a good thing)
15:44 egonw yes, we need people at the gate
15:44 egonw to prevent crap to sneak in
15:44 egonw which is the sad truth
15:44 egonw another truth is that I need to leave now
15:45 egonw and go home
15:45 egonw will be back later
15:45 maclean cya.
16:24 jbrefort joined #cdk
17:23 CIA-42 cdk: shk3 * r14584 /jchempaint/trunk/src/main/​org/openscience/jchempaint/ (4 files in 3 dirs): cleaned up the save/as
17:49 CIA-42 cdk: shk3 * r14585 /jchempaint/trunk/src/main/org/openscience/jch​empaint/applet/JChemPaintAbstractApplet.java: better comment
18:14 edrin joined #cdk
18:15 egonw joined #cdk
18:50 timvdm joined #cdk
19:14 CIA-42 cdk: gilleain * r14586 /cdk/branches/jchempaint-primary/src/main/o​rg/openscience/cdk/renderer/Renderer.java: Added a getModelCenter method to mirror the getDrawCenter method.
19:55 jbrefort joined #cdk
21:00 timvdm joined #cdk

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