Camelia, the Perl 6 bug

IRC log for #bioclipse, 2009-07-10

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

All times shown according to UTC.

Time Nick Message
06:12 olass joined #bioclipse
06:24 egonw joined #bioclipse
06:26 egonw wow
06:27 egonw normally I can reasonably anticipate effects... but this cought me by suprise...
06:29 olass egonw: you clearly did not explain enough
06:29 olass and I thoung jonalv knew git as well
06:29 olass so if he makes that comment
06:30 olass well
06:30 olass glad you sorted that out
06:30 egonw ah, ok...
06:30 egonw phew :)
06:30 egonw anyway... good point...
06:30 egonw we need a git.bioclipse.net
06:30 egonw listing an overview of the main git repositories...
06:30 egonw whatever those URLs are
06:34 olass right
06:41 egonw btw, the cheminfo repos is at:
06:41 egonw http://github.com/egonw/bioclip​se.cheminformatics/tree/master
06:41 zarah egonw's link is also http://tinyurl.com/ms7fct
06:42 egonw olass: can you please try to fork it to your account?
06:42 olass if you explain how, yes
06:43 olass do I need to create a repos?
06:43 olass is there a doc on the cdk wiki for this?
06:44 egonw this is very simple...
06:44 egonw no, no docs...
06:44 egonw you'll see why...
06:44 egonw http://github.com/egonw/bioclipse.core/tree/master
06:44 zarah egonw's link is also http://tinyurl.com/kn37tf
06:44 egonw please go to that link
06:44 olass yes?
06:44 egonw mom
06:45 egonw do you see the 'fork' button
06:45 egonw right next to 'bioclipse.core' ?
06:46 olass grr, look at stefans commit message
06:46 egonw yeah, sorry about that
06:46 olass you should not be sorry, he should
06:46 olass pressing fork
06:46 egonw nice incentive to make a commit ? :)
06:46 egonw that's all
06:46 olass ok, and now?
06:46 egonw well, regarding the forking that is...
06:46 egonw OK
06:47 egonw did you install a git client
06:47 egonw ?
06:47 olass I have git on CL yes
06:47 egonw OK, do you see the "Public Clone URL" ?
06:47 egonw don't use that one
06:47 egonw that's the read only one
06:47 egonw use the 'Your Clone URL'
06:48 egonw lot's of guides here:
06:48 egonw http://github.com/guides/home
06:48 olass how do I use it?
06:48 egonw check 'Setting Up'
06:49 egonw you'll need to share your SSH key
06:49 egonw just like with SF
06:49 egonw follow the 'Setting Up' step by step
06:49 olass I have not shared my SSH key with SF
06:50 olass probably need to generate an SSH key
06:50 egonw http://github.com/guides/tell-git​-your-user-name-and-email-address
06:50 zarah egonw's link is also http://tinyurl.com/6hgh2h
06:50 egonw do that one first
06:50 olass will follow http://github.com/guides/setting-up-a-​remote-repository-using-github-and-osx
06:50 zarah olass's link is also http://tinyurl.com/d7bs3c
06:50 olass ok?
06:51 olass should I use my SF account there or my github user account?
06:51 egonw yes, but only 'Create your public keys'
06:52 egonw yes -> not to your last question
06:52 olass what email should I use with git?
06:52 olass who sees that?
06:52 egonw the world
06:52 olass ok, so it will give me spam
06:53 olass right
06:53 olass will not use farmbio then
06:53 egonw see my blog about attribution, copyright and license
06:53 olass why?
06:53 egonw it is actually sort of important, because people will rely on being able to reach you on that address
06:53 olass hmm
06:53 olass will use my gmail account then I guess
06:53 egonw (but I will not complain if you will use your SPAM address)
06:54 egonw I'm sure they will find you if they need to
06:54 egonw sounds good
06:55 olass so I did global user name
06:55 olass and global email
06:55 olass enough?
06:55 egonw http://github.com/guides/tell-git​-your-user-name-and-email-address
06:55 zarah egonw's link is also http://tinyurl.com/6hgh2h
06:55 egonw that one also speaks of github.user
06:56 egonw and github.token
06:56 egonw but don't remember having done that, but I suggest you do those too...
06:56 egonw just like
06:56 egonw git config user.name "Ola Spjuth"
06:56 egonw and
06:56 egonw git config user.email foo.bar@gmail.com
07:01 olass hmm Ola Spjuth should be olas, right?
07:01 olass since that is my user name
07:01 olass ok, told github my ssh
07:01 olass now?
07:02 egonw no, user.name is just what shows up in the commit messages
07:02 egonw User.Name <user.email>
07:02 olass ok
07:02 egonw I have:
07:02 egonw user.email=egon.willighagen@gmail.com
07:02 egonw user.name=Egon Willighagen
07:02 egonw use:
07:02 egonw git config --list
07:03 olass ok
07:06 olass egonw: should I now do: Specify GitHub as a remote repository?
07:07 olass no
07:07 olass egonw: please advice
07:07 olass next step
07:11 olass egonw: ping
07:12 olass should I now enter: git remote add github git@github.com:git_username/projectname.git ?
07:12 egonw sorry... had to make breakfast for the kids
07:12 egonw back now
07:12 olass (a lot of git in that sentence)
07:12 egonw neither
07:12 egonw skip that lot
07:12 egonw http://github.com/olas/bioclipse.core/tree/master
07:12 egonw go back there
07:12 olass ok...
07:13 zarah egonw's link is also http://tinyurl.com/lq4bb8
07:13 olass I'm there
07:13 egonw look up the your clone URL
07:13 egonw hit that copy/paste button
07:13 egonw which will put that url on your clipboard
07:13 olass ok
07:13 egonw and then do
07:13 egonw on the cmd line
07:13 egonw git clone <PASTE>
07:13 egonw git clone <Your Clone URL>
07:13 olass first mkdir bioclipse.core, right?
07:13 egonw no
07:14 egonw git clone will do that
07:14 olass so, where will it put things?
07:14 olass mkdir repos, right?
07:14 olass doing so now
07:15 egonw I have things in:
07:15 egonw $HOME/var/Projects/GitHub
07:15 olass ok
07:15 olass downloading
07:16 mgerlich joined #bioclipse
07:16 olass egonw: you need to write wiki pages for the bioclipse wiki how to checkout and commit to bioclipse with git
07:17 olass so other users can do this
07:17 olass for example Lars
07:17 egonw people will not commit your your copy directly
07:17 egonw instead, you will pull there changes
07:17 egonw are apply there patches
07:18 egonw this allows you to check
07:18 olass well, there must be docs on the bioclipse wiki
07:18 egonw things
07:18 egonw yes
07:18 olass maybe you can copy somr from cdk wiki?
07:18 egonw I'll point there
07:18 olass well, the page must exist so you can search bioclipse wiki
07:19 olass ok
07:19 olass done cloning
07:19 egonw yes, yes
07:19 egonw working on it...
07:19 olass that was fast!
07:19 egonw good, notice the speed!
07:19 egonw yes :)
07:19 olass almost as fast as subclipse
07:19 olass ...NOT
07:19 olass ok, what now?
07:20 egonw import the content into an Eclipse environment
07:20 olass a new workspace, I assume
07:20 egonw yes, that's what I would do too
07:21 egonw ok...
07:21 egonw I just deleted my bioclipse.core
07:21 egonw and then forked from yours
07:21 egonw to make the hierarchy right
07:21 olass great
07:22 olass hmm
07:22 olass I guess I should NOT copy projects into workspace but merely link?
07:22 egonw rigth
07:22 egonw veyr important indeed
07:22 olass egonw?
07:22 egonw yes?
07:22 egonw oh no...
07:23 olass ok
07:23 egonw am I forked into a invisible dimension?
07:23 egonw ah, no, good
07:23 olass ok, I got plugins and features
07:23 mgerlich joined #bioclipse
07:23 egonw excellent
07:23 egonw I should try this too
07:23 olass will try to start
07:23 egonw I hope it is the right set
07:24 egonw if not, I have today to fix it
07:24 olass will try to run
07:24 olass setting target platform
07:25 olass hmm, some crap here that should be moved away from core
07:25 olass services.embrace for example
07:26 olass but not so important
07:26 olass hmm, I still have errors
07:26 olass trying a clean
07:27 egonw olass: please send me a list...
07:27 egonw I can still remove those...
07:27 egonw but...
07:27 olass no worries
07:27 olass we can do that later
07:27 egonw the repository size is OK, so you could 'git rm' yourself too...
07:27 egonw though it would not be erased from the repo history then
07:27 olass I still have errors
07:27 olass checking into this
07:27 egonw no, we cannot do that later
07:27 egonw easily
07:27 olass mom, want it to run first
07:28 olass might be other things
07:28 olass ok, this is not good
07:28 olass net.bioclipse.ioformats require org.openscience.cdk.annotation and cdk.ioformats
07:28 egonw oh, crap
07:29 egonw did not think of that...
07:29 egonw damn
07:29 olass what!? cdk.ui requires  org.openscience.cdk.ioform​ats;bundle-version="1.2.2" ??
07:29 olass when did that sneak in?
07:29 egonw cdk.ui?
07:29 egonw you have that?
07:29 egonw I though I removed that from bioclipse.core...
07:30 egonw net.bioclipse.ioformats and org.openscience.cdk.[annotation|ioformats] is fine
07:30 olass no it is not
07:30 olass they are cdk classes
07:30 olass core should not depend on cdk
07:30 olass right?
07:31 olass that was not how I envisioned it
07:31 olass net.bioclipse.encryption is missing too
07:32 olass and net.bc.webservices require  net.bioclipse.cdk.business;bundle-version="2.0.0"
07:32 olass indeed bad
07:32 olass egonw: we have some work to do to clear out these errors
07:32 egonw yes
07:32 egonw indeed
07:33 egonw no, those two cdk plugins are OK
07:33 egonw there is no real CDK stuff in there...
07:33 egonw annotation is just two classes
07:33 egonw and ioformats is just about file formats
07:33 olass then put the classes in bioclipse
07:33 egonw and is actually indepdenent from any CDK interface
07:33 egonw well, they *are* from the CDK project
07:33 olass you should recator them out
07:33 olass refactor
07:34 olass else things are very confusing
07:34 egonw and make a new project for them?
07:34 egonw possible... I'll put it on my TODO.defered
07:34 olass well, the current situation is not good
07:34 egonw ok, so what are the important things:
07:34 egonw encryption missing
07:34 egonw net.bc.webservices
07:35 egonw is that in the core_feature ?
07:35 olass dunno
07:35 olass looking
07:35 olass no
07:35 olass cheminfo
07:35 olass hmm
07:35 olass how do we handle the case when a git repo requires other git repos?
07:36 egonw on the eclipse workspace level
07:36 olass ok
07:36 egonw bioclipse.core must be self consistent
07:36 egonw bioclipse.cheminformatics would require bioclipse.core
07:36 egonw etc, etc
07:36 olass yes
07:36 olass so remove webservices for now then
07:36 olass to cheminfo
07:36 egonw ok, mom
07:36 egonw need to recover encryption first
07:37 egonw OK, doable...
07:37 egonw anything else missing?
07:37 egonw or just encryption ?
07:39 olass well, org.openscience.cdk.[annotation|ioformats] maybe?
07:39 egonw that's not missing, is it...
07:39 olass what you mean?
07:39 olass it's not there
07:40 egonw oh, right
07:40 egonw crap...
07:40 olass but plugins are depending on it
07:40 egonw sorry
07:40 egonw yes, will need to add that too...
07:40 olass yes
07:40 olass and remove webservices and webservices.test
07:40 egonw ack
07:40 olass and remove net.bioclipse.tests
07:41 olass that requires ALL test plugins
07:41 olass and ALL other plugins
07:41 egonw ok
07:41 egonw good
07:41 egonw OK, I removed the cdk stuff rather early on...
07:41 olass also remove services and services.embrace
07:41 egonw so that is gonna take a while...
07:42 olass ok
07:42 olass do I need to make a fresh clone later?
07:42 olass ok, let me know when I can do something
07:42 * olass is getting coffee
07:43 egonw yes, you will
07:52 egonw oh... wait...
07:52 egonw cdk.annotation and cdk.ioformats
07:52 egonw they actually do not have history...
07:53 egonw so, I should just copy them back in...
07:53 egonw OK, problem 1 solved
07:53 egonw that leaves encryption and the stuff to remove
07:53 egonw that's easier....
07:56 egonw while git is working... http://wiki.bioclipse.net/ind​ex.php?title=Git_Development
07:56 zarah egonw's link is also http://tinyurl.com/mqpa8m
08:00 mgerlich joined #bioclipse
08:02 egonw OK, here goes...
08:02 egonw the building of a new bioclipse.core
08:05 edrin joined #bioclipse
08:05 edrin identify garte
08:05 egonw oops'
08:05 egonw :)
08:06 edrin i hate this
08:06 edrin oh
08:07 edrin already changed
08:07 edrin egonw: and what do you think?
08:07 edrin egonw: will they eatit this time?
08:08 * egonw thinks "why the hell is Google ignoring me... ?!? just because I am not from Cambridge ?!?!" http://depth-first.com/articles/2009/07/​10/linked-chemical-data-google-techtalks
08:11 egonw brilliant photo... the cross marks the hero ?
08:12 olass they asked for a photo... had nothing else...
08:12 olass we need a photo bank
08:12 olass and a group picture
08:18 edrin egonw: what photo are you talking about?
08:18 egonw http://www.uu.se/news/news_​item.php?typ=pm&amp;id=701
08:18 zarah egonw's link is also http://tinyurl.com/mg7nsh
08:19 edrin hehe
08:20 shk3 joined #bioclipse
08:21 edrin http://www.uu.se/news/news_​item.php?id=690&amp;typ=pm
08:21 zarah edrin's link is also http://tinyurl.com/mtp9ee
08:24 egonw olass: is actionHistory needed?
08:24 olass mom
08:24 olass don't think so
08:24 olass optional
08:24 egonw oh, phew...
08:24 olass yes, optional
08:24 olass egonw: all should be listed in core_feature
08:25 egonw yes, sorry... lazy
08:25 olass ah
08:25 egonw ok, back from scratch
08:25 olass mom
08:25 olass net.bioclipse.xom?
08:25 olass hmm
08:26 olass no, not in cor efeature
08:26 olass but required in core.tests
08:26 olass egonw: what do you think?
08:27 olass XMO is used in core.tests at getPlugins()
08:27 olass I sort of guess you wrote that
08:27 olass hence we will need net.bc.xom in bioclipse.core git repo
08:27 egonw ok
08:28 shk3 found the first bug in bc2.0.0 :-(
08:29 olass shk3: incorrect, there are MANY bugs
08:29 olass many are reported
08:29 egonw shk3: make sure to file it
08:29 egonw shk3: there should be a '2.0.0' version
08:30 olass egonw: did you remove all RC options?
08:30 shk3 just filing it now
08:30 egonw olass: no, should I?
08:30 shk3 yeah, first bug _I_ found in release
08:30 olass egonw: does it make sense to keep them?
08:30 egonw I think that will mess up the bugzilla...
08:30 egonw with bugs for non-exesting versions
08:30 olass ok...
08:32 egonw ok, another 'git filter-branch' interation...
08:33 egonw olass: meanwhile...
08:33 egonw please delete your bioclipse.core repos
08:33 egonw I'll delete mine too...
08:33 egonw people should not be cloning that one...
08:33 olass ok
08:34 egonw check the Admin tab (light grey, top of page)
08:34 olass hmm... how?
08:34 olass ok
08:34 egonw then down the bottom of that page
08:35 olass done
08:35 egonw https://sourceforge.net/apps/mediawiki/cdk/index.​php?title=Thoughts_on_requirements_for_the_applet
08:35 egonw ChEBI team write up on how they use a chemical editor
08:36 egonw impressive analysis
08:37 olass indeed
08:44 olass egonw: how are things going?
08:45 egonw 95%
08:45 egonw then checking
08:45 egonw so, closing in
08:45 olass ok
08:45 egonw 30 minutes, I think
08:45 egonw all for keeping history :)
08:47 olass history is important
08:47 olass it makes us what we are
08:49 egonw yes, we we don't keep a copy of our past
08:49 egonw git does
08:49 egonw git tracks commits
08:49 egonw we at most make a picture now and then
08:49 egonw with a cross above our heads
08:50 egonw olass: please wipe your new eclipse workspace
08:50 olass done
08:50 olass in tel now
08:50 egonw ack
08:54 edrin1 joined #bioclipse
08:54 edrin1 x
08:54 edrin1 what the hell!?
08:55 egonw ha, your new pwd :)
08:55 edrin1 no
08:55 edrin1 but i was typing on my other client
08:55 edrin1 and now i realize that teh messages did not go in the channel!
08:57 edrin1 i killed him!
08:57 edrin1 the ghost
08:57 edrin1 egonw: http://ajax.sys-con.com/node/980762
08:57 zarah edrin1's link is also http://tinyurl.com/lrk7jm
08:58 edrin1 http://wave.google.com/
08:58 edrin1 http://www.waveprotocol.org/draft-protocol-spec "Google Wave Federation Protocol Over XMPP"
08:58 zarah edrin1's link is also http://tinyurl.com/mq2fvk
08:59 egonw yes, yes...
08:59 egonw was already thinking how we could hook in the XMPP services there...
08:59 egonw but first...
08:59 egonw now
08:59 egonw XMPP IO-DATA update for pointing to the root element
08:59 egonw I think this is important
08:59 egonw next thing, I suggest:
08:59 egonw working things out to be compatible with RDF
08:59 edrin1 sure,
08:59 egonw RDF does not have a XML schemata...
09:00 egonw so we need to do some thinking there...
09:00 edrin1 yes
09:00 egonw olass: I'm cloning the new bioclipse.core git repos at my account in github
09:00 edrin1 zarah: @msg
09:00 zarah edrin1: eh... wha'?
09:00 egonw 93MB
09:01 egonw so will take a while for me to download at ~160KB/s
09:01 edrin1 me notices http://www.waveprotocol.org/patent-license
09:01 zarah edrin1's link is also http://tinyurl.com/mnzzr8
09:04 rojasm joined #bioclipse
09:05 egonw olass: it's at 28% now
09:13 * egonw is installing the 0.5.0 egit eclipse-git plugin of two weeks ago
09:13 egonw now in the official org.eclipse.egit namespace, now being a Eclipse Incubation project
09:16 egonw building workspace...
09:18 egonw olass: OK, I got no compile errors after removal of an unused Require-Bundle
09:18 egonw but notice another unused plugin
09:20 egonw olass: ping
09:20 egonw let me know when you are back
09:20 egonw please fork, like before
09:20 egonw and then I'll delete mine, and fork from you
09:23 egonw I also added a 2.0.0 tag now
09:24 * egonw is making a .zip file from the git repository
09:24 olass ok, back
09:24 egonw and suggest that (after checking it looks good) upload that to SourceForge as source.zip
09:24 egonw actually... why should we...
09:25 olass egonw: ok, how do I do it again?
09:25 egonw people can simply hit the 'Download' botton themselves...
09:25 egonw go here:
09:25 egonw http://github.com/egonw/bioclipse.core/tree/master
09:25 zarah egonw's link is also http://tinyurl.com/kn37tf
09:25 olass what is the URL I should fork?
09:25 olass ok
09:25 egonw git the 'fork' button
09:25 olass don
09:25 olass http://github.com/olas/bioclipse.core/tree/master
09:25 zarah olass's link is also http://tinyurl.com/lq4bb8
09:26 egonw excellent
09:26 olass and now: git clone?
09:26 olass from CL?
09:26 olass or anything else?
09:26 egonw yes
09:26 egonw and then import into eclipse workspace
09:27 olass downloading
09:27 egonw looks good
09:28 egonw we now have a very easy way to make source distribs :)
09:31 egonw OK, forked you
09:31 olass would be nice to do the git checkout with buckminster
09:31 egonw yes, let's talk about that when Arvid is back
09:31 olass yes
09:32 egonw cool
09:32 egonw that's done
09:32 egonw olass: welcome to the git world!
09:32 egonw btw...
09:32 egonw installed the egit plugin for eclipse...
09:33 egonw and that seems to work fine so far
09:33 olass strange
09:33 olass Can't set target platform
09:33 olass "widget is disposed"
09:33 olass wil lreboot eclipse
09:34 olass ah, that worked
09:34 olass no compile errors now
09:34 olass will try to run
09:35 olass yes, runs
09:35 olass egonw++
09:36 olass ok, egonw, what now?
09:36 olass I need cheminfo repo too
09:36 olass and then I'd like to know how to make commits
09:36 olass etc
09:37 * olass is installing egit
09:39 olass egonw: 0.5.0?
09:39 egonw yes
09:39 olass other people using gtx on mac?
09:40 egonw gitx
09:41 egonw yes
09:42 * egonw will now look at bioclipse.cheminformatics
09:42 egonw while that is in progress...
09:42 egonw I suggest I will teach Ola how to make commits...
09:43 egonw http://github.com/egonw/bioc​lipse.core/graphs/punch_card
09:43 zarah egonw's link is also http://tinyurl.com/nebx2v
09:43 egonw yes, Bioclipse is clearly product by payed employees
10:01 egonw olass: http://github.com/egonw/bioclip​se.cheminformatics/tree/master
10:01 egonw that one works
10:01 egonw too
10:02 zarah egonw's link is also http://tinyurl.com/ms7fct
10:14 olass egonw: do I need to fork the cheminfo?
10:14 egonw no need for that
10:14 egonw well...
10:14 olass so use the clone url?
10:14 egonw depends
10:14 egonw I guess you will want to make patches
10:14 egonw then you fork
10:14 olass ok
10:15 olass I will VERY likely make patches
10:15 egonw I can see what patches you made and commited to GitHub and apply them to the main repos on my account
10:15 egonw :)
10:15 olass I guess that is the way things work
10:15 olass ok, forking
10:16 olass and then git clone
10:16 olass hmm, what does the git clone really do?
10:16 olass in what way is that different from a fork?
10:16 egonw it's the same
10:17 olass so why do I need a for kon github?
10:18 olass so I can commit there, and you can pull from that repo?
10:18 egonw so that you can share your changes with the rest of the world
10:18 egonw right :)
10:18 olass ok, so no use for pele for this
10:18 olass I thought you used it at some points
10:18 egonw no, pele not needed
10:18 egonw ah, yes
10:18 egonw but that was because the whole of Bioclipse git repos was about 300MB
10:19 egonw which is about the free quota one has on GitHub
10:19 olass hmm, I am now using 114 Mb
10:19 olass 300 is not that much
10:20 olass will it suffice?
10:20 egonw I hope so
10:20 olass oh, you can pay
10:20 olass hmm, do you pay?
10:21 egonw no
10:21 olass you have a lot of repos
10:21 olass hmm, but no private repos I guess
10:21 egonw no
10:25 olass egonw: the feature patch and experimental feature and cdkws09-feature should be removed
10:25 olass from bioclipse.core
10:25 egonw yes, please do
10:26 egonw was not sure about that
10:26 egonw this is how you do it:
10:26 olass just wipe them?
10:26 egonw yes
10:26 egonw or 'git rm -r'
10:26 olass they should go in other git repos
10:26 egonw both should be the same
10:26 egonw right
10:27 olass net.bioclipse.databases is not in core either
10:29 olass egonw: how should we treat the data features?
10:29 olass I'd like to be responsible for them
10:29 olass (for now)
10:30 olass and also balloon
10:30 olass will you set this up?
10:30 egonw it's a lot of work
10:30 egonw I can do it...
10:30 egonw but you would need to be patient
10:30 olass or tell me how to do it
10:30 egonw see my email
10:30 olass sorry, I need VERY verbose explanations
10:31 olass type this, then this etc
10:31 olass but I can be patient
10:31 egonw well, *all* the command are there
10:31 egonw ok, except the git-svn command
10:32 egonw OK, will go for lunch first now
10:32 egonw bbl
10:37 olass egonw: Commented in page http://wiki.bioclipse.net/ind​ex.php?title=Git_Development
10:37 zarah olass's link is also http://tinyurl.com/mqpa8m
10:37 olass asking for clarification at some points
11:18 egonw olass: I'm updating the wiki
11:18 egonw and back from lunch
11:20 rojasm joined #bioclipse
11:27 olass great
11:28 shk3 joined #bioclipse
11:35 egonw olass: please email me what plugin and feature folders you want together...
11:35 egonw then I can help with conversion to a git repository
11:36 egonw shk3: same for you... if you want to move speclipse/nmrshiftdb into a git repository too
11:36 egonw rojasm: and you too
11:36 olass ok, I will email
11:36 shk3 currently discussing with christoph which way to go
11:36 egonw shk3: ack
11:37 egonw shk3: I think it would be a valid option to keep using the bioclipse svn
11:37 egonw so there is no stressing need to convert to git right now
11:37 * egonw will now remove, however, all plugins that did move to one of the two git repositories
11:38 egonw which will be interesting in itself...
11:39 olass egonw: I suggest keeping plugins in SVN for now
11:39 olass just to be sure
11:39 egonw I rather not, because people have two places where the can commit then...
11:39 olass let us turn off SVN write access then
11:39 egonw for sureness reasons it's not needed, because there is always the svn history
11:39 egonw no, can't do that...
11:40 egonw for all the plugins *not* in a git repos yet
11:45 olass egonw: how do we do for features that share a common plugin?
11:45 olass can (should) we have plugins in multiple git repos?
11:46 egonw yes, but that is basically a bad form of forking
11:46 olass thinking of the qsar repo
11:46 egonw I think we need to decide on a case by case basis
11:46 olass it has chart, statistics, and statistics.jama in it
11:46 olass we could put those in the core repo
11:47 egonw I'd say one repository...
11:47 egonw does not have to be core
11:47 olass even though it does not go in the core platform
11:47 olass feature
11:47 egonw because you can still have any git repos checked out
11:47 egonw without the need to have all of the plugins/features in your workspace
11:49 egonw olass: so, ok if I remove the bioclipse.core and bioclipse.cheminformatics plugins and features from SVN now?
11:52 olass I guess...
11:57 olass hmm, where should the update sites go? I suggest we add them later, without history (it is not important there)
11:57 egonw agreed on the history
11:58 olass egonw: sent you email about repos
11:58 olass egonw++ //for the hard work on this
12:00 egonw ok
12:27 sneumann joined #bioclipse
12:36 masak joined #bioclipse
12:45 CIA-54 bioclipse: egonw * r11666 /bioclipse2/trunk/ (27 files in 27 dirs): These plugins are now available from the git repository at: http://github.com/egonw/bioclip​se.cheminformatics/tree/master
12:45 zarah CIA-54's link is also http://tinyurl.com/ms7fct
12:57 maclean joined #bioclipse
13:01 CIA-54 bioclipse: egonw * r11667 /bioclipse2/trunk/ (27 files in 27 dirs): These plugins are now available from the git repository at: http://github.com/olas/bioclipse.core/tree/master
13:01 zarah CIA-54's link is also http://tinyurl.com/lq4bb8
13:11 * egonw will now first take a break...
13:11 egonw bbl
13:15 masak egonw++ # bioclipse.cheminformatics and bioclipse.core
13:20 maclean This question may have been asked already, but will each developer have to have a clone (fork) of each repository
13:21 maclean for something like (n-m)! repositories for n developers each managing one of m plugins/features.
13:26 masak maclean: two answers:
13:26 maclean ok.
13:27 masak maclean: (1) each developer having a clone is how git, and distributed revision control systems in general, work(s)
13:27 maclean this, I know.
13:27 masak maclean: (2) it's hard to believe, but it actually makes merging simpler, not harder.
13:27 maclean hmmmmm.
13:28 masak it's all due to the way git braids things into a coherent history after the fact.
13:29 maclean yes, it's all very clever. But say I am a (potential) developer new to bioclipse, and I want to make a new plugin, but need to get the code for the rest of bioclipse.
13:29 maclean where do I go?
13:30 maclean Am I meant to get core from ola, cheminformatics from egon, structuredb from jonalv...
13:30 masak ah. good question.
13:30 masak I had half-assumed that there would be a 'bioclipse' user on github.
13:30 maclean I saw Buckminster mentioned on the chat earlier today.
13:31 maclean Ahh. A sort of 'centralised' repository, eh? :)
13:31 egonw back
13:31 egonw yes, you have to do some shopping
13:31 egonw but, right now that's more or less the same...
13:31 egonw you put this plugin in the workspace and that one... but not that one
13:32 maclean This is very true.
13:32 egonw now you just put all in from a particular set of a few git repositories
13:32 egonw also, you now also make a  fork
13:32 egonw just without history
13:32 egonw now := in the svn world
13:32 egonw and just without any commit rights
13:32 maclean right.
13:33 maclean so, this (hypothetical) new developer clones all these repositories, creates his own plugin (locally) and puts that it its own repository.
13:34 egonw yes
13:34 maclean but without commit rights, he can't commit to (say) core. he has to make a patch, and submit that.
13:34 egonw (or pushes the Download button to pull a .zip or .tar file instead of cloning)
13:34 egonw yes
13:34 maclean okay.
13:35 egonw so that he can choose how he wants to share it with the world
13:35 maclean I'm just thinking of the little guy, here :)
13:35 egonw he can also put a git repos on his private homepage
13:35 egonw but 'git format-patch' works for the midgets
13:36 maclean Should there be a mechanism for keeping track of repositories for new ('third-party') plugins?
13:36 egonw and if even that does not work, there is always "just send me your code, and I'll figure it out"
13:36 egonw was thinking about a google map for that...
13:36 maclean :)
13:36 egonw http://maps.google.com/maps/ms?ie=UTF8&amp;hl=​en&amp;msa=0&amp;ll=56.559482,13.623047&amp;sp​n=17.913934,78.134766&amp;t=h&amp;z=4&amp;msid​=103733862046497992978.00046dce090029ceed81d
13:36 zarah egonw's link is also http://tinyurl.com/lcv8x9
13:36 maclean And perhaps a twitter hashtag...
13:37 egonw like that one, but then with icons for repositories or update sites or so...
13:37 maclean Could do, could do. Could start with a simple html list on a webpage, with links, though...
13:38 egonw would do too
13:38 maclean Nothing wrong with the old-fashioned way, to start.
13:39 egonw what ever is quicker
13:39 maclean Anyway, I'll let you get back to the heavy-lifting of repository management :)
13:39 egonw actually...
13:39 egonw doing some jcpprim porting work now
13:39 egonw would very much like to finish that today
13:39 egonw only controlbasic and controlextra left...
13:40 egonw but this requires some more code clean ups...
13:40 egonw as we just discussed...
13:40 maclean 'good job' as the americans say. egonw++
13:41 maclean I should get on with using my freshly compiled signature translator.
13:41 egonw partially followed that...
13:41 * maclean is also using bioclipse to make unit test case molecules
13:41 egonw but did not quite get what you were doing...
13:42 egonw excellent!
13:42 egonw happy to hear that
13:42 maclean I like the new atom types in the outline, btw. nice idea.
13:43 egonw thanx
13:43 egonw ola and I came up with the same need simultaneously
13:44 egonw it's very useful to see what bits of your drawing Bioclipse/CDK understands
13:44 maclean true.
13:44 maclean and for atom merges, I suppose.
13:44 egonw go on...
13:44 maclean to make it clear that the atoms are not just on top of each other.
13:44 egonw did not think if that ...
13:45 maclean There isn't the same feedback as in swing-jcp for merging
13:45 egonw because you see the type change?
13:45 egonw or while doing the merge?
13:45 maclean yes
13:45 maclean no, after.
13:45 egonw ack
13:45 egonw yes, I think Bioclipse-JChemPaint has much more flexibility in terms of functionality...
13:45 egonw it can't replace the applet
13:46 egonw but it could replace the application...
13:46 maclean true, but in this case, it seems like the swing version has the functionality :)
13:47 maclean unless I'm mis-remembering.
13:47 * egonw is going to figure out who the author(s) is/are of the Renderer.calculateBOunds() methods
13:47 maclean me, probably.
13:47 egonw what functionality?
13:47 maclean feedback on merging
13:48 egonw maclean: thing is... I need to know for sure, and all who contributed in all the methods history...
13:48 egonw oh, btw... did you see Mark's question?
13:48 maclean I did, yes.
13:48 egonw please tell him I do not know the answer to that yet... and Arvid is on holiday now...
13:49 egonw but that I will explore asap
13:49 maclean ok, will do.
13:49 egonw but not before I finish the porting
13:49 egonw which is why I really would like to finish that today
13:49 CIA-54 bioclipse: shk3 * r11668 /bioclipse2/trunk/plugins/net.biocli​pse.nmrshiftdb/META-INF/MANIFEST.MF: removed some ids from plugin dependencies
14:21 masak egonw: so, there'll be a bioclipse.bioinformatics repo? I'm longing to transfer my branch. :)
14:22 egonw yes
14:22 egonw are you gonna keep history?
14:22 masak you mean, will I squash the commits or not?
14:23 masak it doesn't matter much, I guess. but I thought I'd keep the commits as they are.
14:23 egonw can you copy one folder to another repository then?
14:23 masak egonw: no, I'll probably export them as patches and apply them on the new repo.
14:23 egonw for one folder?
14:24 egonw how does that work?
14:24 masak sorry, which folder would that be? I've been all over the place with this branch.
14:24 egonw right
14:24 egonw that's why I ask :)
14:24 egonw there are also commits changing bioclipse.core stuff
14:24 masak aye.
14:24 egonw affecting your plugin folders...
14:24 egonw so was wondering how to get a clean set of commit patches
14:25 masak I see now that that will be difficult. :/
14:25 egonw I have been doing this the 'git filter-patch' route....
14:25 egonw alternatively...
14:25 masak my branch is suddenly two branches in different repositories.
14:25 egonw instead of deleteing folders
14:25 egonw this also allows isolation of folders...
14:25 egonw which you could then make a patch set for...
14:25 egonw but have not tried this approach yet...
14:26 egonw which also involves aggregating back into one repos those individual folders again
14:31 masak I'll see what I'll do.
14:31 egonw let me know if you come up with something
14:31 egonw my approach of deleting works nice for last blobs
14:31 egonw but not for the smaller ones
14:31 masak I haven't migrated my workspace from git-svn to git yet. I think I'll wait until the bioclipse.bioinformatics branch is out.
14:32 egonw ah, but creation of that was what I was just talking about...
14:32 egonw creating that is pretty much a pain
14:32 egonw it took me some 15 hours to do bioclipse.core and bioclipse.cheminformatics
14:32 egonw learning and doing and redoing
14:32 egonw and redoing once more
14:33 masak ah, ok.
14:33 masak so what are you suggesting?
14:33 egonw that you come up with something clever
14:33 egonw maybe perly ? :)
14:33 egonw masak: please email me which features and plugins are involved
14:37 masak in my branch, you mean?
14:37 masak I'll have a look.
14:51 masak egonw: email sent.
14:56 egonw olass: can you reply to Toni ?
14:56 egonw masak: ack...'
14:57 egonw with only 4 plugins, i think we should explore the 'git filter-branch' option to isolate individual plugins and add that...
14:57 egonw I will explore...
14:57 egonw need to learn that anyway
14:59 olass egonw: I'll reply to Toni
14:59 egonw thanx
15:04 edrin1 did you already start your release party?
15:06 edrin1 olass: did you already have your release party?
15:06 olass yes
15:06 olass we had it on wednesday
15:06 edrin1 ah ok
15:07 edrin1 no party pic on the blog?
15:11 masak edrin1: there were photos taken, but I haven't seen any being sent around yet.
15:11 * egonw wonders how to extract part of a commit patch...
15:11 egonw the bits for one particular file...
15:15 olass edrin1: the blog is quite formal
15:15 olass not a place for party pics
15:15 olass I can put one on the wiki
15:15 egonw oh, cool
15:16 egonw git++
15:16 egonw it's really easy!
15:16 egonw git format-patch --start-number 1 4ffc24c919b1e3043f1dfbabdff76d4c9bd617a1 -- src/main/org/openscience/cdk​/geometry/GeometryTools.java
15:16 egonw masak: this is important for you our earlier discussion too!
15:16 egonw it allows just giving the changes in a certain range for a given path !
15:17 egonw so, it should be trivial to make a patch series for a particular folder!
15:19 edrin1 olass: sure :)
15:21 masak egonw: oh, good.
15:21 masak yes, that's all we need.
15:22 masak egonw: I'll try that as soon as there's a bioclipse.bioinformatics repo to try it on.
15:22 egonw :)
15:22 edrin1 masak: ah, the bioinformatics repo
15:22 egonw first I need to make dinner, finish jchempaint git/cdk-master patch and some OWL stuff
15:22 masak egonw: there's no rush.
15:23 masak I'm doing other things in the meantime.
15:23 edrin1 masak: do you have already a feature list for this?
15:23 egonw yes, time is mean
15:23 masak edrin1: yes, a growing one.
15:23 edrin1 masak: is it publicly observable?
15:24 edrin1 zarah: ping
15:24 zarah edrin1: pong
15:25 edrin1 masak: ?
15:25 masak edrin1: no, because I value rebasing.
15:25 edrin1 sorry, i dont understand
15:25 edrin1 what do you mean?
15:26 masak edrin1: git allows you to 'rebase', which makes it easier to merge.
15:26 edrin1 ah
15:26 masak http://www.kernel.org/pub/softwa​re/scm/git/docs/git-rebase.html
15:26 zarah masak's link is also http://tinyurl.com/27437e
15:26 masak but it's only possible (more or less) in a non-public repository.
15:27 edrin1 no, i mean, something like a plan, what kind of stuff/plugins you want to create
15:27 masak oh, ah.
15:27 masak no, I don't have that. I should, though.
15:27 masak openness is good, and I need the input from others.
15:27 masak right now what I have is simple sequences stuff: DNA, RNA, Protein.
15:27 masak conversion between them, reading/writing to files.
15:28 edrin1 masak: I always wanted a free alternative to Vector NTI :-)
15:28 masak what's that?
15:29 edrin1 that's a very good and very expensive program from Invitrogen that is used by many experimental bio researchers to clone their stuff...
15:29 edrin1 http://www.invitrogen.com/site/us/en/home/Pr​oducts-and-Services/Applications/Cloning/Vec​tor-Design-Software/Vector-NTI-Software.html
15:29 zarah edrin1's link is also http://tinyurl.com/5k624b
15:30 edrin1 masak: look at this: http://www.invitrogen.com/etc/medialib/en/image​s/mainbody/punchout.Par.59207.Image.-1.-1.1.gif
15:30 zarah edrin1's link is also http://tinyurl.com/lmm592
15:30 masak looks impressive.
15:30 masak I guess it'll be a while before we get there in Bioclipse.
15:30 masak but let's make it possible, by all means.
15:30 edrin1 yes
15:32 edrin1 masak: if we have some/few features of the vector NTI, like drawing vectors according to its sequence in a window, annotating it graphically, visualizing selected restriction site positions...
15:32 edrin1 in this somehow stable...
15:32 edrin1 Bc users will increase exponentially
15:32 masak sounds like fun features to implement.
15:33 * masak likes doing graphics
15:33 edrin1 my institute is going to buy 5 Vector NTI licenses soon...
15:33 edrin1 we have around 50 (?) researchers here...
15:34 edrin1 but we cannot afford more than 5 licenses...
15:34 edrin1 i think 1 license is ~ 1500 per year...
15:34 masak ouch.
15:36 maclean joined #bioclipse
15:36 edrin1 well, i am sure this is very much to do to create something halfway like this...
15:39 masak it's good to hear about user stories early, anyway.
15:39 edrin1 masak: yes
15:39 edrin1 i hope i can help here a lot
15:40 edrin1 .
15:40 edrin1 lets see, maybe we have some time to discuss this at some occasion
15:44 masak edrin1: yes, I'd like that.
15:44 masak sometime early next week, perhaps?
15:44 masak (I'll be away Thu-Fri next week)
15:44 edrin1 can not say yet
15:46 egonw rection
15:46 egonw vertically
16:42 olass egonw: thank good you missed an e in the beginning at 17:47
16:43 maclean evertically?
16:48 egonw inded
16:50 shk3 joined #bioclipse
16:59 egonw OK, controlbasic works too now!
16:59 shk3 joined #bioclipse
16:59 egonw maclean: is Mark still around?
17:00 egonw ok, never mind
17:00 maclean I have no idea - I'm at home :)
17:00 egonw will email
17:00 egonw ic
17:00 egonw ok
17:00 egonw one module left...
17:00 egonw good thing is...
17:00 egonw controlbasic and renderbasic
17:00 egonw are self consistent
17:00 egonw no deps on renderextra or controlextra
17:00 egonw bbl
18:46 edrin joined #bioclipse
20:14 edrin joined #bioclipse
20:25 egonw hi edrin
20:39 edrin hi egonw
20:40 edrin still online and working?
20:42 egonw online, but not working...
20:42 edrin :)
20:42 edrin same here
20:44 egonw well, there is considerable overlap :)
20:44 egonw was doing some jchempaint work
21:08 olass joined #bioclipse
21:33 egonw @tell masak please read this wiki page, and the "Create patches for individual plugins/features" in particular: http://wiki.bioclipse.net/index.php?title=Git_De​velopment#Create_patches_for_individual_plugins.2Ffeatures
21:33 zarah egonw's link is also http://tinyurl.com/mqpa8m
21:33 zarah Consider it noted.
21:43 edrin egonw: will bc move to git very soon?
21:50 edrin uhm
21:51 edrin obviously it did alrady

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