Camelia, the Perl 6 bug

IRC log for #bioclipse, 2007-10-23

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

All times shown according to UTC.

Time Nick Message
00:05 CIA-18 joined #bioclipse
06:34 viklund joined #bioclipse
07:30 bjarni good morning
07:35 viklund moring
07:35 viklund morning...
07:36 egonw joined #bioclipse
07:36 olas joined #bioclipse
07:37 olas hi
07:37 olas hi bjarni
07:38 bjarni is it possible to resize the fs on the laptop i got online? it appears to be an hfs+ fs...
07:38 bjarni i'd like to install another os, which requires some fs resizing i guess
07:42 olas I dunno
07:42 olas google or ask magnus
07:43 olas are you at BMC now?
07:43 bjarni yes
07:43 olas drop by here and we'll talk to magnus
07:45 bjarni mm, I'll drop by in a bit
07:45 egonw moin
07:45 egonw bjarni: what fs does it have now?
07:45 egonw there are gparted live CDs
07:46 egonw which generally cover the most general FSs
07:46 bjarni egonw: hfs+
07:46 bjarni apparently I can't overwrite the current system, so I'll have to shrink it
07:46 egonw http://gparted-forum.surf4.​info/viewtopic.php?pid=1592
07:46 egonw there are some comments tehre...
07:47 egonw "this has been on the TODO for a long time" :)
07:48 bjarni :-/
07:49 bjarni well, if hfs supports resizing there should be a resizehfs or something...
07:54 jonalv joined #bioclipse
07:57 olas hi jonalv
07:57 jonalv hi
08:41 masak joined #bioclipse
08:41 CIA-18 bioclipse: ospjuth * r4040 /trunk/playground/plugins/net.bioclip​se.biomoby/src/net/bioclipse/biomoby/ (3 files in 2 dirs): Saving results works (but not good)
08:49 viklund xmonad!
08:52 olas ??
08:53 masak viklund: what about it?
08:53 viklund ;) it's a window manager written in haskell
08:53 viklund it's nice
08:53 masak ok...
08:53 viklund I'm running it now, compiz leaked too much memory
08:54 jonalv oh?
08:54 jonalv I was looking forward to compiz...
08:54 jonalv :(
08:54 masak .oO( leaking forward to compiz )
08:55 viklund xmonad only takes 5380k in virtual memory, resident memory is half of that
08:55 viklund compiz went to 500 to 700 Mb in a day...
08:56 jonalv wow
08:56 masak viklund: but how's the user experience in xmonad?
08:56 viklund It's a lot easier than ion2, which was the last tiling windowmanagers I tried.
08:56 viklund it's very simple to use, not many commands at all
09:03 viklund I also find the default tiling setup really good. One big window to the left (stretches from top to bottom) and several tiled windows to the right.
09:04 masak viklund: btw, how's the web services stuff coming along?
09:04 viklund currently, not at all... ;(
09:05 viklund it's a bit downprioritized, got too much on my hands, courses and stuff to take care of
09:06 masak ah.
09:27 bjarni mumble
09:29 bjarni who here has actually installed an operating system on a macbook?
09:30 jonalv :)
09:31 bjarni after reading a bit it appears this with running operating systems is not what they are built for...
09:31 masak bjarni: why? trouble?
09:31 masak ah.
09:31 masak bjarni: as I said, I know this guy who runs BSD on his
09:32 masak so it's certainly possible
09:32 bjarni I don't know, I might need to use the os X installation to edit the partition table
09:32 bjarni of course it's possible
09:32 masak well, in the worst case, you'll just have to contend with Mac OS X :)
09:32 bjarni one problem is that I don't have an account on the os X installation
09:32 masak with MacPorts, it's not so bad
09:33 masak bjarni: are you able to get into single user mode?
09:33 bjarni unknown
09:34 masak it's something you do at startup to get there -- kinda lika holding down most of the keys and whistling at the same time
09:34 bjarni the bootloader never says anything, so I don't know what to press...
09:34 masak bjarni: want an URL with key combos?
09:34 bjarni ok
09:35 masak http://davespicks.com/writing/​programming/mackeys.html#boot
09:35 * masak misses lambdabot
09:35 bjarni is he dead now?
09:36 * jonalv too
09:36 masak bjarni: no, just gone, and I'm not sure it's a 'he'
09:37 bjarni yay \o/
09:37 masak it has an extension called 'vixen' which does an ELIZA-like thing, only as a h0t ch1xx0r
09:37 bjarni localhost:/ root#
09:37 bjarni :D
09:37 masak bjarni: better? :)
09:37 bjarni y�s :D
09:43 bjarni fsck fsck...
09:45 lambdabot joined #bioclipse
09:46 jonalv @botsnack
09:46 lambdabot :)
09:46 masak @tell masak wohoo!
09:46 lambdabot You can tell yourself!
09:46 masak :)
09:46 bjarni diskutil just hangs...
09:47 masak bjarni: ready to be here in 15?
09:47 bjarni ok, I'll continue with this after lunch :-/
09:47 * bjarni -> bmc
09:47 jonalv @viksen hi
09:47 lambdabot hello there
09:48 masak @vixen what is your name?
09:48 lambdabot My name is Emily
09:49 masak @vixen hello Emily
09:49 lambdabot hi
09:49 jonalv @vixen hello sara
09:49 lambdabot hello there
09:49 masak @vixen so, what do you think of Bioclipse?
09:49 lambdabot i do occassionally, i guess
09:49 masak eh.
09:49 masak not quite Turing-ready yet
09:49 jonalv no
09:50 EskilA joined #bioclipse
09:50 olas @vixen do you like jonalv or masak?
09:50 lambdabot sure, i definitely don't hate
09:50 masak haha  :)
09:50 masak @vixen remember that jonalv calls you sara
09:50 lambdabot liverwurst gets me really gassy...
09:50 olas what !?
09:50 masak :P
09:50 jonalv strange...
09:51 masak @vixen please never mention liverwurst again
09:51 lambdabot you have very nice manors
09:51 jonalv eh
09:51 masak "manors"?
09:52 masak @join #farmbio
09:52 lambdabot Not enough privileges
09:52 masak they have to be cleared first
09:53 masak dons++ has cleared #bioclipse for us
09:53 jonalv aha
09:55 viklund > let fibs=1:1:zipWith (+) fibs (tail fibs) in take 10 fibs
09:55 lambdabot [1,1,2,3,5,8,13,21,34,55]
09:55 viklund Yay!
09:56 masak viklund: you forgot the initial 0
09:56 viklund nope, I like the fibonacci series better when it's starting with two ones...
09:56 masak viklund: I don't
09:57 viklund ok
09:58 viklund let fibs=0:1:zipWith (+) fibs (tail fibs) in take 10 (drop 1 fibs)
09:58 viklund oops
09:58 viklund > let fibs=0:1:zipWith (+) fibs (tail fibs) in take 10 (drop 1 fibs)
09:58 lambdabot [1,1,2,3,5,8,13,21,34,55]
09:58 viklund better?
09:58 viklund masak?
09:58 masak viklund: yes and no.
09:58 masak you still ignore the 0
09:58 viklund ;)
09:58 masak but nvm
09:59 viklund but it's in the definition now!
10:01 masak lunch &
10:53 olas left #bioclipse
11:57 CIA-18 bioclipse: jonalv * r4041 /trunk/playground/plugins/ (2 files in 2 dirs): continued work with adopting the usermanager into accepting many accounts of the same account type
12:11 ChanServ joined #bioclipse
12:39 egonw joined #bioclipse
13:52 masak viklund: let's say I have a hypothetical Bioclipse 2 with a scripting language built in from the start
13:52 viklund masak: mm
13:52 masak we're thinking about possible syntaxes here for things
13:53 masak maybe you could act as a sanity filter
13:53 viklund ok
13:53 masak ISequence seq = new DNASequence("ACGTCATGCATGCATGCATGCTAGCTACGTA");
13:53 masak sane?
13:53 viklund (i've fallen in love with xmonad=
13:53 viklund )
13:54 masak viklund: goodie
13:54 viklund looks ok, the I might be a bit strange though, but it works
13:55 masak viklund: we're growing ever more aware of
13:55 masak the tension between a pure/sane model and ease of use for the new Bioclipse(2) user
13:55 viklund ;)
13:55 masak it's not an easy problem
13:56 viklund no, it isn't...
13:56 masak the way around ISequence wpould be using a scripting language that had "mu" instead of "ISequence"
13:56 masak it'd still _be_ an ISequence, but we wouldn't bother the user with it
13:58 viklund "mu"?
13:58 viklund muSequence?
13:58 masak no, more like `my seq = new DNASequence("...");`
13:58 viklund ahh
13:58 jonalv joined #bioclipse
13:59 viklund a static, strong type system then...
13:59 egonw why define a new scripting language?
13:59 egonw why not just expose an existing API?
13:59 masak egonw: that's also one of the issues at hand
13:59 masak the jury is still out on that one
14:00 jonalv The arguing was that we wanted to do: sequence.save() (to file)
14:00 masak the chief guiding light is Easy Of Use, and the question is whether existing scripting languages can give us that
14:00 jonalv but I argue that saving to file is a controller thing and thus should not ve confused with the model
14:00 egonw yes, I can imagine that Bioclipse2 will have a definate API
14:00 egonw which allows easy scripting
14:00 viklund ISequence seq = new ...
14:00 masak egonw: definitely
14:01 viklund i think that works...
14:01 masak that's what I want to make sure, at least
14:01 masak viklund: goodie
14:01 viklund I guess it's not too diffucult to go from "I... seq = new" to "my seq = new"
14:01 viklund or rather not more difficult than doing it right away...
14:01 masak (actually, I don't think one should take the Ease Of Use idea too far. people will still copy working code from somewhere and adapt it)
14:02 egonw moreover... no one wants to learn a new language
14:02 masak egonw: especially one that is only usable in Bioclipse
14:03 egonw but just use Java/JavaScript/Pyton/Ruby instead
14:03 masak hence the many advantages of using js or Ruby
14:03 masak egonw: yes
14:03 masak personally, I find that a more important requirement than avoiding certain syntactical constructs
14:07 masak viklund: have you used SMILES?
14:09 masak IMolecule mol = cdk.parseSmiles("CCCCO");
14:09 masak sane? :)
14:10 jonalv maybe just Molecule mol?
14:10 jonalv There ain't a law for naming interfaces starting with I you know...
14:10 masak jonalv: so you're willing to sacrifice a programming convention for Ease Of Use?
14:11 masak fair enough
14:11 masak as long as we don't think we're going to use Molecule for anything else, might as well call IMolecule Molecule instead
14:11 viklund masak: no
14:12 viklund not the sane part, the SMILES part
14:12 masak viklund: what about the sane part, hen?
14:12 masak s/hen/then/
14:12 masak SMILES is just a molecule specification format
14:12 viklund it's a bit assymtetric I think, compared to the sequence
14:13 viklund maybe something like:
14:14 viklund IMolecule mol = new fromSmiles("CCCCO")
14:14 viklund or such...
14:14 masak and use a class called "fromSmiles"?
14:14 masak yeech...
14:14 jonalv go the other way
14:14 viklund yes well
14:14 * egonw suggests that viklund to have a look at Rich' Rubidium
14:14 jonalv don't use new at all
14:14 viklund that might not be that good, I know
14:14 masak egonw: what's that?
14:15 egonw google for 'rubidium rich apodaca'
14:15 masak @google rubidium rich apodaca
14:15 lambdabot http://depth-first.com/
14:15 lambdabot Title: Depth-First
14:15 masak oh, that one
14:15 egonw oh, that's cool
14:15 masak :)
14:15 viklund it's just that I thought having "new DNASequence(...)" and then "cdk.parseSmiles(...)" isn't really that logical from a users point of view.
14:15 egonw @google open source chemoinformatics
14:16 lambdabot http://almost.cubic.uni-koeln.de/cosi/
14:16 lambdabot Title: COSI - Chemoinformatics Open Source Initiative
14:17 viklund using new in one place and a factory method parse in another. I mean, your creating an object from a string in both cases
14:17 masak egonw: I've been following the depth-first blog for a few weeks now
14:17 masak it looks very promising
14:17 masak I think getting Juby into Bioclipse would be a great boon
14:19 viklund http://eclipsefp.sourceforge.net/ ftw!
14:19 lambdabot Title: EclipseFP - Functional programming support for Eclipse
14:20 masak viklund: I still think Ruby is a preferable alternative
14:20 viklund why so practical? Take a walk on the wild side!
14:22 egonw viklund: right... as if biologists are so wild...
14:22 egonw oh wait... what about the chemists?? they are wild :)
14:22 viklund :p
14:23 masak and they seem to be already using Ruby
14:23 masak viklund: I say one step at a time :)
14:23 viklund yes
14:23 jonalv what is it about ruby that I hear?
14:23 masak of _course_ FP will take over the world
14:23 masak eventually.
14:23 masak if not for other reasons than because we will need to build real AI somehow
14:24 viklund this syntax is perl but whatever:
14:24 viklund perlish...
14:24 masak all input is appreciated.
14:25 viklund IMolecule mol = new Molecule(smiles => '...')
14:25 jonalv viklund: Can you give me a sequence format name except for fasta?
14:26 viklund genbank
14:26 masak EMBL
14:27 masak viklund: that syntax works in all languages which have syntactic support for hashes
14:27 viklund or similar...
14:27 masak right
14:27 viklund so would work in ruby (juby?)
14:27 masak would work in JRuby, yes
14:27 masak ...I think
14:29 masak viklund: seemingly, the data model around Molecules can quickly get very complex
14:29 masak egonw: I was told to seek your views on that, too
14:29 masak check out http://wiki.bioclipse.net/index.php?title=​A_Meta_Language_for_Bioclipse#Create_1000_​virtual_molecules_around_a_scaffold_from_a​_fragment_library_with_random_selection
14:30 masak for starters
14:30 lambdabot Title: A Meta Language for Bioclipse - BioclipseWiki, http://tinyurl.com/29o6bt
14:30 masak @thx
14:30 lambdabot you are welcome
14:30 masak viklund: actually, olas++ has used hash-like syntax in his examples on the wiki
14:30 masak only with = instead of =>
14:36 jonalv What more than biojava works with sequences?
14:40 viklund hmm, yes...
14:40 viklund jonalv: huh?
14:40 egonw writing a meta language is doable... Jmol did it too
14:40 egonw but that has a relatively small command set
14:40 jonalv did?
14:41 viklund I'll come back in a sec, X-server restart
14:41 egonw ok, yes, development of extensions is still ongoing...
14:41 jonalv I'm scetching a classdiagram for this bc2 structure so I need some examples and then I will nedd LOTS of feedback
14:41 masak egonw: it's doable. all I'm saying is that it's a tradeoff
14:43 masak betweem (perceived) Ease Of Use and taking something that's already established (Ruby, js, Java)
14:46 bjarni ugh, I'm too tired for this... bioclipse wiki says to install subclipse in eclipse but that fails...
14:47 bjarni howtos for installing linux on a macbook say to use bootcamp to resize partitions, but bootcamp has er expired...
14:47 masak bjarni: what fails with subclipse?
14:51 bjarni \o/
14:55 bjarni fail mode deactivated :O
15:27 viklund joined #bioclipse
15:28 viklund back...
16:23 test joined #bioclipse
16:23 Topic for #bioclipseis now An open source workbench for chemo- and bioinformatics - http://www.bioclipse.net/ | BiclipseWiki => http://wiki.bioclipse.net/ !! | logs => http://colabti.de/irclogge​r/irclogger_logs/bioclipse & http://moritz.faui2k3.org/irc​log/out.pl?channel=bioclipse | Dieter wrote an article for Eclipse Magazine: http://eclipse-magazin.de/itr/ausg​aben/psecom,id,359,nodeid,228.html
16:23 Topic for #bioclipseis now An open source workbench for chemo- and bioinformatics - http://www.bioclipse.net/ | BiclipseWiki => http://wiki.bioclipse.net/ !! | logs => http://colabti.de/irclogge​r/irclogger_logs/bioclipse & http://moritz.faui2k3.org/irc​log/out.pl?channel=bioclipse | Dieter wrote an article for Eclipse Magazine: http://eclipse-magazin.de/itr/ausg​aben/psecom,id,359,nodeid,228.html
16:37 CIA-18 bioclipse: jonalv * r4042 /trunk/playground/plugins/ (22 files in 6 dirs): fixed last test and did some refactoring from the name Keyring into Usermanager and from SuperUser to User
16:49 viklund joined #bioclipse
19:40 CIA-18 joined #bioclipse
19:54 viklund joined #bioclipse
19:59 viklund /leave
19:59 viklund left #bioclipse

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