Camelia, the Perl 6 bug

IRC log for #bioperl, 2009-07-16

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

All times shown according to UTC.

Time Nick Message
00:31 kyanardag_ joined #bioperl
01:08 rbuels ho ho ho
01:09 rbuels i'm a bad man
01:09 rbuels let's see what hackles my latest email will raise
01:10 deafferret mmmmmmm hackles
01:10 deafferret DAMMIT, MYSQL, BUILD THIS F'N INDEX OR DON'T!!
01:11 deafferret YOU THINK I'M GETTING PAID BY THE HOUR TO SIT HERE?!?!?!  oh... wait...
01:12 rbuels lol
01:12 rbuels deafferret: i volunteered you to help me break bioperl up into smaller distributions
01:12 rbuels that'll give you something to do while building indexes
01:13 rbuels 'CREATE INDEX' => 'svn cp ....; vi Build.PL'
01:14 deafferret rbuels: k. one distro per package?
01:15 * deafferret jumps to it
01:15 deafferret how about we put all the packages that start with A in BioPerl-A.tgz ?
01:15 deafferret or group them by checksum
01:15 deafferret BioPerl-efaca5a6c4a4s6d.tgz
01:15 rbuels hahahha
01:16 deafferret gawd damn... seriously. I should commute. this is silly
01:16 rbuels screen.
01:16 deafferret how long can I stare at this before I go even crazier?
01:16 * deafferret builds a time machine to launch screen
01:16 rbuels hmmm.
01:16 rbuels ctrl-z, bg
01:16 rbuels ?
01:17 * deafferret refuses to admin he clicked the Index button in phpMyAdmin
01:17 rbuels ofmg
01:17 deafferret lol
01:17 rbuels heh
01:19 deafferret gonna commute. will hop on from home
01:20 deafferret SWEAR TO ME THAT YOU'LL WAIT FOR ME
01:20 * deafferret cries
01:21 rbuels vnc?
01:21 rbuels vncserver?
01:21 deafferret holy crap! it worked
01:21 deafferret apparently threatening to leave makes all the difference, 45m later
01:21 rbuels and now my suggestions are moot.
01:21 rbuels moot!
01:22 deafferret dammit... relaunching all those damn PuTTy's
01:23 deafferret I never wanted to be a DBA... I wanted to be a lumberjack!! Leaping from tree to tree as I flot down the mighty something or other
01:23 deafferret float, even
01:24 rbuels mighty Columbia
01:24 rbuels i think
01:26 deafferret partitioning looks pretty cool for this date junk I got stuck doing today
01:26 deafferret MySQL partitioning that is
01:28 rbuels partitioning tables?
01:28 rbuels into....more tables?
01:29 deafferret umm.... so MySQL partitioning lets you tell MySQL to physically split up a table under the hood, based on a key that you specify
01:29 deafferret so, in my case, partition these 3M rows by date. then I can litterally drop the 2003 partition in a couple seconds, rather than hours of delete
01:29 rbuels yeah.
01:30 deafferret bad part is you have to keep rolling new years in manually
01:30 deafferret every few years
01:30 rbuels eew.
01:30 deafferret ya... -shrug-
01:31 deafferret I'm surprised they didn't design that for 100% automation
01:32 deafferret grr... yup. this is what I thought -- this index lets me select faster, but delete is even slower
01:32 deafferret so I'm in a pruning nightmare of 20M rows
01:32 deafferret so I guess I'll drop this .... hmmm
01:35 rbuels screen?
01:35 rbuels commute?
01:36 deafferret In a former life, I had a big pruning framework that dropped all indexes, deleted one day at a time (or whatever, depending), and just looped all night in small bites....
01:36 deafferret and put the indexes back, re-opt'd the table
01:37 rbuels hmm
01:37 rbuels would be nice to have some cpan modules for doing things like this
01:37 rbuels i have to go afk for a bit
01:37 rbuels i have to commute.  housesitting for somebody.
01:37 rbuels bbiab
01:37 * deafferret waves
02:00 deafferret damn. 22m to drop an index
02:02 deafferret huh? 1m to drop the 2nd index
02:06 deafferret oh, sweet!! deletes FLY now that I got those damn indexes off it
02:10 philsf joined #bioperl
04:15 gokul joined #bioperl
11:32 kyanardag_ joined #bioperl
14:51 basu joined #bioperl
16:02 basu hello
16:27 rbuels mornin'
16:28 * deafferret waves
16:28 rbuels or, um, evenin' in europe
16:28 rbuels i guess that would be
16:28 rbuels "evenin' guv'nor"
16:29 rbuels :-p
16:49 rbuels where is pyrimidine's blog?
16:50 * rbuels looks penetratingly at deafferret, cause calling him would be useless.  Cause he's deaf.
16:51 deafferret google FAIL
16:51 rbuels ENOPYRIMIDINE
16:52 rbuels (one of ewilhelm's email addresses is enobacon@something.  which is amusing.  LOLZ)
16:52 deafferret what's ENO?
16:52 * deafferret knows emo.
16:52 * deafferret no knowns eni
16:52 deafferret eno either
16:52 rbuels it's like an old C-style error constant
16:53 deafferret oh, right
16:53 deafferret C
16:53 rbuels E ~ error
16:53 deafferret jokes
16:53 rbuels C is very funny.
16:53 rbuels ha. ha.
16:53 * deafferret screams "BUNCHA NERDS!!" and leaves
16:53 * deafferret punches rbuels and takes his lunch money
16:53 * rbuels cries and goes to the garage to read comic books
16:53 deafferret http://scsys.co.uk:8001/31229
16:53 deafferret stupid CPAN
16:54 rbuels unnh.
16:54 rbuels yeah i saw you paste that into perl-help
18:11 deafferret rbuels: so what bioperl-l post was I supposed to read? "bioperl reorganization"? you boat rocker, you
18:14 deafferret laugh   I bet somebody whose name rhymes with "Jay Hannah" would probably do grunt work to help with this also
18:18 rbuels lol
18:18 rbuels pyrimidine quite properly chided me for not reading up on what's been done on this prior to now
18:19 deafferret mmmm chiding
18:19 rbuels with a side of cole slaw
18:19 deafferret hard to read this w/o my eyes glazing over
18:19 rbuels the upshot is just "yes we should break it out into smaller modules but it's a lot of work and we don't really know how"
18:20 rbuels i have a strategerie for this i think
18:20 rbuels for semi-automatically extracting sets of modules and getting the dependencies right
18:20 deafferret Why isn't Bio::Tree::* it's own CPAN release and if other CPAN releases depend on Bio::Tree::* , then they list that in their deps list -- just like the rest of CPAN
18:20 rbuels yes, that is what should be done eventually
18:21 deafferret cool. so. Bio::Tree on github?
18:21 rbuels but the monolith has to be chipped away in layers
18:21 rbuels hah.
18:21 rbuels Bio::Graphics was an example of something that could be chipped off the outside of the monolith
18:21 rbuels and it was.
18:21 rbuels exposing deeper layers
18:21 rbuels there are other things that can be chipped off now, but they have to be identified
18:23 rbuels first step i think would be to make a Task::BioPerl that depends on everything that used to be in bioperl, including bio::graphics
18:24 rbuels well, it would just be bioperl-live and Bio-Graphics right now
18:24 rbuels and then as distros are extracted from bioperl-live, you can add those as dependencies to Task::BioPerl
18:24 rbuels thus extracting them, put also preserving a single point of install
18:24 rbuels i'll write an email about this.
18:25 deafferret Bundle::BioPerl?
18:25 rbuels Bundle:: is not what the cool kids are using anymore
18:25 rbuels so says #toolchain anyway
18:25 * deafferret desperately wants to be cool
18:25 * rbuels smokes a cigarette and coughs
18:26 * deafferret does too
18:26 * deafferret feels cooler
18:28 deafferret oh shit, you went completely off the reservation at 2am
18:29 rbuels "off the reservation" haha
18:29 deafferret "because it is all like this (except for BioPerl)" rolf
18:32 deafferret "small and focused"... that's what she said
18:32 rbuels hey, i use it well.
18:35 deafferret "sheer property of legacy."  what does that mean?
18:35 deafferret "However a large set of modules in bioperl were effectively donated by the author, so they will fall to the core devs to maintain by sheer property of legacy."
18:39 rbuels meaning these modules get written and abandoned in bioperl
18:39 rbuels and if bioperl wants to have these modules, it has to maintain them
18:39 rbuels this is actually *the* big problem in academia, projects are short-term and there is no incentive to maintain software
18:40 rbuels because in academia, the currency is publications, not dollars
18:40 deafferret sure, but unless we're talking about Bio::Root dogma then whoever cares should be able to fix whatever, as long as the tests keep working
18:40 rbuels yeah you would think
18:41 * deafferret agrees with your academia jive
18:41 deafferret PS, corn is a fucker
18:41 rbuels i saw a website about that once, i think
18:41 rbuels that and octopuses
18:41 rbuels octopi?
18:42 deafferret anyhoo, do we have a BioPerl project today? or am I back to Erlang?
18:42 rbuels a good bioperl project would be to find another chunk of code to extract
18:43 deafferret Bio::Unrelated::To::Other::Things perhaps?
18:43 rbuels yeah.
18:43 rbuels i have to work on this grant though
18:43 rbuels that's what i was supposed to be doing last night
18:43 rbuels instead of ranting on bioperl-l
18:43 deafferret perhaps I need to whet my SVN commit bit teeth on a bugzilla
18:44 rbuels not a bad idea
18:45 rbuels or you could take my horrible dependency-finding scriptie and make it use  some Devel::MagicallyFindDependencies or whatever magic
18:45 rbuels so that it is actually not an ugly hack
18:45 rbuels and then it could be used to inspect and find pieces of bioperl that can be bitten off
18:46 deafferret and this scriptie lives where?
18:46 rbuels in the sgn svn repo, but i can nopaste it to you
18:46 rbuels or i could github it
18:48 deafferret http://headrattle.blogspot.com/2009/05/​perl-template-toolkit-and-graphviz.html
18:49 rbuels wow that's cute
18:50 * rbuels looks at the code
18:52 rbuels my script is basically the same thing, except for perl modules and instead of a graphviz it generates requires and build_requires syntax for a Build.PL
18:53 rbuels but i have a feeling there is something more reliable than grep-like stupidity for extracting what the dependencies are
18:53 rbuels there's all kinds of devel:: magical things on cpan
18:54 deafferret ya, you'd think. I spent a few hours at YAPC::NA hypothesizing about that for enlightened perl extended core hand waving
18:56 * rbuels waves hands magically and searches cpan
18:56 deafferret they have grant proposals on CPAN now? sweet!
18:57 rbuels link?
18:57 * deafferret was kidding. are you drunk?
18:57 rbuels oh.
18:57 rbuels i get it now.
18:57 rbuels no, it's only noon
18:58 rbuels i won't be drunk until at least 2
18:58 rbuels lol
19:01 rbuels perl -MDevel::Dependencies <perl_file>
19:02 rbuels perl -MGrant:
19:02 rbuels perl -MGrant::NSF -e write > chlamybase_grant.txt
19:02 perl_splut heheh
19:07 deafferret laugh... search for easy-to-fix-bugzilla not going well  :)
19:13 * deafferret chews on 2515
19:30 deafferret dev.open-bio.org -- interesting... it never likes my password the first time. It like it the second time.
19:30 deafferret svn checkout --depth immediates is SLOW... getting a list of all branches
19:30 deafferret apparently the server is spinning through, and then not sending, all the stuff I don't want
19:32 deafferret ooo  bioperl-live-branches/yapc10hackathon
19:40 deafferret oh crap. I'm in. truly, BioPerl is doomed
19:45 deafferret hmm... it liked my password fine first time that time. must of been me
20:04 driley_ joined #bioperl
20:36 * rbuels hands deafferret ssh-keygen
20:36 * rbuels hands deafferret ssh-agent
20:36 * rbuels hands deafferret ssh-add
20:36 * rbuels is away, don't talk to him
20:36 rbuels !
23:30 kyanardag_ joined #bioperl

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