Camelia, the Perl 6 bug

IRC log for #bioperl, 2013-02-15

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

All times shown according to UTC.

Time Nick Message
04:12 census joined #bioperl
04:26 carandraug joined #bioperl
04:26 carandraug joined #bioperl
04:33 census joined #bioperl
04:33 census hi
06:59 carandraug joined #bioperl
08:51 sivoais joined #bioperl
10:15 ObseLeTe joined #bioperl
14:03 looper joined #bioperl
15:59 census joined #bioperl
16:06 census hi! :)
17:35 scottcain joined #bioperl
17:36 scottcain_ joined #bioperl
17:37 scottcain Hi fellow bioperlers,
17:38 scottcain I'm taking a look at the new (to me) FeatureIO repository, and I have a few quesitons.  rbuels any familiarity with it?
17:39 rbuels scottcain: yes, some
17:39 scottcain Anything I should know?  Will it be wickedly faster since it uses Bio::GFF3?
17:40 rbuels scottcain: i don't think it's much different from the old stuff, just split out
17:41 scottcain Hmm, doesn't feel that way at a cursory glance--it uses new-fangled stuff like given-when (never used before; I feel so old).
17:42 rbuels really?
17:42 rbuels oh, there
17:42 rbuels huh
17:42 rbuels scottcain: it's got the same interface as the old stuff though i think
17:42 rbuels doesn't it?
17:43 rbuels scottcain: i think pyrimidine wrote most of that stuff
17:43 scottcain I'm assuming so--don't actually have it working in my hands yet though.
17:43 * rbuels wouldn't use given/when
17:43 scottcain why not?
17:43 rbuels scottcain: doesn't really add that much over if/else
17:44 rbuels scottcain: and it's not as backcompatible
17:44 scottcain now you sound old :-)
17:44 * rbuels shrugs
17:44 rbuels heh
17:44 scottcain but I agree with you.
17:46 * rbuels does not have strong opinion about it
17:46 rbuels programming is so 1997
17:46 rbuels i just go be a social media consultant
17:46 scottcain sounds like a good gig.
17:47 rbuels lol
17:52 rbuels scottcain: the parser in the new featureio looks more correct than the old one, but it's not going to be particularly fast
17:52 rbuels scottcain: it's not the low-level parser i wrote, which is probably about as fast as a pure-perl parser can get
17:52 rbuels was toying with the idea of doing an XS one, actually
17:52 scottcain I guess more correct is better, as long as I didn't write code depending on it being less correct :-)
17:53 rbuels maybe i should write the XS parser instead of working on jbrowse
17:53 * rbuels chuckles
17:54 * scottcain forgot that his perl environment is hosed since upgrading the OS. now it's easier to just go fire up an AMI to develop on.
17:54 scottcain "-bash: /usr/bin/perldoc: Permission denied" WTF?
17:55 * scottcain wonders why the heck perldoc isn't set to executable.
17:55 scottcain I know I didn't do that.
18:01 census joined #bioperl
18:03 deafferret my dissertation relies on those bugs, please don't correct. kthx
18:04 census what bugs?
18:05 deafferret census: just joking based on some stuff you missed while you were away
18:05 census oh!
18:05 census what is your dissertation on?
18:05 deafferret no grad school would have me ;)
18:06 census how do you know?
18:06 census what do you want to work on?
18:06 deafferret making money :)
18:07 census economics?  finance?
18:07 deafferret i code business solutions for a living. I'm a bio* hobbiest.
18:07 deafferret well, I mostly just hang around and make unfunny jokes.
18:07 deafferret and quote AWESOME rap artists
18:08 deafferret http://bioperl.org/wiki/User:Jhannah
18:10 scottcain anybody know off hand if Test::Warn is in perl core 5.10?
18:10 * deafferret should know how to know that
18:10 scottcain (yes, I'm heading off to look for myself now)
18:10 scottcain deafferret: me too.
18:11 deafferret scottcain: http://stackoverflow.com/questions/204​9735/how-can-i-tell-if-a-perl-module-i​s-core-or-part-of-the-standard-install
18:11 deafferret corelist looks cool
18:12 census cool!
18:12 census may i ask a perl q?
18:12 census :)
18:12 deafferret don't ask to ask, just ask
18:12 deafferret those who ask to ask risk axing
18:14 scottcain corelist says "Test::Warn was not in CORE (or so I think)"
18:14 scottcain Nice that it doesn't want to be too definitive :-)
18:14 census axing = ban?
18:14 deafferret census: again, just kidding.
18:15 scottcain Oh, geez, don't do that.
18:16 deafferret scottcain: i'll stop when you stop being so awesome.
18:16 scottcain heh
18:18 * scottcain thinks deafferret may have scared census out of asking his question.
18:19 scottcain Well, at least now he can't ax him :-)
18:19 deafferret i'm just a sucker for alliteration. however you spell that.
18:36 census Hey!
18:36 census #perl LOVES to ban
18:36 census I got scared
18:36 census THank you scottcain :)
18:37 scottcain ls
18:37 scottcain Hah--wrong window.
18:41 scottcain_ joined #bioperl
18:44 census huh?
18:50 census anybody still here?
18:55 deafferret census: pong
18:55 deafferret ya, I've given up on #perl several times
18:56 census Here is my code: http://scsys.co.uk:8002/230568?ln=on&amp​;tidy=on&hl=on&submit=Format+it!
18:57 census locations.txt = Joseph,Bonsigmore,New York,1913
18:57 census The problem is that Joseph may go by Joey or Joe.
18:58 census So in the url, there might not be a match for that reason.
18:58 census Or worse yet, there might be a Joe and a Joseph, in which case I want to go to the >= 2 loop rather than the == 1 loop
18:58 census so I've made a database of potential nicknames
18:58 census for example
18:58 census nicknames.txt
18:58 census row 1= Joe,Joseph
18:58 census row 2= Joey,Joseph
18:59 census What I'd like to do for Joseph,Bonsigmore,New York,1913 is check Joe,Bonsigmore,New York,1913; Joey,Bonsigmore,New York,1913; and Joseph,Bonsigmore,New York,1913
19:00 deafferret yup. so a foreach my $nickname (@$nicknames{$thisname}) { ...
19:00 census yes, something like that
19:00 census some people have no nickname
19:00 census like Keith is just Keith
19:01 census or it might go the other way around.
19:01 census locations.txt = Ben,Bonsigmore,New York,1913
19:01 deafferret yup. if there aren't any thats fine -- the foreach simply does nothing
19:01 census I need to check that  Ben = Benjamin, Ben = Benny or Ben = Benjamin
19:01 census I'm not sure how to write this.
19:02 census And here's the the thing:
19:02 census Suppose for Joseph,Bonsigmore,New York,1913, the hits are: 0 for Joe,Bonsigmore,New York,1913; 1 for Joey,Bonsigmore,New York,1913; and 0 for Joseph,Bonsigmore,New York,1913
19:02 census I need to know to scrape the Joey one for Joseph
19:02 census OR suppose they are:
19:03 census Suppose for Joseph,Bonsigmore,New York,1913, the hits are: 1 for Joe,Bonsigmore,New York,1913; 1 for Joey,Bonsigmore,New York,1913; and 0 for Joseph,Bonsigmore,New York,1913
19:03 census I need to know to put that in the >= 2 section because 1 + 1 + 0 >= 2
19:05 deafferret you currently have one 80-line loop. That's too big to be healthy. Time to split out some subroutines.
19:05 deafferret separate your fetching logic from your nicknaming logic from your parsing logic
19:05 deafferret and it'll be more obvious where to stick your new code
19:06 deafferret refactoring++
19:08 census i'm not a big programmer
19:08 census so . . .
19:08 census i know what the program is doing
19:08 census but adding functions for the sake of it (subroutines, however it is called) does make me a little nerous
19:08 census i can tell you in words each part of the program
19:08 census but if the subroutine woudl help you, i'm all ears
19:09 deafferret you'll need to get over your nerves. Your program is becomming too large for a single loop. :)
19:09 deafferret this isn't about helping me, I'm trying to help you make your job easier.
19:10 deafferret if your code is 20,000 lines and you have ONE loop it will be completely impossible to add the next feature to it.
19:10 deafferret well... possible, but please don't.
19:10 deafferret best to learn the techniques everyone uses so that your code can scale gracefully. one of those basic tools is a subroutine
19:11 census ok . . .
19:11 census i mean the code is only for my eyes ultimately
19:11 census and i know what every part does.
19:11 deafferret yes, and YOUR eyes will start to bleed as it continues to grow. :)
19:11 census but can we get back to the nicknames please?
19:11 census i know where to stick the new code more or less
19:12 census i don't know how it should go
19:12 census lines 51-101 are virtually unaffected
19:12 deafferret your searchHits 0 vs. 2 vs. 1 is complicated BECAUSE you're now trying to figure out how to shoe-horn your NEW nickname problem INTO your previous solution. If you separate the two problems away from each other then you have two small solutions instead of a growing mound of spaghetti :)
19:12 census ok.  we can do that part later one
19:13 census the 0 v.s. 2 v.s. 1 later on
19:13 census but first, even just dealing with the nick.txt
19:13 deafferret foreach my $nickname (get_nicknames($thisname) { .... }   sub get_nicknames { my ($this) = @_; # put lookup here return @all_nicknames; }
19:14 pyrimidine joined #bioperl
19:14 deafferret subroutines are your friend. :)  every piece of software you use has hundreds/thousands of them
19:14 * pyrimidine has a flashback of Carl Sagan: billions and billions of stars
19:15 deafferret hi py<TAB>!
19:15 deafferret fancy seeing you here!
19:15 pyrimidine :)
19:15 pyrimidine been a little busy lately
19:17 pyrimidine scottcain: working on FeatureIO?  Feel free to rip, tear, burn as needed.
19:18 pyrimidine I'm planning on focusing on the bioperl release, then the splitting out stuff
19:18 pyrimidine probably starting with help on FeatureIO
19:51 scottcain hi pyrimidine, carne pointed out a problem merging bioperl commits with featureio commits that related to something I did ages ago.  I think I fixed it in the featureIO master (and even added a test!)
19:51 pyrimidine scottcain: testz r good
19:52 pyrimidine s/good/gud/
19:53 scottcain Of course, i have no idea if the current incarnation of FeatIO works with the Chado loader.  That's still on my todo list.
20:37 pyrimidine joined #bioperl
20:37 pyrimidine joined #bioperl
20:52 census deafferrret: I figured it out! :)
20:54 deafferret census: good work :)
20:59 census thank you!!
21:03 scottcain_ joined #bioperl
21:05 pyrimidine joined #bioperl
21:06 pyrimidine joined #bioperl
21:08 pyrimidine joined #bioperl
21:08 pyrimidine joined #bioperl
21:18 scottcain joined #bioperl
21:45 scottcain_ joined #bioperl
22:10 scottcain_ joined #bioperl

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