Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-02-22

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

All times shown according to UTC.

Time Nick Message
00:43 brandi joined #bioperl
00:44 brandi left #bioperl
06:53 bag joined #bioperl
14:06 brunov joined #bioperl
14:06 brandi joined #bioperl
14:14 brandi joined #bioperl
14:14 brandi left #bioperl
16:09 pyrimidine joined #bioperl
16:37 perl_splut joined #bioperl
17:39 deafferret @
18:49 bag_ joined #bioperl
21:06 brunov joined #bioperl
21:07 kyanardag_ joined #bioperl
21:18 deafferret say cheese!
21:18 deafferret $row->{utr} =~ s/5pUTR/FpUTR/;   # So TT can cope with the hash keys
21:32 rbuels cheese!
21:32 rbuels that's almost as good as tr/I/1/
21:47 deafferret sigh.... with CGI.pm how do I get Apache to run cheese.pl and save and launch the client's .xls handler as filename cheese.xls ?
21:52 rbuels deafferret: print "Content-type: application/excel\n"
21:52 rbuels or whatever the real content type is
21:53 rbuels i don't remember what the mime type for excel is
21:53 deafferret rbuels: ya, but then the filename on the client side is cheese.pl
21:53 rbuels oh.
21:53 deafferret I think "Content-Disposition", "attachment;filename=   might do it
21:53 rbuels maybe
21:53 rbuels yeah i think so.
21:53 rbuels cause the browser will open a dialog of what to do with it
21:54 rbuels and they can just open it
21:57 deafferret print $q->header( -type => 'application/xls', '-Content-Disposition', "attachment;filename=cheese.xls"
21:57 deafferret );
21:57 rbuels you can do that too
21:57 * deafferret once again fails to understand Lincoln's penchant for -arg =>
21:58 rbuels well, it keeps away confusion with subroutines ...
21:58 rbuels but it's a pain on the implementation side
21:59 rbuels (meaning for the bloke who's actually writing the function taking the args)
21:59 deafferret ... wut? are you drunk?   ;)
21:59 rbuels meaning -arg is not going to be mistaken for arg() like arg => might be
21:59 rbuels doesn't that happen sometime?
21:59 rbuels s
21:59 rbuels ?
21:59 rbuels never actually happened to me ...
21:59 rbuels but i've heard stories.
22:00 deafferret arg => 7   gets mistaken for    arg() => 7     ???
22:00 rbuels i seem to recall somebody once suggesting that it might be under certain circumstances in some situations
22:00 deafferret uh-huh
22:01 * deafferret boos
22:01 rbuels yep.
22:01 rbuels thank you, thank you.
22:01 deafferret I've never seen that anywhere except Steinware  :)
22:02 rbuels oh, my wife and i had some of that in our wedding registry
22:03 deafferret and were your guests confused, thinking that your wedding was actually a sub-routine? I think not.
22:18 deafferret I submit to you sir, that if the only way you can tell the difference between a wedding and an argument is the '-' character; that you should not be married.
22:19 rbuels I submit to you sir, that weddings and arguments are so intimately related as to be very easily confused!
22:21 * deafferret weeps for the institution
22:22 * rbuels pines for the fjords
22:24 * deafferret wonders why gzip -d on 33GB take CPU time away from other processes
22:24 * deafferret brags about this 129TB mount point over here
22:26 rbuels wow, that's a lot of T
22:26 rbuels is that unomaha or $otherjob
22:26 deafferret Firefly (UNL). Mostly full of LHC data   :(
22:28 * deafferret shakes his fist at Higgs bonons everywhere
22:28 deafferret err bosons
22:28 rbuels bonons!
22:28 rbuels i love it
22:28 rbuels i'll tell that one to my father in law
22:28 deafferret soon, they will realize they were looking for the wrong particle
22:28 rbuels he's in the physics dept at cornell, does a little peripheral work on the lhc
22:28 deafferret then THEY WILL ALL BE SORRY
22:29 rbuels the Hannah Bonon
22:29 deafferret plural
22:29 rbuels the Hannah Bonons
22:29 rbuels ?
22:29 deafferret yes. there are 2
22:29 deafferret we will discover the first in 2013. the second in 2023 just before the great phant crash
22:29 pyrimidine deafferret: -arg syntax is to distinguish named parameters
22:30 * deafferret startles
22:30 pyrimidine heh
22:30 * deafferret thought this was #rbuels-silliness
22:30 deafferret pyrimidine: wut, pray tell, is the difference between 'named' and 'unnamed' parameters?
22:30 * rbuels groans
22:30 * deafferret takes copious notes
22:31 pyrimidine nope. this is actually something incorporated into perl6
22:31 rbuels eh?
22:31 * rbuels looks
22:31 pyrimidine look up perl6 signatures
22:32 pyrimidine also, one can use a variant of this with Method::Signatures w/ perl5
22:32 rbuels oh, yes i knew there were named parameters ...
22:32 rbuels but with a leading hyphen?
22:32 deafferret is someone trying to maintain multiple APIs --  (1,2,3) and (-one => 1, -two => 2, -three => 3)  ?
22:33 * rbuels seems to remember that this was the case, but does not trust his memory
22:33 * deafferret thought 'unnamed parameters' died at, say, 3 parameters.
22:33 rbuels oh ho ho.
22:33 rbuels perhaps it "dies"
22:33 rbuels but it's a living death
22:33 pyrimidine rbuels: the hyphen makes it easier to distinguish a named parameter from an unnamed one
22:33 deafferret foo(@my_seventeen_carefully_ordered_gems)  ???
22:35 pyrimidine the problem is, in p5, the fat comma is essentially just a comma, so "'one' => 1" is the same as "'one', 1"
22:35 rbuels that's true, you could just be doing => as a fat comma
22:35 * deafferret probably gets it
22:36 rbuels come to think of it, i think we had our Steinware pattern as "Fat Comma" in our registry .....  it's all coming together
22:36 pyrimidine the only difference is during the initial parse during compilation; a fat arrow allows the lhs to be uncommented in certain circumstances
22:37 pyrimidine (one of the reasons MooseX::Types has an issue with them)
22:37 pyrimidine but in all other respects it's just a comma
22:38 pyrimidine Oh, can't forget MooseX::Method::Signatures, if you fancy MooseX::Declare
22:39 pyrimidine anyway, in perl6, there is no fat comma; 'one' => 1 is a Pair, not a list of two values, so one can now distinguish them
22:40 pyrimidine and, with signatures, one can do something like:
22:40 pyrimidine sub foo (Int :x) {...}
22:40 pyrimidine and call it as 'foo(x => 1)'
22:40 pyrimidine or 'foo(1)'
22:43 pyrimidine perl6 will own your asses
22:43 pyrimidine :-D
22:46 brunov joined #bioperl
22:46 deafferret so BioPerl6 is going to let people call methods with named and/or unnamed parameters, and document all that?
22:47 deafferret at @work we just outlawed unnamed parameters for any list > 2. problem solved...
22:48 deafferret .oO( why am I getting 18,19, or 20 results from these 3 reports?? )
22:49 pyrimidine deafferret: I think there is a way to have required named params with perl6, would have to check the spec.  But (unless specified), I believe one could have it either way.
22:49 pyrimidine ok, commuting &
23:02 deafferret ooo... burned by a missing group by arg
23:02 * deafferret uses more -s
23:07 brunov p54life!
23:07 * brunov ducks
23:08 deafferret don't think anyone in here is p6 addicted  :)
23:08 deafferret pyrimwhateveridine is the only person I know who actually uses it
23:09 brunov I played around with it, it's quite nice. I can't stand the slowness of it though. It's a deal breaker for me.
23:09 deafferret slow doing what?
23:09 brunov anything
23:10 brunov I'd say it's an order of magnitude slower than perl
23:10 brunov *perl5
23:10 deafferret Rakudo?
23:10 brunov at least when I tried it, which was somewhere around... beginning of 2009 I think.
23:11 brunov yes
23:11 rbuels brunov: they're working on performance
23:11 rbuels brunov: it'll get fast.
23:11 rbuels just wait
23:11 deafferret brunov: probably runs OK on the right GPUs  :)
23:11 brunov rbuels, I surely hope so.
23:11 driveby_bot joined #bioperl
23:11 driveby_bot /home/svn-repositories/bioperl: r16869 (kortsch) : bug fix bugzilla #3014
23:11 driveby_bot Diff: http://tinyurl.com/yzgc7he
23:12 brunov I love what Larry has done to the language though
23:12 brunov at least the bits I've seen and understood
23:13 brunov last year's advent calendar was a really nice read
23:15 rbuels the language design is certainly groundbreaking.
23:16 rbuels the really big thing is parrot, if it takes off as a common platform for dynamic languages
23:16 rbuels because *then*, everybody will be able to use eachothers libraries
23:49 driveby_bot joined #bioperl
23:49 driveby_bot /home/svn-repositories/bioperl: r16870 (kortsch) : Remove unix command from constructor.
23:49 driveby_bot Diff: http://tinyurl.com/ybxn4pl
23:54 deafferret so under perl6/Parrot people will use software *I* write? sounds like a bug already.  :(
23:55 deafferret MWOOOOHAHAHAHHAAHAHAHAH  http://clab.ist.unomaha.edu/s​vn_co/CLAB/RT922/report7.html

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