Perl 6 - the future is here, just unevenly distributed

IRC log for #bioperl, 2014-01-30

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

All times shown according to UTC.

Time Nick Message
02:13 looper_ joined #bioperl
11:49 carandraug joined #bioperl
14:48 average joined #bioperl
15:40 [Coke] joined #bioperl
15:40 [Coke] Hio.
15:44 [Coke] Wonder if anyone from Perl 6 has swung by to see what sorts of things (aside from moar speed) the bioperl folks would need from perl 6 to make it a choice.
15:46 carandraug joined #bioperl
15:46 carandraug joined #bioperl
18:20 carandraug joined #bioperl
18:29 jhannah [Coke]: here's the only bioperl/6 thing I'm aware of: https://github.com/cjfields/bioperl6
18:32 carandraug I'm hoping for Moose based BioPerl
18:33 jhannah carandraug: that'd be https://github.com/cjfields/biome
18:33 carandraug jhannah, I know.
18:33 jhannah ooo! and he touched it recently :)
18:33 jhannah good touch! :)
18:34 carandraug every time I look into BioPerl source I just give up. It's never gonna be released. But to start over is also too much work
18:35 carandraug specially when I only have 1 month left to finish the thesis
18:35 jhannah what's never going to be released?
18:36 carandraug jhannah, a new thing based on bioperl-live
18:36 jhannah "released" meaning what? CPAN? Do you need help with something?
18:36 jhannah I'm in a releasing kind of mood today. :)
18:37 jhannah #shipit
18:37 carandraug jhannah, yes, that's what I mean with releases. I don't need help with anything. Just complaining that I find contributing to BioPerl is a bit discouraging
18:38 jhannah oh? is there something I can do for you?
18:38 carandraug not because the rest of people mind you. Not at all
18:38 jhannah I could throw a few hours at something if it would help you somehow
18:40 carandraug jhannah, I appreciate the offer but it is not needed. I could fix the bugs if I wanted, I just find it a waste of time because I'm unsure of whether the project is going
18:40 jhannah well, it's going wherever whoever cares about it takes it :)
18:40 jhannah Perl's been "dying" for 26 years now. :)
18:41 jhannah be the change you want to see in the world.  lol
18:42 carandraug and I found a bug earlier this week which left me appalled could still exist http://permalink.gmane.org/gmane.​comp.lang.perl.bio.general/26997
18:44 carandraug how is it possible that something such as "@seqs = map {Bio::SeqIO->new(-file => $_)->next_seq;}  @files;" went unfixed for so many years? I started to fix it, and each time I did so, the exact same issue was still somewhere so I gave up after a while
18:44 jhannah hmm... I don't understand this website. Did anyone reply?
18:44 jhannah well, I can help you fix that one if you want. step 1: create a failing test it t/
18:44 jhannah *in
18:44 carandraug jhannah, yes, someone replied. Apparently is a know thing that BioPerl sometimes modifies $_ so people better not use it. Really?
18:46 carandraug jhannah, thank you. But I don't need help fixing it. I can do it on my side and I did start doing it. But I want to share my code with others, so there's not really a point of fixing it.
18:46 jhannah why can't your code be on github, shared with the whole world?
18:46 carandraug jhannah, it is in github
18:47 carandraug my side means my repositories, not on my computer
18:48 carandraug that one in specific is a private repository until it gets accepted for publication. Hopefully I'll submit next week. And it does make heavy use of BioPerl, and all the logic of what I'm doing I contributed to BioPerl
18:48 jhannah cool :)
18:49 jhannah good luck with the pub!
18:49 carandraug jhannah, I have a few paragraphs on the text bashing at people who make their own projects instead of contributing upstream, or releasing them as a proper library on CPAN
18:49 carandraug my supervisor will probably edit them a lot
18:50 carandraug but I want people to just grab my paper, run scons, and be done with it. Having to install the latest revision from bioperl is not something that helps with reproducibility.
18:52 carandraug bottom line was, I did start fixing the issue in BioPerl. But after a while it dawned on me that the fix would never be released. And even if it was released, since the problem is so common in BioPerl, even after I fixed in the places that affected me, the recommendation would still be to avoid $_. So I simply did the easy thing, and used a for loop without $_
18:52 looper joined #bioperl
18:53 jhannah -nod- ya, sounds like you've chosen your battles wisely.
18:54 jhannah huh, that's dumb. side-step, move on. :)
18:54 carandraug jhannah, by the way, mailman is nicer than gmane http://lists.open-bio.org/pipermail​/bioperl-l/2014-January/071354.html
18:55 jhannah that said, tilting at windmills can be fun sometimes too. And why not change something if you care enough about it to follow through? -shrug-
18:55 carandraug the fact that I;m running out of time is teaching me better. Had this been a year ago, I'd have spent the whole week tracking down that issue
18:56 jhannah ya, I hear ya. it's easy to yak shave forever
18:57 jhannah Chris++ # http://lists.open-bio.org/pipermail​/bioperl-l/2014-January/071356.html
18:58 jhannah "feel free to flame-thrower the problem" :)
18:58 * [Coke] grabs a copy of bioperl6 and checks against inevitable bit rot.
18:59 carandraug jhannah, like I said, I would if I could now. But I think there's more important things needed in BioPerl. Mainly a clear plan of what it will be.
19:00 jhannah ah. tilting at ALL the windmills simultaneously ;)
19:00 jhannah a crusade only of the brave
19:01 carandraug it's not all. But unless BioPerl gets changed a lot, I don't believe it will last long other than something used to run ancient code.
19:01 jhannah perhaps. like "Perl is dead," I've been hearing that since I first heard of BioPerl. Time will tell.
19:02 carandraug jhannah, last year I started to create separate repositories for small parts of BioPerl. While it seemed a good idea for everyone, no one continued
19:02 jhannah hell, all of bioinformatics is dead. still. :)
19:02 jhannah ah, yes. exploding the monolith. another worthy cause. I'd contribute to that effort if I was working with it at all
19:03 jhannah if people are going to change things, please split them up
19:03 carandraug it's not dead. And many still use perl for it. All at my university use R and Perl for it. But not one uses BioPerl. They all have written their own routines to parse fasta files
19:05 [Coke] ah, that actually answers my question. :)
19:05 carandraug jhannah, exploding it is simple. I did a few and wrote very simple instructions to to do it again while conserving the history. What is needed a clear idea of what each piece should be. I even finished some nice bundle for distzilla and PodWeaver
19:05 carandraug [Coke], what was your question?
19:05 [Coke] I assume the main thing a perl6 variant would need is to be comparable on speed?
19:05 carandraug I was not here when you made it
19:06 [Coke] or do all these one offs have something that sets them apart?
19:06 [Coke] carandraug: I didn't ask it explicitly.
19:06 carandraug one offs?
19:07 [Coke] "They all have written their own routines"
19:07 [Coke] ^^
19:07 carandraug [Coke], ah! No. They all suffer from Not Invented Here syndrome
19:08 carandraug either that or sheer ignorance that there's things such as CPAN
19:10 [Coke] ah. ok.
19:11 carandraug I did send an email around the department asking about why they don't use BioPerl. The only one that replied me was that they don't know how to use it. I offered to teach them but then we never managed to find a time that suited us all
19:11 carandraug it's surprising that they find it easier to write their own parser than read bioperl documentation
19:17 jhannah ya. the software development practices in all the academia I was ever in were abysmal.
19:20 carandraug jhannah, it's so bad, that coding guidelines are still accepted for publication
19:20 carandraug http://www.plosbiology.org/article/info​%3Adoi%2F10.1371%2Fjournal.pbio.1001745
19:21 carandraug "Best Practices for Scientific Computing"
19:22 carandraug it's interesting how it's needed 13 people to write that article
19:23 jhannah heh
19:25 carandraug and the main author is from the Mozilla foundation and apparently the owner of the software carpentry website that he talks about at the end as a place to learn how to code. I really hope that people will read it though I think a lot more could be taught to them.
19:27 carandraug and I read a really good opinion yesterday about reusing code, the reasons why there's no incentives for it, and how it's delaying research
19:27 carandraug if you wish to read https://www.siam.org/news/news.php?id=2131
19:30 jhannah heh "write the codes"
19:31 jhannah "just coding"  :)
19:32 carandraug jhannah, apparently it's very common in certain fields. Kinda of "this is the code I use to parse a fasta file, and this is the code I use to parse a genbank file. Together, they're 2 codes"
19:33 jhannah -nod-   good articles. thanks :)
19:35 carandraug jhannah, we are not alone http://stackoverflow.com/questions/277​3761/armchair-linguists-code-vs-codes-​or-why-i-write-code-and-my-manager-ask
19:37 jhannah every time Brits call it 'maths' I giggle
19:38 jhannah "I believe the proper term is codez, as in pls send me teh codez -- you must correct your manager at once."
20:57 [Coke] joined #bioperl
21:16 carandraug joined #bioperl

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