Perl 6 - the future is here, just unevenly distributed

IRC log for #bioperl, 2015-02-23

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

All times shown according to UTC.

Time Nick Message
02:46 ilbot3 joined #bioperl
02:46 Topic for #bioperl is now Be patient! People chat here daily, but not necessarily the minute or hour you wandered in. Leave your IRC client connected. | http://bioperl.org/wiki/IRC#Getting_help | http://www.bioperl.org/wiki/Using_Git | nopaste to gist.github.com
02:57 PetrVod` joined #bioperl
04:08 scottcain joined #bioperl
04:13 scottcain_ joined #bioperl
05:09 greencow joined #bioperl
09:45 _TK_ joined #bioperl
09:48 fruitorama joined #bioperl
12:11 scottcain joined #bioperl
13:24 scottcain_ joined #bioperl
13:59 PetrVod` joined #bioperl
15:39 greencow joined #bioperl
16:13 fruitorama joined #bioperl
17:27 fruitorama joined #bioperl
17:41 saiaku joined #bioperl
17:42 saiaku hi, is anyone aviable for a sec?
17:42 pyrimidine saiaku: good timing.  just happened to look in at the same time you posted
17:42 saiaku hi pyrimidine
17:43 saiaku Well, I have studied Java and C++/C. I'm studying Biocomputing and I'd like to know whether learning Perl will be uselful or it will be replaced by Java/Python/R :/
17:44 pyrimidine I know Perl and Python and R (and some Java and C).  They all have their advantages/disadvantages
17:45 saiaku I understand that analyzing big genomas (Big files and REGEX) is Perl's main goal, but I would like to know if it goes any further than that.
17:45 saiaku Cause I'm hearing that other languages are aiming to replace it so I'm not sure about it
17:46 pyrimidine saiaku: perl and python mainly wrap more optimized code that does the hard-core (e.g. computationally intensive analyses)
17:47 saiaku And Java?
17:47 saiaku Cause I'm hearing Python + Java makes a good combo and would replace Perl
17:47 saiaku Python + R*
17:58 pyrimidine sorry, had to step away for a sec
17:58 pyrimidine I agree a bit w/ Python there, not so much Java
17:59 pyrimidine but you definitely can't only use Python, or only use Perl, if you plan on doing any compute-intensive work.  You'll have to mix it up
18:00 pyrimidine The one 'killer' feature Python has over Perl (and IMHO currently makes it useful) is the ease of making native bindings
18:01 pyrimidine well, shouldn't say 'the one killer feature'; there are others.
18:02 pyrimidine Perl OTOH is quite fast, and generally faster than Python at many tasks, particularly text munging
18:02 pyrimidine but Perl XS (c/C++ bindings) suck
18:03 pyrimidine hope saiaku got that :)
18:19 PetrVod joined #bioperl
18:45 saiaku joined #bioperl
18:45 saiaku sorry pyrimidine  I had to go
18:46 pyrimidine np
18:46 saiaku So perl would probably be replaced by Python and R right?
18:46 pyrimidine No
18:46 saiaku I mean I have to choices at the moment
18:46 saiaku Learn C at a much higher level, or learn Perl
18:47 saiaku And atm I'm studying biocomputing so well
18:47 saiaku I'd have to decide
18:48 pyrimidine Python is a good choice for higher level (it has some basic types, nice layer for C lib bindings).
18:48 pyrimidine Perl5 is pretty good as well but has fallen behind on some of those things
18:49 pyrimidine Keep in mind Python also has problems.  GIL is a big one; it can be worked around but it's an issue
18:49 saiaku Wut problems is having Perl 5?
18:54 pyrimidine Perl5 can have C/C++ bindings but they're not easy to set up
18:54 pyrimidine via Perl XS
18:54 pyrimidine OOP is pretty hammered on (but it's that way in Python as well, there is no truly private data in objects)
18:55 pyrimidine main thing going for Python is community
18:56 pyrimidine Most current BI code is (simply put) written in Python.  Still a decent amount of Perl out there, but Python outnumbers it now I believe
18:56 pyrimidine Big Perl plus: CPAN
18:58 pyrimidine Python has a package installation system, but it seems (IMHO) fairly inconsistent
18:58 pyrimidine pip vs pip2 vs easy_install vs ....
18:59 pyrimidine If you have to pick a version of Python to code in, that's really tricky.  Most users seem to still be on Python 2.7
18:59 saiaku So python will outnumber perl in biocomputing?
19:00 pyrimidine currently, yes.  In five years?  Who knows.  I've started working in Perl 6 (which solves a lot of the Perl 5 problems) and I find it much nicer than Perl5 or Python, frankly
19:00 pyrimidine But it's an alpha release
19:01 saiaku C is not used in biocomputing right?
19:02 pyrimidine I wouldn't say that.  Most low-level code is still written in C/C++.  I don't see as much Java as I used to.
19:03 pyrimidine The rule I find works best is, if you build a very useful tool that's easy to install, the underlying implementation may not matter as much.  Perl/Python/C/C++ is everywhere
19:03 pyrimidine Pick what works best for your needs.
19:05 pyrimidine my 2c: Python might fit better based on your C/C++ background
19:06 pyrimidine good example of Python mixed w/C++ code: khmer: https://github.com/ged-lab/khmer
19:07 pyrimidine Nice example of Perl mixed with C bindings (w/ a bit of bioperl): Bio-Samtools: https://github.com/GMOD/GBrowse-Adaptors/tree/master/Bio-SamTools
19:08 pyrimidine saiaku: Now, regarding R: it depends on your needs, but Python has pandas, numpy, etc.  Not bad to learn R, but it's becoming less and less necessary depending on what you are trying to do
19:09 saiaku pyrimidine: so BioPerl is a framework inside Perl?
19:09 pyrimidine yep.
19:09 pyrimidine and biopython is the same for Python, BioJava for Java, etc.
19:11 saiaku I see
19:11 pyrimidine so much of the frameworks are in place for any of the langs you need.
19:11 pyrimidine *langs you choose
19:12 saiaku BioC ?
19:12 saiaku :P
19:13 pyrimidine saiaku: there is SeqAn (for C++): http://www.seqan.de
19:14 pyrimidine just released a v2.0
19:14 saiaku I see
19:14 saiaku I'd really like to go deeper into C
19:14 saiaku But I can't decide x_x
19:19 pyrimidine saiaku: depends on the area but there is lots of C code to use.  e.g.: htslib/samtools/bcftools (for NGS)
19:20 pyrimidine For a higher level language, python is a safe bet if you are just getting started.  I just happen to have started w/ Perl5, so I use it more frequently on tasks
19:24 saiaku Well I just want to decide on the course
19:24 saiaku So I want to find out what suits best for biocomputing
19:24 saiaku C or Perl
19:24 saiaku I know C, but the course goes deeper on it
19:25 saiaku Working with yacc lacc makefiles debugging improve compilation times...
20:07 PetrVod` joined #bioperl
20:16 fruittie joined #bioperl
20:18 greenlion joined #bioperl
21:30 greenlion joined #bioperl
21:31 greencow joined #bioperl
22:11 greencow joined #bioperl
22:14 greenlion joined #bioperl
22:17 greencow joined #bioperl
22:19 greenlion joined #bioperl

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