Time  Nick      Message
23:03 pdurbin   huh. nope. haven't seen that
23:03 pdurbin   "1.7% other"
23:03 sivoais   have you evaluated <http://www.renjin.org/>? It's R on the JVM. I'm wondering if I should support it in Statistics::NiceR
23:02 sivoais   that's the best way... one langauge is not enough! :-)
22:56 pdurbin   it is a many splendored app
22:44 sivoais   pdurbin: wait, there's an Rserve involved too!
22:42 sivoais   haha, what about a Perl script that depends on Java! I've done that before with Inline::Java :-) I used HtmlUnit that way.
22:41 sivoais   I feel like that could be done with configuration management
22:40 semiosis  generally speaking though i'm offended by a java app that depends on perl
22:40 semiosis  tell me more about your requirements for this installer :)
22:38 semiosis  idk
22:35 pdurbin   https://github.com/IQSS/dataverse/tree/master/scripts/installer
22:34 semiosis  installer?
22:34 pdurbin   semiosis: should we write our installer in jruby? Right now it's in bash and perl.
22:33 semiosis  pdurbin++
22:33 pdurbin   semiosis: when you get around to it, give me a shout
22:33 semiosis  i learned jruby contributing to logstash and liked that a lot.  it would be my go-to scripting language on the jvm if i needed one
22:33 pdurbin   people at work are looking at Spark
22:31 sivoais   <http://en.wikipedia.org/wiki/Apache_Spark>
22:31 sivoais   it looks like Scala is becoming a larger force in data science with things like Spark
22:31 pdurbin   I haven't even looked at the Java 9 link sivoais dropped in here earlier.
22:31 semiosis  i've wanted to play with clojure since i learned common lisp in college but haven't gotten around to it
22:31 pdurbin   Java 8 has a lot of nice stuff.
22:30 semiosis  i'm not really a fan of scala. tried it a few years ago and was thoroughly unimpressed.  i'm sure it's come a long way since then, but then again, so has java
22:30 pdurbin   semiosis: you too
22:30 pdurbin   sivoais: you could write an addressbookmvc app in one of them?
22:29 pdurbin   heh. "Abstraction Without Regret"
22:29 sivoais   pdurbin: neither yet. I just look at them for ideas. ;-) Haven't written anything in them yet.
22:28 pdurbin   sivoais: are you more into scala or clojure?
22:28 sivoais   though I don't like functional languages emphasis on writing parsers with parser combinators... it's just recursive descent.
22:27 semiosis  iirc twitter uses a lot of scala
22:26 pdurbin   we have an app at work in Scala: https://github.com/IQSS/DataTags
22:22 sivoais   It's an approach for code generation
22:22 sivoais   this <http://scala-lms.github.io/>
22:22 sivoais   heh, yeah or Scala. Scala has some really interesting compiler theory
22:21 sivoais   you should also take a look at the Java repos from <https://github.com/tokuhirom>
22:21 semiosis  jruby :)
22:21 pdurbin   If I were to sneak some non-Java onto a JVM at work I'm not sure it would be Perl 6. :)
22:20 semiosis  there's so many better languages on the jvm
22:19 sivoais   :-)
22:19 pdurbin   maybe lombok first though :)
22:19 pdurbin   I know! I should try it.
22:18 sivoais   Perl6 cuts that down a great deal more. And it runs on the JVM too. :-)
22:18 pdurbin   :)
22:17 sivoais   there are ways to make $p->y++ work, but that's just getting fancy ;-)
22:17 sivoais   my $p = Point->new( x => 10 ); $p->y( $p->y + 1 );
22:16 sivoais   you can do
22:16 sivoais   package Point; use Moo; has [ qw(x y) ] => ( is => 'rw', default => sub { 0 } );
22:16 sivoais   you don't but if you do
22:16 pdurbin   but yeah, reduced boilerplate in general
22:15 pdurbin   do you need getters and setters in Perl?!
22:14 sivoais   Lombok looks like it does for Java what Moo/Moose do for Perl.
22:14 pdurbin   fair enough
22:10 semiosis  pdurbin: that link to the @Data annotation page really demonstrates it as well as any of our code though
22:08 semiosis  pdurbin: it's proprietary code.  maybe we could do a google hangout or something, but it's not on a public repo
22:07 pdurbin   semiosis: you'll have to link me to a class on github you think benefits well from lombok, please
22:05 semiosis  value objects with literally no methods (they're all generated at compile time)
22:05 semiosis  lombok really shines in our apps at work.  we have db entities with almost no code, just annotations.
22:04 * pdurbin will have to check out lombok
22:01 semiosis  just to clarify
22:00 semiosis  pdurbin: i use lombok in every java project i work on.  i only use hawtjni in https://github.com/semiosis/libgfapi-jni (which uses lombok as well)
22:00 semiosis  i use it in every java project i work on, yes
21:58 pdurbin   semiosis: and you're using it in your glusterfs java client thing?
21:51 sivoais   yeah... it would be neat to have a DSL to generate both
21:48 semiosis  however i think hawtjni has the edge in performance
21:48 semiosis  yep, i probably would have used jna instead of jni if i were on my own, but I made friends with the developer of hawtjni and he helped me get it all set up
21:47 sivoais   I think Project Panama will be using an FFI (like Java Native Access)
21:47 sivoais   semiosis: that's cool! Just C though... C++ is a weird beast *shudders*
21:24 semiosis  http://projectlombok.org/features/Data.html
21:23 semiosis  oh and logging too :)
21:22 semiosis  getters, setters, equals, hashcode, tostring, constructors, all taken care of by lombok, except where there's business logic
21:22 semiosis  our code has practically no boilerplate
21:21 semiosis  imagine
21:21 semiosis  cant imaging going back to java without it
21:21 pdurbin   semiosis: you use lombok? and like it?
20:27 semiosis  it gets a little more complicated when you want to pass structs around, but it's still all java code
20:26 semiosis  that's pretty much how you use hawtjni
20:26 semiosis  this page, http://www.infoworld.com/article/2607953/java/article.html, says "You just specify what library to load, what function you want to access, and its parameters and return value and the API should do the rest."
20:25 semiosis  https://github.com/fusesource/hawtjni
20:24 semiosis  sivoais: i use hawtjni for the glusterfs c library binding.  you might find that interesting
17:58 sivoais   Generic specialisation and better C API calls are what are the most interesting out of those for me
17:55 sivoais   all of these look really cool <http://www.infoworld.com/article/2895889/java/oracle-java-9-future-lego-like.html>
16:18 semiosis  pdurbin: also, lombok! http://projectlombok.org/
16:16 semiosis  they gave me an open source license for the glusterfs java filesystem project :-D
16:15 semiosis  i'm sure you can get a free (beer) license either for DVN as an open source project or through Harvard
16:14 semiosis  pdurbin: jetbrains all the way.  webstorm for html/js, or intelij idea for everything.
14:53 pdurbin   hopefully we'll find someone to contribute sinatra and dancer examples
14:52 pdurbin   codex++
14:52 pdurbin   thanks for the links at least :)
14:52 pdurbin   ok
14:49 codex     to me a pull request has only one use case - when someone can better your software by at least 10% or more, and is willing to take ownership of that part of the code for some period of time
14:48 codex     :)
14:48 codex     yea, i don't subscribe to that whole thing
14:43 pdurbin   I'm looking for pull requests. :)
14:43 pdurbin   codex-- ;)
14:32 codex     http://www.perldancer.org/example
14:31 codex     pdurbin: https://www.digitalocean.com/community/tutorials/how-to-install-and-get-started-with-sinatra-on-your-system-or-vps
13:36 pdurbin   prologic++++
13:35 prologic  pdurbin: I’m halg way done :)
03:39 * aditsu  made a language called CJam :p
01:08 pdurbin   :)
00:50 sivoais   "The most important thing in the programming language is the name. A language will not succeed without a good name. I have recently invented a very good name and now I am looking for a suitable language." -- Donald Knuth
00:49 sivoais   re: names of programming languages (can't find a proper source on this quote...):