Perl 6 - the future is here, just unevenly distributed

IRC log for #perl11, 2014-10-21

| Channels | #perl11 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:41 travis-ci RPerl build passed. Will Braswell says 'Grammar, Create Number Test File'
02:41 travis-ci http://travis-ci.org/wbraswell/rperl/builds/38563317 https://github.com/wbraswell/rperl/compare/0f1cd7b5454f...e7821b01f3dc
04:11 Bender joined #perl11
04:11 renormalist joined #perl11
04:11 dalek joined #perl11
04:11 sivoais joined #perl11
04:11 willthechill joined #perl11
04:11 davido_home joined #perl11
04:11 bulk88 joined #perl11
04:11 ingy joined #perl11
04:12 McFist joined #perl11
06:44 davido__ joined #perl11
06:45 rurban joined #perl11
07:16 basiliscos joined #perl11
07:54 davido__ joined #perl11
09:57 basiliscos joined #perl11
10:13 stelf|pc joined #perl11
10:13 stelf|pc hi all. just cloned the p2 git to a centos 6.5 and tried to build it.
10:13 stelf|pc got this Makefile error
10:14 stelf|pc in the libuv build
10:14 stelf|pc ./configure: line 12279: syntax error near unexpected token `;'
10:14 stelf|pc make: *** [3rd/libuv/Makefile] Error 2
10:14 willthechill rurban: somebody needs p2 help
10:14 stelf|pc Makefiles are really scarry, though I'd like to see if I can use p2 for some smaller projects..
10:18 willthechill stelf|pc: sorry I can't help, you'll have to wait for rurban
10:19 stelf|pc okay. i'm really eager to see all the goodness in practice
10:19 stelf|pc if i get disconnected, i'll come back and ask again.
10:20 rurban can you do an autoreconf in libuv?
10:20 basiliscos joined #perl11
10:20 rurban building p2 is not really complicated. you can do most things manually
10:22 stelf|pc okay. done.
10:22 stelf|pc and now I start the p2 build over?
10:22 rurban centos 6.5 is really old. I think I have it also somewhere
10:23 stelf|pc it is old, but it is what the admin keeps running on this dev box that I can mess with
10:23 rurban autoreconf; ./configure --disable-dtrace --enable-shared --prefix=`pwd`
10:23 rurban make
10:24 stelf|pc configure fails with the same error
10:24 stelf|pc ./configure: line 12279: syntax error near unexpected token `;'
10:24 rurban and then back to p2 and make
10:24 rurban that's a bad problem upstream with libuv then.
10:25 rurban does centos have libuv in epel?
10:25 stelf|pc i don't think so....
10:25 stelf|pc i'll try to fix the configure
10:25 rurban I've got a centos 5.11 here
10:27 stelf|pc hmm.. removed all ; before then statements in the configure
10:28 stelf|pc and it builds now.
10:28 stelf|pc is probably some kind of autoconf/configure version mismatch
10:29 stelf|pc make test passes
10:29 stelf|pc good!
10:29 rurban I got: lib/potion/aio.so: undefined symbol: uv_fs_event_init
10:34 rurban You are from Sofia?
10:34 stelf|pc yes.
10:35 rurban Great :)
10:35 rurban but p2 is still missing a lot. you cannot use it yet. potion is usable though
10:36 stelf|pc it is nice and warm outside, quite nice for a late october. we call it 'a gipsy summer'
10:36 stelf|pc btw, is p2 supposed to be able to run the examples/*.pl stuff ?
10:37 rurban Don't think so. Maybe some
10:37 rurban This is there for benchmarking against perl5
10:37 stelf|pc oh. ok.
10:38 rurban the .p2 samples should work
10:38 rurban fib
10:38 rurban and only 2 files from the roast5 core testsuite
10:41 rurban I'm still extending the potion vm a bit, not much time for p2 syntax and libs yet
10:41 stelf|pc so the pn files are reference and not supposed to run..?
10:43 rurban the .pn are for bin/potion
10:43 rurban different syntax, same vm
10:44 travis-ci RPerl build passed. Will Braswell says 'Grammar, Create Number Tests'
10:44 travis-ci http://travis-ci.org/wbraswell/rperl/builds/38590354 https://github.com/wbraswell/rperl/compare/e7821b01f3dc...30042afd00ef
10:44 rurban like with parrot. p2 can run many languages. we just need a syntax file for each language. perl6 and ruby e.g.
10:45 rurban we even have ruby and perl6 syntax already. need to merge it soon
10:48 stelf|pc okay. so, no perl5 grammer yet, but the vm is good to run most tasks
10:50 rurban the p5 gramamr is there. just some pieces are missing, like heredoc
10:50 rurban but mostly simple library functions, like chdir and such
10:51 rurban chdir would be a method on a string
10:51 stelf|pc the pn tests produce some quite impressive times
10:52 rurban nbody is the slowest so far. array access
10:52 stelf|pc it is very exciting to expect a stable p5 grammar. even though most XS stuff would not run, a lot of modules have pureperl implementations good for large set of jobs
10:52 rurban almost as slow as perl5 (or the old clisp for which test was written)
10:53 stelf|pc to be honest, i found perl11 just after reading one more of these 'perl is dead' articles. got me really pissed.
10:53 rurban the ffi should be better than XS
10:53 rurban extern …
10:54 rurban I just need to add ffi argument translation to the compiler. this needs to be fast
10:55 rurban but perl is dead :)
10:56 stelf|pc more like undead
10:57 rurban I'll need to fork it to keep it on life-support somewhen next year eventually
10:59 willthechill is perl dead?  no, because perl11 is alive
10:59 rurban well, perl11 is not good enough yet. still in early development
11:00 willthechill not YET
11:00 willthechill :)
11:00 stelf|pc imho... perl (for years) is not p5 only. with all the vms and language versions its more of a realm of its own
11:02 stelf|pc people find a way to optimize/hack the concept in so many different aspects, that the question is if this quantum-parallel dev is ever going to be merged in a single reference version
11:02 willthechill the future of perl is unclear
11:02 willthechill too many paths to know which will win just yet
11:03 willthechill http://rperl.org/the_book_of_rperl.html
11:03 rurban look at gperl also. it misses eval but is also pretty impressive
11:03 rurban but for gperl you'd need llvm, and that isn;t on centos 6.5 I fear
11:12 stelf|pc rperl is like perlcc but improved in a way, is it ?
11:13 willthechill rperl is for low magic, perlcc is for high magic
11:13 willthechill rperl will give very high optimization for low magic perl 5 code
11:13 willthechill perlcc using B::C will give no optimization for high magic perl 5 code
11:14 stelf|pc should I try it with project-with-lots-of-DBIC-IOASYNC-and-Moose stuff ?
11:14 willthechill perlcc using B::CC will give low optimization for high magic perl 5 code, but is not complete
11:18 stelf|pc as far as I can tell from http://rperl.org/the_low_magic_perl_commandments.html large-scale projects that tend to use 'half of CPAN' may not be greatly optimized...
11:20 willthechill RPerl v1.0 is not meant to compile CPAN modules
11:20 willthechill future versions of RPerl will add support for medium and high magic perl 5 code, such as is found on CPAN
11:49 travis-ci RPerl build passed. Will Braswell says 'Grammar, Create Number Tests, Part 2'
11:49 travis-ci http://travis-ci.org/wbraswell/rperl/builds/38595081 https://github.com/wbraswell/rperl/compare/30042afd00ef...13a5f30b75e0
11:49 willthechill yay!
12:03 rurban joined #perl11
12:37 stelf|pc joined #perl11
16:40 basiliscos joined #perl11
19:37 willthechill joined #perl11
23:51 basiliscos joined #perl11

| Channels | #perl11 index | Today | | Search | Google Search | Plain-Text | summary