Perl 6 - the future is here, just unevenly distributed

IRC log for #pdl, 2014-08-16

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

All times shown according to UTC.

Time Nick Message
00:28 jberger yay, my only real contribution to pdl! (unpdl)
00:30 mohawk jberger, in alien-base it looks like there are some PRs and whatnot from march?
00:51 jberger yeah, I have been shirking responsibility
00:51 jberger however, I have actually shirked some responsibility
00:51 jberger I started a github org to handle the further development
00:52 jberger I have two problems, lack of primary motivation (I no longer need my library) and lack of deeper knowledge of the subject (I knew just enough to do what I needed)
00:52 jberger so if there are people who want to join and make improvements, just yell
00:52 jberger I did rope chm in
00:55 mohawk jberger, i'm interested
00:56 sivoais jberger: just put in a PR for Alien::Base. Doing my part! :-)
00:56 mohawk CPAN day-tastic!
00:56 sivoais \o/
00:56 jberger \o/
00:56 sivoais now on to your PRs, mohawk :-)
00:57 mohawk grin
00:57 mohawk is Perl::OSType in core?
00:57 mohawk ah, since 5.13.9
00:57 jberger I have been more interested in fixing this lately: http://www.techempower.com/benchmarks/#sectio​n=data-r9&hw=peak&test=json&l=pa8
00:57 sivoais > Perl::OSType was first released with perl v5.13.9
00:58 sivoais according to corelist
00:58 mohawk ha
00:58 mohawk too slow ;-)
00:58 mohawk does a::b have a min perl version?
00:58 sivoais in META.yml:  perl: v5.8.1
01:00 sivoais jberger: woah. So that's saying that none of the Perl frameworks even ran that benchmark?
01:02 jberger they all failed to run
01:02 jberger :'(
01:03 jberger I have been working with those guys to ensure that the tests at least run in round 10
01:03 jberger then I need to rope in each framework community to make sure their apps pass the tests and (hopefully) perform well
01:04 sivoais ah, like <https://github.com/TechEmpower​/FrameworkBenchmarks/pull/986>
01:04 sivoais nice
01:05 jberger yuppers
01:05 jberger I'm working on using Carton now
01:05 mohawk sivoais, would you be interested in working on perl5-a-b?
01:05 jberger a little bit just to show off
01:05 mohawk i need to get to grips with "modern perl"
01:05 jberger showing off how good the deploy chain can be relative to other languages would be fun
01:06 sivoais That's very odd that they can't get them to run. I always have no issue at all running perlbrew on a clean VM install.
01:06 sivoais mohawk: sure. What do you have planned?
01:06 mohawk ask jberger ;-)
01:07 mohawk he's implicitly asking for volunteers
01:07 jberger explicitly!
01:07 sivoais OK, I'll explicitly say: I volunteer to help Alien::Base. :-)
01:07 jberger great!
01:08 jberger we do have working group rules set out, but I'm ok nominating you
01:09 sivoais these <https://github.com/Perl5-A​lien/Perl5-Alien-Documents>?
01:13 mohawk jberger, i'm volunteering myself too if you'd like
01:20 jberger sivoais: https://github.com/Perl5-Alien/​Perl5-Alien-Documents/issues/2
01:20 jberger mohawk: I appreciate the offer, but I haven't seen any of your code yet
01:21 jberger what I really need is people to shepherd PRs
01:26 sivoais mmm, yes. I'm looking at some of the PRs there.
01:27 sivoais I think this one regarding adding the .dlls to $ENV{PATH} is correct, but I need to confirm <https://github.com/Perl5-Alien/Alien-Base/pull/32>. I'm working on getting libmagic working on Windows, so I'll be slowly tackling issues there.
01:31 jberger see, that's the kind of thing that I don't have the time or inclination to do anymore
01:32 jberger for that specific one, I had wanted to investigate DynaLoader solutions
01:32 sivoais yeah. It's a bit frustrating. But I'm doing it so that I can do cooler things later, heh
01:32 jberger I have been trying to let it do all of the per-platform work
01:33 sivoais yeah, that makes for a cleaner implementation. Plus, if Dynaloader needs patching, that's a win for everyone
01:33 jberger but, if I don't want to do all the work, I'm going to have to relax a bit I think
01:34 jberger my only real design goal is that if AB chooses to install a library (rather than use one that exists on the system) then it cannot affect anything outside of that perl interpreter
01:35 jberger and preferrably only if the Alien:: module is actually loaded
01:38 sivoais Makes sense. Otherwise you get odd bugs that are a pain to track down.
01:38 jberger worse, ab starts to need root (in some cases)
01:38 jberger it can change system behavior
01:39 jberger installing a cpan module should never break anything non-perl
01:52 mohawk jberger, take a look at EUMM's closed & merged PRs if you want to see my code
01:54 sivoais what about the changes in Inline / Inline::C! Lots of good stuff coming out there.
01:55 mohawk i haven't changed Inline in a few weeks
01:55 mohawk i'd forgotten about that ;-)
01:55 mohawk (this evening's Inline::MakeMaker change hardly counts)
01:58 mohawk 11 merged PRs in 18 days: https://github.com/Perl-Toolchain-Gang/ExtU​tils-MakeMaker/pulls?q=is%3Apr+is%3Aclosed
01:58 mohawk not that i'm big-timing or anything ;-)
02:02 jberger :o
02:04 mohawk speaking of shepherding, before i decided to get stuck in and fix it, EUMM had 15 or so branches with code just lying around
02:04 jberger mohawk: ok ok
02:04 jberger :-P
02:04 mohawk now there's 4 + master and i'm still on the case ;-)
02:05 mohawk nuff said
02:33 jberger mohawk: https://github.com/Perl5-Alien/​Perl5-Alien-Documents/issues/3
02:33 mohawk looking
02:33 mohawk oh, that's kind - thanks
02:33 jberger mohawk/ sivoais: you may each want to make a brief statement in your own support
02:33 mohawk ok
02:36 mohawk done
02:44 sivoais same
06:26 lungching joined #pdl
07:03 mohawk joined #pdl
16:13 mohawk joined #pdl
18:00 sivoais someone is asking about a "PDL for strings" at <https://www.linkedin.com/groupItem?view=&amp;item=​5904164747456049153&amp;type=member&amp;gid=106254>. Hopefully that link works. LinkedIn is weird when it comes to URLs.
18:11 sivoais I understand what they want, but that's not what PDL's for. Plus, you can't overload =~
18:11 sivoais but you can do what Regexp::Grammars does :-P
21:38 mohawk sivoais, that url goes to a login page for me, but meh
22:21 sivoais yeah :-/ I answered them.
22:22 sivoais btw, I've been experimenting at using PDL to store categorical variables (very similar to enums and R factors)
22:22 sivoais I'm using Moo to extend PDL.

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