Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2016-05-19

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:49 ilbot3 joined #perl6-dev
01:49 Topic for #perl6-dev is now Perl 6 language and compiler development 2.0 | Logs at http://irclog.perlgeek.de/perl6-dev/today
06:51 RabidGravy joined #perl6-dev
06:53 nine ugexe: my merge last Friday would be a candidate. Unfortunately I don't have Windows, so I can't debug.
08:07 sno joined #perl6-dev
11:21 brrt joined #perl6-dev
12:14 [Tux] jnthn, do you want me to create a new ticket for that utf8-c8 issue with examples?
12:15 timotimo we don't have one yet?
12:15 timotimo in any case, i've started looking at it, but i only managed to make it worse, not better :(
12:15 [Tux] there was one, but that had a wider issue range
12:17 [Tux] https://rt.perl.org/Ticket/Display.html?id=127878 that one was the original, but that has been fixed
12:18 [Tux] https://rt.perl.org/Ticket/Display.html?id=126756 <= that is the new one
12:18 [Tux] which has also been marked fixed
12:18 timotimo ah, yeah, we don't segv any more
12:19 timotimo but we do read past the end of the buffer sometimes, and that causes an extra 0 to appear at the end
12:21 timotimo i managed to make it duplicate the second-to-last byte and then have the last byte
12:21 timotimo and also to drop the last two bytes
12:22 [Tux] This is Rakudo version 2016.04-208-ge39ce3d built on MoarVM version 2016.04-134-g9879233
12:22 [Tux] test            20.691
12:22 [Tux] test-t          13.293
12:22 [Tux] csv-parser      35.050
12:23 timotimo :\
12:32 [Tux] I just filed https://rt.perl.org/Ticket/Display.html?id=128184
12:33 timotimo pairs are swapped o_O
12:33 timotimo teh f ?!?
12:45 brrt joined #perl6-dev
13:33 dalek nqp: 45e0bfb | (Pawel Murias)++ | src/vm/js/nqp-runtime/ (4 files):
13:33 dalek nqp: [js] Plugable and serializable container specs.
13:33 dalek nqp: review: https://github.com/perl6/nqp/commit/45e0bfb34d
13:42 lizmat joined #perl6-dev
13:51 mst joined #perl6-dev
13:57 pmurias joined #perl6-dev
13:57 dalek nqp: 9df5217 | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js:
13:57 dalek nqp: [js] Fix bug in NYI message.
13:57 dalek nqp: review: https://github.com/perl6/nqp/commit/9df5217d32
14:30 pmurias joined #perl6-dev
14:49 pmurias joined #perl6-dev
14:59 pmurias joined #perl6-dev
15:01 lizmat joined #perl6-dev
15:50 ugexe nine: on my windows dev machine i can't reproduce any significant speed difference between blead vs 2016.04, but on appveyor its about a 33% slow down for every command vs 2016.04. My only (wild) guess is now its hitting an io/mem/cpu limit or something
15:52 ugexe on an unrelated note, I always thought "blead" referred to "bleed" as in bleeding edge
15:59 lizmat m: 42.polymod(0)
15:59 camelia rakudo-moar e39ce3: OUTPUT«Attempt to return outside of any Routine␤  in block <unit> at /tmp/HRKsEl47P9 line 1␤␤»
15:59 lizmat that feels LTA
15:59 lizmat m: gather { fail }
15:59 camelia rakudo-moar e39ce3: OUTPUT«Failed␤  in block <unit> at /tmp/ZMEILAfFla line 1␤␤»
15:59 lizmat hmmm
16:12 lizmat m: class A { method a() { gather { fail } } }; A.a   # golfed the issue
16:12 camelia rakudo-moar e39ce3: OUTPUT«Attempt to return outside of any Routine␤  in block <unit> at /tmp/bdyOdNFdxV line 1␤␤»
16:13 jnthn fail is just a kind of return, and yeah, you can't return from a gather 'cus the routine already returned.
16:14 lizmat so that means you can't fail from a gather ?
16:15 lizmat m: class A { method a() { gather { Failure.new } } }; A.a   # golfed the issue
16:15 camelia rakudo-moar e39ce3: OUTPUT«Failed␤  in block <unit> at /tmp/IS6S256t22 line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/IS6S256t22 line 1␤␤»
16:21 jnthn lizmat: Correct, just as you can't from a map if it's being lazily evaluated
16:22 jnthn m: sub foo() { (1..5).map({ fail }) }; foo()
16:22 camelia rakudo-moar e39ce3: OUTPUT«Attempt to return outside of any Routine␤  in block <unit> at /tmp/NfVN4VlSp9 line 1␤␤»
16:22 jnthn Same situation there.
16:22 lizmat hmmm... so that means we'll get LTA error if a module is doing a fail, and it's being used inside a gather ??
16:22 jnthn Why'd that be a problem?
16:23 jnthn fail is a form of return
16:23 jnthn So it's all about the lexically enclosing routine
16:24 jnthn The reason the above ones don't work out is because you're no longer executing foo (or method a) at the point the fail executes
16:25 lizmat yeah, got that now  :-)
16:26 jnthn So if I get what you're asking: no, I don't think we'll have too many LTA situations. That error *would* be nicer if it could indicate that the routine was already returned from
16:33 lizmat ack
16:52 [Coke] grep!
16:56 [Coke] hey all you groovy people, don't forget to stay hydrated and have a good snack. :)
17:10 RabidGravy I'm contemplating beer, does that count?
17:11 geekosaur nope
17:13 [Coke] sounds good to me. :)
17:27 MadcapJake [Coke]: What would you consider to be the most atomic list of rakudobug categories? I'd like to provide templates for each in my issue submitter (that I hope to apply some tuits to)
17:29 [Coke] the categories are all ad hoc. I put the biggest ones I expect people to use in the rakudo README
17:54 lizmat joined #perl6-dev
17:59 lizmat I'm wondering whether "gather" and "fail" should know of each other, and basically have "fail" take the failure, rather than throwing it if called within a gather
18:28 tomboy65 joined #perl6-dev
19:29 masak lizmat: I thought `fail` was lexical-only, like `return`
20:22 lizmat joined #perl6-dev
20:40 brrt joined #perl6-dev
20:43 Ven_ joined #perl6-dev
20:43 sno joined #perl6-dev

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary