japhb jnthn++ # Talk slides from brno 02:08
.ask jnthn Is there a version of your Brno talk slides with the speaker notes? Several slides ask questions that you seem to have answered and/or discussed live.
yoleaux japhb: I'll pass your message to jnthn.
Geth rakudo: 5feb6bbec3 | (Zoffix Znet)++ | docs/archive/2018-03-04--Polishing-Rationals.md
Update "Polishing Rationals" proposal to rev2

  - Change plan to create FatRats in cases where we currently make
   broken Rats to instead make fatty RatStrs
  - A fatty RatStr type-wise is indistinguishable from a regular
   RatStr, but internally has the ability to offer FatRat-like
   precision in contexts where FatRat can be used—basically it
   functions as a non-infectious FatRat.
  (N.B.: still need to do a proof-of-concept trial impl of this idea)
03:54
[Tux] Rakudo version 2018.02.1-120-g5feb6bbec - MoarVM version 2018.02-27-g760b0913e
csv-ip5xs0.845 - 0.849
csv-ip5xs-207.782 - 8.180
csv-parser37.431 - 39.350
csv-test-xs-200.504 - 0.510
test9.323 - 9.849
test-t2.668 - 2.728
test-t --race1.111 - 1.294
test-t-2049.440 - 49.791
test-t-20 --race15.900 - 20.488
08:46
lizmat Files=1236, Tests=76222, 313 wallclock secs (14.83 usr 5.30 sys + 2144.08 cusr 216.84 csys = 2381.05 CPU) 08:47
[Tux]: "test-t-20 --race 15.900 - 20.488" seems to be a very wide range 08:48
[Tux] running it again now. I saw that too 08:49
and I restarted my browser after clearing the cache
still slow though
and my box isn't very busy 08:50
[Tux] stashed a special edition lagavullin for tonight
Rakudo version 2018.02.1-120-g5feb6bbec - MoarVM version 2018.02-27-g760b0913e
csv-ip5xs0.820 - 0.828
csv-ip5xs-207.645 - 7.683
csv-parser36.893 - 37.286
csv-test-xs-200.415 - 0.459
test9.054 - 9.250
test-t2.529 - 2.629
test-t --race1.056 - 1.180
test-t-2045.923 - 46.238
test-t-20 --race15.789 - 16.230
08:56
[Tux] ⇒ $work
lizmat that looks like more familiar values 08:57
Geth rakudo: de30c16218 | (Elizabeth Mattijsen)++ | src/core/set_elem.pm6
Make Int:D (elem) Range:D be independent of size of Range

  - if the Range is an .is-int
  - also fixes GH #1593, because it doesn't vivify the entire (massive) Range
12:17
synopsebot RAKUDO#1593 [open]: github.com/rakudo/rakudo/issues/1593 LTA (elem) doesn't special case Range
lizmat afk again& 12:19
AlexDaniel lol I guess someone is having fun reproducing the bug? XD 17:26
… the issue disappears if I remove this line: my $host-arch = $*KERNEL.hardware; 18:27
which maybe makes sense because it does qqx under the hood
( github.com/perl6/whateverable/issues/296 )
ooooooooooooooooooooooooo 18:40
reproduced locally! YEEEEEEEEEEEeeeeeeeesssssssssssssssss!
Can anybody confirm that R#1595 is reproducible? 18:48
synopsebot R#1595 [open]: github.com/rakudo/rakudo/issues/1595 [weird] Some bug with a combination of $*KERNEL.hardware, slurp, run and Cro
dogbert17 ah 'frustratable' is that you latest superbot :-) 18:50
*your 18:51
AlexDaniel shareable6 is frustratable
shareable6: HEAD
shareable6 AlexDaniel, whateverable.6lang.org/HEAD
timotimo i get loads and loads of enters and afters
i.e. it keeps working 18:52
AlexDaniel ↑ you go to that link, cancel the download, and the whole refuses to work after that. Can't get any more frustrating for me :(
bot*
timotimo: :'(
timotimo: is that rakudo HEAD?
timotimo yeah 18:53
and probably some non-release commit of cro
... a crommit
AlexDaniel timotimo: well, that's really sad. Thanks. 18:54
I see the same behavior on the server and on my machine
↑ you see, someone followed my advice xD 18:56
dogbert17 are you on Cro 0.7.3 ? 18:59
AlexDaniel yes
dogbert17 it works for me unless I press 'Cancel' just after refresh, then I get 'broken pipe' 19:04
Geth rakudo: db010b847a | (Timo Paulssen)++ | src/core/signals.pm6
prepare for cancellable Signal supply

when the next moarvm version bump happens, closing all taps belonging to one signal will remove the underlying signal handler and restore it back to normal.
This way you can temporarily handle Ctrl-C and future Ctrl-C hits will terminate the program.
19:05
dogbert17 but that was under valgrind so I might not count :-)
AlexDaniel ↑ lol someone is having a lot of fun xD 19:06
dogbert17 AlexDaniel: your code works for me as well, same results as timotimo 19:09
timotimo hrmpf. i'd really like people to react to my pull request rfc for udp sockets 20:01
Geth nqp: eacd7346eb | (Zoffix Znet)++ | tools/build/MOAR_REVISION
[MoarVM Bump] Brings 2 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...9-g9c4b389 9c4b389 remove stray debug output 723fc75 make signal handlers cancellable
20:21
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...9-g9c4b389
rakudo: a31579c7ff | (Zoffix Znet)++ | tools/build/NQP_REVISION
[NQP Bump] eacd734 [MoarVM Bump] Brings 2 commits

NQP bump brought: github.com/perl6/nqp/compare/2018....3-geacd734
MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...9-g9c4b389 9c4b389 remove stray debug output 723fc75 make signal handlers cancellable
20:22
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....3-geacd734
8d954027fd | usev6++ | src/core/ThreadPoolScheduler.pm6

The array @last-utils isn't native on the JVM (cmp. commit 581edd58e9), so we can't use nqp::shift_n and nqp::push_n here.
travis-ci NQP build passed. Zoffix Znet '[MoarVM Bump] Brings 2 commits 20:39
travis-ci.org/perl6/nqp/builds/350000931 github.com/perl6/nqp/compare/90d98...cd7346ebfe
Zoffix Thinking of an article I'll write for Rationals eventually and RatStr being an allomorph for Rat, FatRat, and Str just doesn't jive well. Feels shoe-hornish. Like, if I want an allomorph of a Rat and RatStr, why am I forced to lug around a multi-byte string version of it I'll likely never use. 21:21
Gonna do a trial impl and see how insane a new type (and its Str allomorph) make the operator dispatch be 21:22
A non-infectious FatRat (AlloRat? MuskRat? MidRat?) 21:23
yet that feels overkillish
.oO( make Str part of RatStr optional if FatRat part is given )
21:24
But that still leaves a wart that if I type enough digits after the decimal point I may get a Str object (its subclass of course, but still something that gets accepted in Str recepticles) 21:26
timotimo: the PRs been there for like a month amd was mentioned in a weekly, so looks like no one has any objections to merging it :) 21:29
(err, above I meant an allomorph of Rat and FatRat, in the "lug Str part") 21:30
timotimo i'll make the changes i alluded to in the description and merge it 21:31
Geth roast: f6bf6da215 | usev6++ | S32-io/pipe.t
[JVM] Unfudge some now passing tests
21:42
dogbert17 to me it looks as if RT #128895 has been fixed, is there something that I'm missing here? 22:15
synopsebot RT#128895 [open]: rt.perl.org/Ticket/Display.html?id=128895 [BUG] $*ARGFILES.words does not fallback to $*IN
Zoffix dogbert17: yeah, it was fixed by IO Grant and IO::CatHandle has full test coverage 22:25
dogbert17 Zoffix++, do you want me to close it? 22:27
Zoffix Sure, dogbert17++
dogbert17 so the comment from liz that your fixes were reverted are bogus ... 22:30
Zoffix dogbert17: the comment is from 2016, before the IO grant, and references the original, 2015 incantation of IO::CatHandle 22:32
dogbert17 Zoffix: cool, thx for the explanation
and another RT is closed, yay 22:34
Zoffix \o/ 22:35
Geth nqp: c855b116e6 | (Tom Browder)++ (committed using GitHub Web editor) | docs/built-ins.md
remove space for consistent format
22:38
travis-ci Rakudo build failed. usev6 '[JVM] Extend bandaid for supervisor 22:51
travis-ci.org/rakudo/rakudo/builds/350025774 github.com/rakudo/rakudo/compare/a...954027fd11