Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2017-08-31

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

All times shown according to UTC.

Time Nick Message
00:24 geekosaur joined #perl6-dev
01:14 AlexDaniel samcv: any news on RT #131961? Or if not, was there anything interesting you were able to find? I'm thinking maybe we can leave something in a comment on that ticket to help the thing go
01:14 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=131961
01:15 samcv well i am not sure what's causing it. though it now doesn't leak memory when it throws, and now frees the bytes in the buffer
01:15 samcv but otherwise no
01:26 AlexDaniel wait, I didn't bisect this ticket?
01:27 AlexDaniel .oO( why???? )
01:28 AlexDaniel makes no sense. I'll get some tea and will bisect it. Who knows, maybe it'll give a clue
01:32 geekosaur joined #perl6-dev
01:52 ilbot3 joined #perl6-dev
01:52 Topic for #perl6-dev is now Perl 6 language and compiler development 2.0 | Logs at http://irclog.perlgeek.de/perl6-dev/today | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
01:55 AlexDaniel bisected to https://github.com/rakudo/rakudo/commit/51b63bf9f0d0594687908f0756774f29cde39c1f
01:57 AlexDaniel but I don't know if it means anything
01:58 AlexDaniel but I was seeing something like this a few weeks before that commit: https://github.com/perl6/whateverable/issues/153
01:58 AlexDaniel it may be a different issue
01:58 AlexDaniel how can I test it? I guess simply try it with a bigger file?
02:06 AlexDaniel samcv: yes, that's the right commit. Before that it can go through 251MB file just fine
02:07 AlexDaniel … “commit”. One of 15 MoarVM commits I mean
02:10 AlexDaniel .tell [Coke] A *tiny* update on the Malformed UTF-8 issue. There's hope! https://rt.perl.org/Ticket/Display.html?id=131961#txn-1486102
02:10 yoleaux AlexDaniel: I'll pass your message to [Coke].
02:17 samcv AlexDaniel, well decodestream was changed
02:17 samcv it's possible that or the utf8 changes caused it
02:18 samcv i can't compile nqp right now though..
02:18 samcv anyone else have this?
02:21 geekosaur nqp built for me a couple hours ago
02:34 ugexe I wrote a quick example of how to get appveyor build status changes in an irc channel by slapping together Cro and IRC::Client - https://gist.github.com/ugexe/a7724886674ffef74f656f93f0b8a8cf
02:35 Zoffix wooo \o/ ugexe++
02:56 MasterDuke joined #perl6-dev
04:05 AlexDaniel alright, it's https://github.com/MoarVM/MoarVM/commit/a6abd3c6654413d2230470dbaa82b7b3a2b05762
04:06 AlexDaniel why can't I teach bisectable to do it? Why? :-/
04:08 samcv AlexDaniel++ should let jnthn know maybe he can figure it out better than I, since he wrote it
04:08 AlexDaniel .tell jnthn FWIW https://rt.perl.org/Ticket/Display.html?id=131961#txn-1486128
04:08 yoleaux AlexDaniel: I'll pass your message to jnthn.
04:28 Ven`` joined #perl6-dev
04:30 AlexDaniel .tell timotimo ping me when you have 15 minutes for me
04:30 yoleaux AlexDaniel: I'll pass your message to timotimo.
04:42 samcv that nqp commit that disabled repl on tty i think broke roast tests
04:42 samcv that wanted the output of the REPL
04:42 samcv maybe we need an ENV var to enable it or a flag
04:43 samcv like one to force repl and one to force no-repl
05:03 llfourn joined #perl6-dev
05:12 samcv .u { .uniprop('Block') eq 'Malayalam' }
05:12 yoleaux U+0020 SPACE [Zs] ( )
05:12 yoleaux U+0027 APOSTROPHE [Po] (')
05:12 yoleaux U+0028 LEFT PARENTHESIS [Ps] (()
05:14 samcv u: { .uniprop('Block') eq 'Malayalam' }
05:14 unicodable6 samcv, U+0D01 MALAYALAM SIGN CANDRABINDU [Mn] ( ?)
05:14 unicodable6 samcv, U+0D02 MALAYALAM SIGN ANUSVARA [Mc] (?)
05:14 unicodable6 samcv, 114 characters in total: https://gist.github.com/fa65e9e3f6629185e979d453f219912e
05:41 Geth ¦ roast: 640d40ee81 | (Samantha McVey)++ | S05-mass/properties-block.t
05:41 Geth ¦ roast: Fix InMalayam test
05:41 Geth ¦ roast:
05:41 Geth ¦ roast: Previously this test was a CJK base with a Malayalam extending character.
05:41 Geth ¦ roast: Replace it so the base character is also Malayam.
05:41 Geth ¦ roast: review: https://github.com/perl6/roast/commit/640d40ee81
06:05 [Tux] This is Rakudo version 2017.08-25-g7d1ece805 built on MoarVM version 2017.08.1-32-gcd41322e
06:05 [Tux] csv-ip5xs        1.324 - 1.352
06:05 [Tux] test            10.438
06:05 [Tux] test-t           3.513 - 3.640
06:05 [Tux] csv-parser      11.803
06:16 brrt joined #perl6-dev
06:39 Geth ¦ nqp: fc57c9a203 | (Samantha McVey)++ | tools/build/MOAR_REVISION
06:39 Geth ¦ nqp: Bump MoarVM for Unicode 10 update
06:39 Geth ¦ nqp:
06:39 Geth ¦ nqp: * Adds Hangul codepoint names
06:39 Geth ¦ nqp: * Property values are now unique for each property
06:39 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/fc57c9a203
06:39 Geth ¦ nqp: version bump brought these changes: https://github.com/MoarVM/MoarVM/compare/2017.08.1-44-g7d5cd52...2017.08.1-46-g79654248
06:43 Geth ¦ rakudo/nom: 64dd94c250 | (Samantha McVey)++ | tools/build/NQP_REVISION
06:43 Geth ¦ rakudo/nom: Bump MoarVM/nqp for Unicode 10 update
06:43 Geth ¦ rakudo/nom:
06:43 Geth ¦ rakudo/nom: * Adds Hangul codepoint names
06:43 Geth ¦ rakudo/nom: * Property values are now unique for each property
06:43 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/64dd94c250
06:43 Geth ¦ rakudo/nom: version bump brought these changes: https://github.com/perl6/nqp/compare/2017.08-24-gd01668b...2017.08-25-gfc57c9a20
07:24 lizmat joined #perl6-dev
07:40 brrt joined #perl6-dev
08:00 lizmat Files=1223, Tests=67663, 293 wallclock secs (13.91 usr  5.21 sys + 1972.40 cusr 206.74 csys = 2198.26 CPU)
08:03 [TuxCM] joined #perl6-dev
08:19 dogbert2 joined #perl6-dev
08:22 Geth ¦ nqp: 589a3dd5cd | (Samantha McVey)++ | docs/ops.markdown
08:22 Geth ¦ nqp: Add docs for a few variations of index/eqat
08:22 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/589a3dd5cd
08:37 pmurias joined #perl6-dev
08:39 Geth ¦ nqp/master: 4 commits pushed by pmurias++
08:39 Geth ¦ nqp/master: 10dc2d82f9 | [js] Fix lexical_handler_not_found
08:39 Geth ¦ nqp/master: c14acf8950 | [js] Pass the HLL object to nqp.arg
08:39 Geth ¦ nqp/master: e1d0c92702 | [js] Refactor argument handling
08:39 Geth ¦ nqp/master: 7729dadb0c | [js] Pass the HLL object around instead of just the name
08:39 Geth ¦ nqp/master: review: https://github.com/perl6/nqp/compare/589a3dd5cd...7729dadb0c
09:16 Ven`` joined #perl6-dev
09:22 lizmat afk for a while&
09:33 timotimo AlexDaniel: i think i might have the time for you now. what's up?
09:33 yoleaux 04:30Z <AlexDaniel> timotimo: ping me when you have 15 minutes for me
09:41 timotimo correction: i have to head out and do some errands
10:14 Geth ¦ rakudo/jit_nativecall: 3a48c66de1 | (Stefan Seifert)++ | lib/NativeCall.pm6
10:14 Geth ¦ rakudo/jit_nativecall: Unbox HLL objects passed to JITed native functions
10:14 Geth ¦ rakudo/jit_nativecall: review: https://github.com/rakudo/rakudo/commit/3a48c66de1
10:16 dogbert17 timotimo: https://irclog.perlgeek.de/moarvm/2017-08-30#i_15092872
10:17 timotimo dogbert17: i believe we're on a non-release commit of libtommath that already has that
10:24 dogbert17 yes but the fix for RT #129829 would be nice to have as well
10:24 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=129829
10:33 pmurias joined #perl6-dev
11:37 Skarsnik joined #perl6-dev
11:38 Skarsnik Hello
11:47 timotimo AlexDaniel: i have time for you, are you there?
12:14 AlexDani` joined #perl6-dev
12:43 AlexDaniel timotimo: now I am :)
12:43 timotimo OK!
12:46 Skarsnik I just watched Damian Conway talk about three little words. I love the your keyword x)
12:53 huggable joined #perl6-dev
12:56 Zoffix ??? New blog post: "On Troll Hugging, Hole Digging, and Improving Open Source Communities": https://rakudo.party/post/On-Troll-Hugging-Hole-Digging-and-Improving-Open-Source-Communities
12:57 pmurias Zoffix: hole digging seems like a more "permanent" way to get rid of trolls ;)
12:58 Zoffix :)
13:17 skids joined #perl6-dev
13:21 AlexDaniel awesome blog post
13:23 MasterDuke Zoffix++ nice post. fyi, "barelly"
13:26 Zoffix Thanks. Fixed.
13:27 AlexDaniel huggable: ? :is: https://rakudo.party/post/On-Troll-Hugging-Hole-Digging-and-Improving-Open-Source-Communities
13:27 huggable AlexDaniel, Added ? as https://rakudo.party/post/On-Troll-Hugging-Hole-Digging-and-Improving-Open-Source-Communities
13:27 AlexDaniel huggable: <3 :is: https://rakudo.party/post/On-Troll-Hugging-Hole-Digging-and-Improving-Open-Source-Communities
13:27 huggable AlexDaniel, Added <3 as https://rakudo.party/post/On-Troll-Hugging-Hole-Digging-and-Improving-Open-Source-Communities
13:28 pmurias joined #perl6-dev
13:30 lizmat joined #perl6-dev
13:46 |Tux| joined #perl6-dev
13:55 nine OMG there's a quote of me
13:55 * nine blushes
13:58 AlexDaniel by the way, maybe somebody has some interesting ideas on this: https://github.com/perl6/whateverable/issues/40#issuecomment-326300469
13:58 AlexDaniel mainly the question is: if nqp bump in rakudo brought more than one nqp commit, what's the right way to bisect it?
13:59 |Tux| joined #perl6-dev
14:00 |Tux| joined #perl6-dev
14:02 nine Zoffix++ # really awesome blog post
14:02 lizmat someone should put Zoffix' blog post on HN
14:10 brrt someone with a flame-retardant suit
14:10 brrt or a 'thick skin' or what have you
14:15 timotimo i just realized i never told anyone my lightning talk would be okay to record & upload
14:15 timotimo so it probably won't appear on youtube
14:15 timotimo which is a good thing because i could re-make it as a much more structured (and rehearsed) screencast or something
14:18 brrt (also, Zoffix++, i enjoyed that post)
14:21 timotimo yes, yes. good post, Zoffix++
14:21 MasterDuke are the SPW videos going up under a particular account?
14:22 perlpilot Zoffix++  I haven't finished reading it yet, but it resonates well with me.  Good job!
14:22 timotimo MasterDuke: lee manages that account
14:30 MasterDuke timotimo: k. hopefully some links get posted, i'd like to watch them
14:33 timotimo https://www.youtube.com/channel/UCV3oXF6JSn-WwIDa_7KMuEA
14:33 Geth ¦ rakudo: MasterDuke17++ created pull request #1145: Simplify ignore(mark|case) handling in INTERPOLATE
14:33 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1145
14:33 timotimo not sure if the comment about lee managing it is correct
14:33 timotimo lee was the one with the camera in any case
14:33 MasterDuke timotimo++ thanks
14:42 AlexDaniel ? Community Bug SQUASHathon http://blogs.perl.org/users/zoffix_znet/2017/08/youre-invited-community-bug-squashathon.html
14:55 Geth ¦ rakudo/nom: 79604a8803 | (Elizabeth Mattijsen)++ | src/core/Junction.pm
14:55 Geth ¦ rakudo/nom: Refine Junction.INFIX-TWO
14:55 Geth ¦ rakudo/nom:
14:55 Geth ¦ rakudo/nom: one() infix one() shouldn't merge, because any merge would change
14:55 Geth ¦ rakudo/nom: the semantics.  This *is* ok for any(), all() and none().
14:55 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/79604a8803
14:55 brrt joined #perl6-dev
15:05 Geth ¦ rakudo/nom: 8b5d283cb6 | (Elizabeth Mattijsen)++ | src/core/Junction.pm
15:05 Geth ¦ rakudo/nom: Complete support for Junction ~ Junction
15:05 Geth ¦ rakudo/nom:
15:05 Geth ¦ rakudo/nom: Even for two differently typed junctions or if one of the junctions
15:05 Geth ¦ rakudo/nom: is a one().
15:05 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8b5d283cb6
15:08 Geth ¦ rakudo/nom: 215a5fa731 | MasterDuke17++ | src/core/Match.pm
15:08 Geth ¦ rakudo/nom: Simplify ignore(mark|case) handling in INTERPOLATE
15:08 Geth ¦ rakudo/nom:
15:08 Geth ¦ rakudo/nom: By switching to use the relatively new nqp::index(ic|im|icim) ops.
15:08 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/215a5fa731
15:08 Geth ¦ rakudo/nom: 067f5407ef | lizmat++ (committed using GitHub Web editor) | src/core/Match.pm
15:08 Geth ¦ rakudo/nom: Merge pull request #1145 from MasterDuke17/simplify_ignoremark_and_ignorecase_in_INTERPOLATE
15:08 Geth ¦ rakudo/nom:
15:08 Geth ¦ rakudo/nom: Simplify ignore(mark|case) handling in INTERPOLATE
15:08 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/067f5407ef
15:14 timotimo .o( squashable )
15:20 ilmari joined #perl6-dev
15:22 sivoais joined #perl6-dev
15:26 brrt joined #perl6-dev
15:48 Geth ¦ rakudo: MasterDuke17++ created pull request #1146: Document the $m argument in INTERPOLATE's comment
15:48 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1146
15:58 |Tux| .hug Zoffix
16:05 Ven`` joined #perl6-dev
16:14 Zoffix |Tux|: oops. I guess IRC::Client still has a bug where the IRC grammar doesn't match nicks with `|` chars in them
16:14 |Tux| (hug still applies)
16:14 Zoffix Thanks :)
16:19 Geth ¦ rakudo/nom: ea57cbec8a | MasterDuke17++ (committed using GitHub Web editor) | src/core/Match.pm
16:19 Geth ¦ rakudo/nom: Document the $m argument in INTERPOLATE's comment
16:19 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/ea57cbec8a
16:19 Geth ¦ rakudo/nom: e717d14db1 | (Zoffix Znet)++ (committed using GitHub Web editor) | src/core/Match.pm
16:19 Geth ¦ rakudo/nom: Merge pull request #1146 from MasterDuke17/patch-1
16:19 Geth ¦ rakudo/nom:
16:19 Geth ¦ rakudo/nom: Document the $m argument in INTERPOLATE's comment
16:19 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e717d14db1
16:35 vendethiel- joined #perl6-dev
17:14 Skarsnik_ joined #perl6-dev
17:15 Skarsnik__ joined #perl6-dev
17:40 dogbert2 joined #perl6-dev
17:59 lizmat_ joined #perl6-dev
18:37 robertle joined #perl6-dev
19:08 Ven joined #perl6-dev
19:50 Geth ¦ rakudo: MasterDuke17++ created pull request #1147: s/index/eqat/g in INTERPOLATE literal handling
19:50 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1147
20:07 Geth ¦ rakudo/nom: c6aacafd1f | MasterDuke17++ | src/core/Match.pm
20:07 Geth ¦ rakudo/nom: s/index/eqat/g in INTERPOLATE literal handling
20:07 Geth ¦ rakudo/nom:
20:07 Geth ¦ rakudo/nom: The simplification in 215a5fa731 used nqp::index* for the ignore* cases,
20:07 Geth ¦ rakudo/nom: but if that change is made to the regular case some spectests break.
20:07 Geth ¦ rakudo/nom: Both 215a5fa731 and this change pass all spectests, which I think
20:07 Geth ¦ rakudo/nom: indicates that there probably aren't enough (any?) tests for
20:07 Geth ¦ rakudo/nom: regex interpolation with ingnore* modifiers.
20:07 Geth ¦ rakudo/nom:
20:07 Geth ¦ rakudo/nom: Using index* is more efficient, since it doesn't require walking
20:07 Geth ¦ rakudo/nom: through the haystack and checking at each position, but it somehow needs
20:07 Geth ¦ rakudo/nom: to be done more correctly.
20:07 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c6aacafd1f
20:07 Geth ¦ rakudo/nom: d135728b71 | lizmat++ (committed using GitHub Web editor) | src/core/Match.pm
20:07 Geth ¦ rakudo/nom: Merge pull request #1147 from MasterDuke17/use_eqat_instead_of_index_in_INTERPOLATE
20:07 Geth ¦ rakudo/nom:
20:07 Geth ¦ rakudo/nom: s/index/eqat/g in INTERPOLATE literal handling
20:07 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d135728b71
20:18 MasterDuke lizmat: thanks. btw, istr you were working on MATCH a little while ago. do you have any thoughts on my recent comments/questions re INTERPOLATE?
20:18 lizmat none, apart from being glad you're working on it :-)
20:19 lizmat MasterDuke: and that we could use some more tests :-)
20:20 MasterDuke heh, ain't that always the case...
20:20 lizmat indeed  :-)
20:23 MasterDuke well, i'm certainly open to help/advice/correction/etc from anybody, i'd be lying if i said i completely understand how INTERPOLATE is working (or should)
21:27 japhb huggable: speed
21:27 huggable japhb, http://tux.nl/Talks/CSV6/speed4.html
21:27 japhb buggable: speed
21:27 buggable japhb, ?????????????????????????????????????????????????? data for 2017-07-30–2017-08-31; range: 3.471s–4.925s; 21% faster
21:29 AlexDaniel this graph needs a logarithmic scale
21:30 AlexDaniel maybe
21:33 samcv left #perl6-dev
21:33 samcv joined #perl6-dev
21:43 cognominal joined #perl6-dev
21:45 japhb AlexDaniel: the problem with logarithmic scale is what your zero point is, because it can't actually be zero.
21:45 japhb Plus that last pixel or two makes a HUGE difference to the experienced speed out of proportion to the visual difference.
21:46 stmuk_ joined #perl6-dev
21:46 japhb The current code is designed to look kinda like it feels:  The above is "Slow for a while, then a bit faster, then slow again (with one measurement REALLY slow), then getting back to faster again.
21:47 japhb "
21:48 japhb Sadly unicode doesn't offer 16- or 32-unit-high bars
21:48 japhb For increased precision
21:49 japhb Though now that I think of it, I suppose buggable could rapidly answer *twice* and stack the two rows into a double-high chart for more precision.
21:49 japhb Oooh, I think I'm going to use this locally.
21:50 japhb Where is the buggable code?
21:50 japhb buggable: source
21:50 buggable japhb, See: https://github.com/zoffixznet/perl6-buggable
21:53 jdv79 is the optimization techniques deployed in moar similar to a tracing jit?
21:53 jdv79 i just remember the term and not any specifics
21:57 jdv79 is the irc log site going to be fixed for unicode anytime soon?
21:58 jdv79 I'd look into it if its something i could fix.  its not great to keep seeking ? chars all the time when backlogging.
22:02 MasterDuke jdv79: moritz and Zoffix have chatted about it a couple times, i think moritz said it's a mysql thing. could try searching the logs (either here or #perl6, don't remember which) for details
22:02 MasterDuke hm, maybe it was moritz and AlexDaniel
22:03 jdv79 didn't it used to work?
22:04 jdv79 i feel like it did cause i think i saw some unicode stuff in those logs before i updated my fonts so my term didn't bork them
22:04 jdv79 moritz: what's the status of that?  is there a ticket or anything else?
22:08 AlexDaniel jdv79: I looked at it a little bit but didn't make much progress
22:08 MasterDuke jdv79: moritz said he was going to sleep a little while ago, so you'll probably have to wait until tomorrow to chat with him
22:12 AlexDaniel jdv79: here's the last discussion: https://irclog.perlgeek.de/perl6/2017-08-22#i_15056532
22:20 MasterDuke another question about INTERPOLATE: i'm trying to correctly use index instead of eqat. just doing s/eqat/index/ fails this test
22:20 MasterDuke m: use Test; is "foobar" ~~ /$( $_.substr(3) )/, "bar", "Contextualizer with functions calls";
22:20 camelia rakudo-moar d13572: OUTPUT: «ok 1 - Contextualizer with functions calls?»
22:21 MasterDuke instead i get: # Failed test 'Contextualizer with functions calls' # at -e line 1 # expected: 'bar' #      got: 'foobar'
22:21 MasterDuke fyi, this is my patch: https://gist.github.com/MasterDuke17/c9512590df63bd530b640464ad04103f
22:22 MasterDuke afk for a bit, but i'll backlog for suggestions
22:25 Ven`` joined #perl6-dev
22:41 cognominal joined #perl6-dev
22:47 japhb If I want to return something multi-line from a buggable plugin, do I need to return a list or something \n separated?
22:47 japhb Zoffix: ^^
22:48 japhb Also, will each line be prefixed with "$nick, " ?
22:50 japhb FWIW, the code I have now can handle anything from 1-4 rows (the code can handle any number, but I figured > 4 was rude to the channel).  It looks like this:
22:50 japhb ???         ?            ?? ?????????????          dates: 2017-07-30–2017-08-31
22:50 japhb ?????????????     ?      ????????????????          range: 3.471s–4.925s
22:50 japhb ?????????????    ??      ???????????????? ??? ?    speed: 21% faster
22:50 japhb ??????????????????????????????????????????????????
22:54 japhb buggable: help
22:54 buggable japhb, tags | tag SOMETAG | eco | eco Some search term | author github username | speed | testers CPANTesters report ID
23:41 skids joined #perl6-dev

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