Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2017-06-09

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

All times shown according to UTC.

Time Nick Message
00:00 AlexDaniel joined #perl6-dev
00:06 benchable6 joined #perl6-dev
01:23 astj joined #perl6-dev
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 | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
02:41 BenGoldberg joined #perl6-dev
03:18 slick joined #perl6-dev
03:18 slick left #perl6-dev
03:52 astj_ joined #perl6-dev
04:14 eveo m: class Foo { multi method foo (Int) { self.foo }; multi method foo() { state $l = 42; dd $l} }; with Foo.new { .foo: 42; }
04:14 camelia rakudo-moar f0c3bf: OUTPUT: «Int $l = 42␤»
04:14 eveo m: class Foo { multi method foo (Int) { self.foo }; multi method foo(:$foo) { state $l = 42; dd $l} }; with Foo.new { .foo: 42; }
04:14 camelia rakudo-moar f0c3bf: OUTPUT: «Any $l = Any␤»
04:15 eveo Filed as https://rt.perl.org/Ticket/Display.html?id=131542
04:43 eveo ZOFVM: Files=1254, Tests=138776, 115 wallclock secs (20.30 usr  2.99 sys + 2324.89 cusr 150.79 csys = 2498.97 CPU)
04:47 Geth ¦ rakudo/nom: 3ff29d42f0 | (Zoffix Znet)++ | src/core/CompUnit/RepositoryRegistry.pm
04:47 Geth ¦ rakudo/nom: Permit IO::Path given as path to `use lib`
04:47 Geth ¦ rakudo/nom:
04:47 Geth ¦ rakudo/nom: Fixes RT#131540: https://rt.perl.org/Ticket/Display.html?id=131540
04:47 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3ff29d42f0
04:47 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=131540
04:47 Geth ¦ roast: 3dce753d7d | (Zoffix Znet)++ | S10-packages/basic.t
04:47 Geth ¦ roast: Remove trailing whitespace
04:47 Geth ¦ roast: review: https://github.com/perl6/roast/commit/3dce753d7d
04:48 Geth ¦ roast: 0bed372147 | (Zoffix Znet)++ | S10-packages/basic.t
04:48 Geth ¦ roast: Test `use lib` accepts IO::Path objects
04:48 Geth ¦ roast:
04:48 Geth ¦ roast: RT#131540: https://rt.perl.org/Ticket/Display.html?id=131540
04:48 Geth ¦ roast: Rakudo impl: https://github.com/rakudo/rakudo/commit/3ff29d42f0
04:48 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=131540
04:48 Geth ¦ roast: review: https://github.com/perl6/roast/commit/0bed372147
05:49 samcv .tell Zoffix yay! you read my mind about `use lib` accepting IO::Path objects! did i ever complain about that or are you a mind reader
05:49 yoleaux samcv: I'll pass your message to Zoffix.
08:38 geekosaur joined #perl6-dev
08:39 robertle joined #perl6-dev
08:42 [TuxCM] joined #perl6-dev
08:54 geekosaur joined #perl6-dev
09:57 eveo .
09:57 eveo There was a ticket for it.
10:09 eveo Though it was first mentioned here.
10:09 eveo https://irclog.perlgeek.de/perl6/2017-05-02#i_14518610
10:10 eveo .tell mscha It now does. RE: https://irclog.perlgeek.de/perl6/2017-05-02#i_14518610
10:10 yoleaux eveo: I'll pass your message to mscha.
10:10 samcv thanks eveo
10:11 lizmat Files=1204, Tests=61798, 222 wallclock secs (13.24 usr  4.93 sys + 1362.70 cusr 120.20 csys = 1501.07 CPU)
10:24 lizmat_ joined #perl6-dev
10:25 [TuxCM] This is Rakudo version 2017.05-386-g3ff29d42f built on MoarVM version 2017.05-71-gad6ab26f
10:25 [TuxCM] csv-ip5xs        2.915
10:25 [TuxCM] test            12.865
10:25 [TuxCM] test-t           4.359 - 4.475
10:25 [TuxCM] csv-parser      13.155
10:38 lizmat joined #perl6-dev
11:28 cono jnthn: looks like my issue has been fixed and I don't know how :( anything you can recall recently which fix segv ?
11:29 timotimo under what conditions did that segfault happen?
11:29 cono timotimo: https://github.com/perl6/roast/pull/270
11:30 cono if I understood correctly this object was collected earlier and another nursery was trying to collect it again
11:30 cono in gen2roots array was the object which shouldn't be there
11:30 timotimo oh, ouch
11:31 cono But I just ran the test on the recent rakudo[nom]+nqp+moar and it works
11:31 cono very very intereseted what was the issue :)
11:32 timotimo difficult to say
11:32 cono indeed
11:33 cono could be actually this: 239c6153115c63b8151a45193194b9e112a8ac76
11:33 cono in moar
11:34 timotimo hm, that doesn't look like it would impact SC objects or something
11:39 AlexDaniel joined #perl6-dev
12:11 lizmat joined #perl6-dev
12:17 jnthn Synchronous sockets got a complete rewrite. It's possible a bug got squelched.
12:18 jnthn Well, I know one did, but I meant in addition to the one I was shooting for :P
12:19 cono would you mind to look to my test rework then and merge it?
12:26 lizmat joined #perl6-dev
12:31 lizmat joined #perl6-dev
12:44 jnthn cono: Will try and find a moment for that later on
12:44 cono thanks a lot! :)
12:50 eveo NeuralAnomaly: status
12:50 NeuralAnomaly eveo, [✘] Next release will be in 1 day and 1 week. Since last release, there are 67 new still-open tickets (64 unreviewed and 0 blockers) and 384 unreviewed commits. See http://perl6.fail/release/stats for details
12:51 eveo should have toast data for any regressions since 2017.05 releaste today.
12:57 lizmat joined #perl6-dev
13:04 eveo m: $ = 'tmp/z1/z2/z3'.IO.mkdir; indir '/tmp/z1/z2', { say dir }
13:04 camelia rakudo-moar 3ff29d: OUTPUT: «("z3".IO)␤»
13:04 eveo m: $ = 'tmp/z1/z2/z3'.IO.mkdir; indir '/tmp/z1/', { indir 'z2', { say dir } }
13:04 camelia rakudo-moar 3ff29d: OUTPUT: «Failed to get the directory contents of 'z2': Failed to open dir: 2␤  in block  at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
13:04 eveo and indir got IO() coercer :/ wonder why it don't like being nested
13:05 eveo m: $ = 'tmp/z1/z2/z3'.IO.mkdir; indir '/tmp/z1/', { indir $*CWD.add('z2'), { say dir } }
13:05 camelia rakudo-moar 3ff29d: OUTPUT: «("z3".IO)␤»
13:06 eveo even tho that works :S
13:13 lucasb joined #perl6-dev
13:17 dogbert17 eveo: do you understand the last two lines of code on https://docs.perl6.org/type/UInt ? They look a bit bogus to me
13:19 eveo That code block is ridiculous. It's just a subset like any other, so the example should have like 1 basic thing and a link to the subsets docs for those who don't know them. Also the "defined as" bit is wrong. It's actually `where {not .defined or $_ >= 0};` so it allows type objects too.
13:20 eveo And the last two lines are just saying that you can use a UInt type in math operations just like regular Int, which isn't surprising, since it's just a subset
13:21 dogbert17 but why declare UInt $d and then not use it on next next line. And $a - 3 doesn't equal -3 since $a == 5. Bizarre
13:22 eveo dogbert17: ah, yeah :) That too is wrong :D
13:30 eveo ah, indir coercer is 'cause IO::Path.new stringifies $*CWD instead of .absoluting it
13:31 eveo The IO::Path.Str issue will haunt for ages
13:47 eveo ZOFVM: Files=1254, Tests=138776, 123 wallclock secs (23.06 usr  3.20 sys + 2523.04 cusr 137.77 csys = 2687.07 CPU)
13:50 dogbert17 does the following nqp statement return an Int:D? nqp::p6box_i(nqp::threadid($!vm_thread))
13:51 jnthn Yes
13:52 dogbert17 thx jnthn, is doing some minor updates to Thread.pod6
13:57 lizmat joined #perl6-dev
14:12 dogbert17 m: say Tap.^methods
14:12 camelia rakudo-moar 3ff29d: OUTPUT: «(BUILD new close)␤»
14:14 dogbert17 hmm, according to the docs, i.e. https://docs.perl6.org/type/Tap it has more methods, e.g. emit and supply. Is this correct or should I remove them?
14:15 jnthn No, it doesn't have those
14:15 jnthn And shouldn't
14:16 dogbert17 then I'll fix that immediately :)
14:16 AlexDaniel joined #perl6-dev
14:17 AlexDaniel oh no, and again I wrote something without seeing previous responses
14:17 ugexe m: sub foo(Str(Cool) $bar) { say $bar }; foo(IO::Path.new("/xxx")); foo(IO::Path.new("xxx/")) # eveo: i'm trying to say this behavior should be used instead of absolutifying it
14:17 camelia rakudo-moar 3ff29d: OUTPUT: «IO::Path is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting line 1␤  in method new at src/RESTRICTED.setting line 32␤  in block <unit> at <tmp> line 1␤␤»
14:17 AlexDaniel I should first get well, then finally move to another place, set up a proper internet connection there… and then join this kind of discussions
14:18 dogbert17 only 'close' will be left
14:18 jnthn Yup
14:19 jnthn Which is all you can do with a Tap
14:19 jnthn That's why you get it; so you can turn it off again later :)
14:20 dogbert17 fixed
14:29 eveo ugexe: that's wrong tho. You can't .Str a path if you intend it to use it as path still
14:29 ugexe its not being used as a path at that point
14:29 eveo ugexe: the Str candidate is still there. So all the strings will go through it.
14:30 eveo ugexe: what is it used as then? It's discarding .CWD attribute. my $p = "foo"; chdir "bar"; foo $p; will give wrong path
14:30 ugexe correct
14:31 eveo So why is the wrong path desirable?
14:31 dogbert17 more https://docs.perl6.org/type/Tap questions. There is a short example on the page which contains the line 'closing => { say "Tap closed" },'. Is that line legit?
14:32 ugexe because who to say its wrong? the CUR is supposed to decide this
14:33 eveo ugexe: the user is to say. If I do `use lib IO::Path.new: 'foo', :CWD</program1>` and you discard the "/program1" part the "foo" won't be pointing to the right location
14:34 ugexe should pass an already absolutified IO object
14:34 eveo :/
14:34 eveo I'm just gonna move the absolutifcation to `use lib` code in Actions
14:35 eveo ZOFFLOP: t/spec/S11-modules/require.t
14:38 ugexe m: require "lib/Test".IO.absolute.IO
14:38 camelia rakudo-moar 3ff29d: OUTPUT: «===SORRY!===␤No such method 'IO' for invocant of type 'Any'␤»
14:39 ugexe lol
14:39 ugexe m: say "lib/Test".IO.absolute.IO
14:39 camelia rakudo-moar 3ff29d: OUTPUT: «"/home/camelia/lib/Test".IO␤»
14:40 Geth ¦ rakudo/nom: b0fc9eecab | (Zoffix Znet)++ | src/core/io_operators.pm
14:40 Geth ¦ rakudo/nom: Simplify &indir code
14:40 Geth ¦ rakudo/nom:
14:40 Geth ¦ rakudo/nom: - The ||, &&, and ?? !! already codegen to nqp::if/nqp::unless
14:40 Geth ¦ rakudo/nom: - No functional changes
14:40 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b0fc9eecab
14:40 Geth ¦ rakudo/nom: 9151ebaaec | (Zoffix Znet)++ | src/core/io_operators.pm
14:40 Geth ¦ rakudo/nom: Fix $*CWD inside &indir when using relative paths
14:40 Geth ¦ rakudo/nom:
14:40 Geth ¦ rakudo/nom: $*CWD gets .Str stringified in IO::Path.new, so it should always
14:40 Geth ¦ rakudo/nom: be set to a IO::Path whose .path is an absolute path.
14:40 Geth ¦ rakudo/nom:
14:40 Geth ¦ rakudo/nom: This fixes the issue with nested indirs with relative Str paths
14:40 Geth ¦ rakudo/nom: as args and any other IO::Path creation inside indir
14:40 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9151ebaaec
14:41 Geth ¦ roast: 0757d88130 | (Zoffix Znet)++ | S32-io/indir.t
14:41 Geth ¦ roast: Test $*CWD is absolute inside indir
14:41 Geth ¦ roast:
14:41 Geth ¦ roast: Rakudo fix: https://github.com/rakudo/rakudo/commit/9151ebaaec
14:41 Geth ¦ roast: review: https://github.com/perl6/roast/commit/0757d88130
14:43 eveo spec isn't set?
14:43 eveo m: BEGIN dd $*SPEC; require "lib/Test".IO.absolute.IO
14:43 camelia rakudo-moar 3ff29d: OUTPUT: «IO::Spec::Unix <element> = IO::Spec::Unix␤===SORRY!===␤No such method 'IO' for invocant of type 'Any'␤»
14:43 eveo hm.
14:44 eveo Ah, it ends up being (require "lib/Test").IO.absolute.IO ?
14:45 ugexe fwiw we *do* have a way to pass in arguments to a cur in string form, which could possibly be used to handle the cwd issue (but may require a new cur class) ala `path#cwd<...>#some/path`
14:46 eveo but there's no cwd issue there?
14:46 * eveo is really confused
14:48 ugexe thats what a io::path argument to cur in string form should probably look like, so the cur can decide to absolutify it (with the proper cwd) or use the relative path part
14:52 bisectable6 joined #perl6-dev
14:53 ugexe https://github.com/ugexe/rakudo/blob/nom/src/core/CompUnit/Repository/AbsolutePath.pm#L13 # if $file.is-absolute {
14:53 ugexe https://github.com/ugexe/rakudo/blob/nom/src/core/CompUnit/Repository/FileSystem.pm#L91 # unless $file.is-absolute {
14:54 * dogbert17 is confused by the Supply/Tap docs
14:54 ugexe granted we probably should get rid of the first one
14:54 eveo gah. Google killed my toaster VM when it was almost done :( Was doing letter 'S' already
14:54 dogbert17 eveo: out of money?
14:54 eveo (the final toaster won't be on pre-emptible VM, but this one is)
14:55 eveo dogbert17: no, it's "pre-emptible". Meaning they can kill it at any time in exchange for it being zillion times cheaper
14:55 eveo Doesn't happen often, so for dev work it's a big win
14:55 dogbert17 aha, good to know
14:56 dogbert17 eveo: are you any good with supplies and taps?
14:57 eveo Not really.
14:57 dogbert17 :(
14:57 dogbert17 am a bit suspicious about this line from the Concurrency pod:
14:57 dogbert17 "Calling done on the supply object calls the done callback that may be specified for any taps, but does not prevent any further events being emitted to the stream, or taps receiving them."
14:58 ugexe which part?
14:58 dogbert17 but does not prevent  ...
14:59 dogbert17 perhaps I'm misunderstanding the text but to me it implies that I can do emit's on a supplier after having call 'done' and things will continue to work
15:00 ugexe maybe its referring to race condition
15:00 Geth ¦ rakudo/nom: 326faed6c5 | (Zoffix Znet)++ | 2 files
15:00 Geth ¦ rakudo/nom: Rework `use lib IO::Path...` feature
15:00 Geth ¦ rakudo/nom:
15:00 Geth ¦ rakudo/nom: Remove CURR interface change and instead convert any IO::Paths
15:00 Geth ¦ rakudo/nom: given to `use lib` to Str directly in World.
15:00 Geth ¦ rakudo/nom:
15:00 Geth ¦ rakudo/nom: This reverts commit 3ff29d42f026fde6ed4483aa022c6d9512c61700.
15:00 Geth ¦ rakudo/nom:
15:00 Geth ¦ rakudo/nom: Discussions:
15:00 Geth ¦ rakudo/nom: https://github.com/rakudo/rakudo/commit/3ff29d42f026fde6e#commitcomment-22461302
15:00 Geth ¦ rakudo/nom: https://irclog.perlgeek.de/perl6-dev/2017-06-09#i_14710384
15:00 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/326faed6c5
15:00 eveo ugexe: ^ that addresses your concerns, I believe.
15:00 dogbert17 ugexe: perhaps, guess I'll leave that line alone
15:01 ugexe dogbert17: well, what happens when you try to explicitly .emit after .done? if you cant then I agree its confusing
15:04 dogbert17 ugexe, m: my $supplier = Supplier.new; my $supply = $supplier.Supply; my $tap = $supply.tap( -> $v { say $v }); $supplier.emit("OK"); $supplier.done; $supplier.emit("What will happen");sleep 1;  # bad golf
15:04 dogbert17 m: my $supplier = Supplier.new; my $supply = $supplier.Supply; my $tap = $supply.tap( -> $v { say $v }); $supplier.emit("OK"); $supplier.done; $supplier.emit("What will happen");sleep 1;  # bad golf
15:04 camelia rakudo-moar 9151eb: OUTPUT: «OK␤»
15:05 dogbert17 it is indeed confusing
15:07 ugexe well
15:07 ugexe `await $supplier.done`
15:07 ugexe thats probably what needs to be clarified
15:07 ugexe nevermind thats not right heh
15:10 dogbert17 so we have that and then the 'closing' callback mystery in the example here https://docs.perl6.org/type/Tap
15:11 dogbert17 running the code in the example does not trigger the 'closing => { say "Tap closed" },' line
15:19 ugexe dogbert17: looks like supply is passing `closing` to something that expects `on-close`
15:19 ugexe well, nevermind that cant be it either (its positional)
15:21 ugexe i'd say this is a bug
15:24 dogbert17 oops
15:27 dogbert17 ugexe: the closing callback is mentioned in the signature here, https://docs.perl6.org/type/Supply#method_tap, but is is not mentioned in the docs below
15:27 ugexe should turn it into a roast test
15:28 ugexe https://github.com/rakudo/rakudo/blob/6d12cecc955ed997a832cff14346eb082456af19/src/core/Supply.pm#L102
15:28 ugexe https://github.com/rakudo/rakudo/blob/6d12cecc955ed997a832cff14346eb082456af19/src/core/Supply.pm#L12
15:28 lizmat joined #perl6-dev
15:28 ugexe so the docs *should* be right - it just isn't working
15:28 robertle joined #perl6-dev
15:29 Geth ¦ roast: a520037d7f | (Zoffix Znet)++ | S32-num/power.t
15:29 Geth ¦ roast: Test power ops with uncommon No chars
15:29 Geth ¦ roast:
15:29 Geth ¦ roast: Covers RT#126732: https://rt.perl.org/Ticket/Display.html?id=126732
15:29 Geth ¦ roast: Closes https://github.com/rakudo/rakudo/pull/1095
15:29 Geth ¦ roast: review: https://github.com/perl6/roast/commit/a520037d7f
15:29 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=126732
15:31 dogbert17 wonder why it's not working, the code looks innocent enough
15:32 Geth ¦ roast: ed4d510255 | (Zoffix Znet)++ | S32-num/power.t
15:32 Geth ¦ roast: Fill missing test descriptions
15:32 Geth ¦ roast: review: https://github.com/perl6/roast/commit/ed4d510255
15:33 dogbert17 looks at https://github.com/rakudo/rakudo/blob/6d12cecc955ed997a832cff14346eb082456af19/src/core/Supply.pm#L77
15:35 eveo The docs don't match the code.
15:35 eveo .tap is using &!closing and doesn't take :&closing arg
15:36 dogbert17 yeah, so who's right, the code or the docs?
15:37 eveo the roast
15:37 eveo So either the docs are wrong or the feature is unspecced
15:39 dogbert17 I'm leaning towards unspecced
15:41 dogbert17 would be nice to be able to correct the docs though
15:43 lizmat joined #perl6-dev
15:45 ugexe `self.Supply::tap(|c, closing => { ... });` this was in the older code, so I suspect the docs at least *were* correct at one point
15:52 dogbert17 ugexe: what about https://github.com/rakudo/rakudo/blob/6d12cecc955ed997a832cff14346eb082456af19/src/core/Supply.pm#L102 is it a leftover?
15:58 lizmat joined #perl6-dev
16:10 ugexe well that is the code that the snippet i pasted above replaces
16:13 eveo "Deprecated use of %*LANG<MAIN> assignment detected in use"
16:13 eveo damn, how to supress that :/ it's in the code path to support old slang interface
16:15 eveo "No such method '!make' for invocant of type 'Perl6::Grammar+{BioInfo::Grammar}'"
16:16 eveo Well, guess it don't matter much since I've no idea how to make the new way work anyway
16:18 eveo Weird. Slang::Tuxic doesn't have the same issues and I basically copied over its method. https://gist.github.com/zoffixznet/fce7475f31492f911deae8fa6ecbbccc
16:21 eveo s:3rd/'$MAIN-grammar'/$MAIN-actions/ # doesn't avoid the issues tho
16:26 dogbert17 ugexe: I wonder what the intended workings of this are supposed to be
16:29 lizmat joined #perl6-dev
16:30 eveo .ask TimToady any idea what to do about "No such method '!make'" when trying to mix in a slang? Having trouble fixing BioInfo module: https://github.com/MattOates/BioInfo/issues/7
16:30 yoleaux eveo: I'll pass your message to TimToady.
16:33 ugexe m: my $supplier = Supplier.new; my $supply = $supplier.Supply.on-close({ say "closing" }); my $tap = $supply.tap(-> $v { say "the value is $v" }); $supplier.emit(100); $tap.close;
16:33 camelia rakudo-moar 326fae: OUTPUT: «the value is 100␤closing␤»
16:35 dogbert17 aha, so the behaviour has changed a bit from what it used to be
16:37 ugexe i wonder if https://github.com/rakudo/rakudo/blob/nom/src/core/Supply.pm#L102 should be like #L234
16:38 dogbert17 who would know, jnthn, lizmat?
16:39 ugexe probably. but you could always try and see what happens :)
16:39 lizmat dogbert17: doesn't ring a bell
16:39 lizmat afk again&
16:48 * jnthn tries to figure out what the Supply question is :)
16:48 dogbert17 ugexe: I'll give it a shot
16:48 jnthn It's certainly the case that Tap takes a thing to run when the tap is closed in its constructor.
16:49 jnthn Not that normal code would ever produce a Tap instance by itself
16:49 jnthn The Supply.on-demand factory method lets the user pass in something to do when the tap is closed, so they can do a resource cleanup
16:49 Zoffix_ joined #perl6-dev
16:50 jnthn .on-close is the more typical user-level way to insert cleanup into a Supply chain
16:50 [TuxCM] joined #perl6-dev
16:50 jnthn Well, Supply.on-demand is user-level too, but you might have a Supply from elsewhere
16:50 dogbert17 we were discussing this line, https://github.com/perl6/doc/blob/master/doc/Type/Tap.pod6#L15
16:51 dogbert17 which does not seem to to anything
16:51 jnthn Ah
16:51 jnthn method tap does not take a closing named argument any more
16:52 jnthn Hasn't since the big Supply refactor of 2015 :)
16:52 jnthn You'd use .on-close({...}).tap(...) now
16:52 dogbert17 aha, the fog begins to lift :)
16:53 dogbert17 we also questioned the following text from the concurrency pod: "Calling done on the supply object calls the done callback that may be specified for any taps, but does not prevent any further events being emitted to the stream, or taps receiving them."
16:54 dogbert17 the second part of that text "but does not prevent ..." feels slightly wrongish
16:55 jnthn Yeah
16:55 eveo sqlite> select count(*) from toast where status != 'Succ' and rakudo = '2017.05';
16:55 eveo 169
16:55 eveo sqlite> select count(*) from toast where status != 'Succ' and rakudo = 'e538cbc';
16:55 eveo 128
16:55 eveo Well, at least overall looks like a lot more modules are passing on HEAD than on 20167.05 :)
16:56 jnthn Almost every Supply you'll encounter today is "sane"
16:56 jnthn Or "sanitry2
16:56 TimToady eveo: that looks more like assumptions from before the Cursor/Match merge
16:56 yoleaux 16:30Z <eveo> TimToady: any idea what to do about "No such method '!make'" when trying to mix in a slang? Having trouble fixing BioInfo module: https://github.com/MattOates/BioInfo/issues/7
16:56 jnthn Which means that the protocol emit* [done|quit] is enforced
16:56 TimToady I'd just change .'!make' to .make and see if it works
16:57 jnthn So you'd have to really go out of your way to see emit after a done :)
16:57 eveo Ohhhhh... I didn't even notice there was a !make in BioInfo. I assumed it was something from Rakudo/nqp
16:58 TimToady there is no '!make' in either nqp or rakudo anymore
16:59 eveo Thanks.
16:59 dogbert17 m: my $supplier = Supplier.new; my $supply = $supplier.Supply; my $tap = $supply.tap( -> $v { say $v }); $supplier.emit("OK"); $supplier.done; $supplier.emit("What will happen");sleep 1;  # bad golf
16:59 camelia rakudo-moar 326fae: OUTPUT: «OK␤»
16:59 lizmat joined #perl6-dev
17:00 TimToady user modules that rely on private nqp implementation methods sometimes get burned  :)
17:00 eveo :)
17:01 dogbert17 so emit after done won't do anything it seems
17:01 eveo "Incompatible MROs in P6opaque rebless for types Perl6::Grammar+{BioInfo::Grammar} and Perl6::Grammar"
17:01 eveo Well, not getting the make error anymore :)
17:02 dogbert17 thx jnthn and ugexe, will fix up the example
17:05 TimToady eveo: perhaps it's "is"-ing some class it shouldn't be anymore?
17:09 eveo Ah, OK $*LANG.^mixin(Some::Grammar) is not the same as %*LANG<MAIN>.^mixin(Some::Grammar)
17:10 eveo All tests pass \o/
17:16 TimToady you probably want $*LANG.slang_grammar('MAIN').^mixin(Some::Grammar) or so
17:17 eveo Works. Thanks.
17:18 TimToady it currently is still implemented as a hash internally, but with the abstraction layer we're free to change that now
17:19 TimToady once all the module catch up :)
17:23 eveo SQLite db with ecosystem toastings for 2017.05 and e538cbc (almost HEAD): https://temp.perl6.party/2017-06-09.toast.sqlite.db
17:23 * eveo starts finishing off a viewer for it
17:29 lizmat joined #perl6-dev
17:32 eveo joined #perl6-dev
17:33 eveo Look at what I got \o/
17:33 * eveo points at host
17:35 timotimo nice
17:52 TimToady bisect: 3 !! 4
17:52 bisectable6 TimToady, On both starting points (old=2015.12 new=326faed) the exit code is 1 and the output is identical as well
17:52 bisectable6 TimToady, https://gist.github.com/ce4c901a95671e67e5687865ea452a04
17:52 TimToady coulda sworn that used to complain about a missing ??, but maybe that was STD
17:55 eveo c: all 3 !! 4
17:55 committable6 eveo, https://gist.github.com/9b3e459206823af9e5068a6b2d5c2f7f
17:56 eveo All are either TTIAR or "ConfuseD"
18:00 lizmat joined #perl6-dev
18:30 eveo "Your !! was gobbled by the expression in the middle; please parenthesize"
18:30 eveo aw c'mon! I specifically didn't use parens because it was obvious where they should end -_-
18:30 eveo ZofBot: DEATH TO PARENS!
18:30 ZofBot eveo, 22 csys = 1316
18:31 lizmat joined #perl6-dev
18:33 * lizmat is developing a distaste for Ziggo's ISP qualities
18:34 lizmat looks like the connection goes down about once an hour  :-(
18:40 eveo :(
19:02 lizmat joined #perl6-dev
19:03 AlexDaniel joined #perl6-dev
19:11 lucasb joined #perl6-dev
19:18 mst lucasb: at least you're not using http://www.verelox.com/
19:19 lucasb mst: hello
19:19 mst bah
19:19 mst tabfail
19:19 lucasb mst: but.. sorry, what?
19:19 mst lizmat: ^^
19:19 lucasb ah, ok
19:19 mst lucasb: my tab complete glommed onto your join
19:26 eveo .tell dogbert17 don't know if you want to bother with it, but here's a bug that I can only repro on my 32bit box. Also the apparent dependence on command line args and path suggest this might be a Rakudo bug: https://github.com/perl6/DBIish/issues/93
19:26 yoleaux eveo: I'll pass your message to dogbert17.
19:32 lizmat joined #perl6-dev
19:52 dogbert17 eveo: the problem shows up on my machine as well
19:52 yoleaux 19:26Z <eveo> dogbert17: don't know if you want to bother with it, but here's a bug that I can only repro on my 32bit box. Also the apparent dependence on command line args and path suggest this might be a Rakudo bug: https://github.com/perl6/DBIish/issues/93
20:02 lizmat joined #perl6-dev
20:04 eveo sweet
20:04 eveo A reproduceable bug is a fixable bug :)
20:04 dogbert17 eveo: running it under strace uncovers the following:
20:04 dogbert17 write(2, "Invocant of method 'Bridge' must"..., 581Invocant of method 'Bridge' must be an object instance of type 'Int', not a type object of type 'Int'.  Did you forget a '.new'?
20:04 dogbert17 in method _row at /home/dogbert/repos/rakudo/install/share/perl6/site/lib/DBDish/SQLite/StatementHandle.pm6 (DBDish::SQLite::StatementHandle) line 51
20:04 dogbert17 in method row at /home/dogbert/repos/rakudo/install/share/perl6/site/lib/DBDish/StatementHandle.pm6 (DBDish::StatementHandle) line 77
20:04 dogbert17 in block  at /home/dogbert/repos/rakudo/install/share/perl6/site/lib/DBDish/StatementHandle.pm6 (DBDish::StatementHandle) line 94
20:05 dogbert17 in block <unit> at -e line 1
20:05 eveo I saw these errors come up before I started doing .elems too
20:08 dogbert17 --ll-exception is a good thing to use here
20:33 lizmat joined #perl6-dev
21:04 lizmat joined #perl6-dev
21:43 lizmat joined #perl6-dev
21:52 geekosaur joined #perl6-dev
21:59 lizmat joined #perl6-dev
22:06 lizmat joined #perl6-dev
22:44 MasterDuke joined #perl6-dev
23:37 robertle_ joined #perl6-dev

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