»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
lookatme o/ 00:15
El_Che lo 00:22
lookatme lol 00:30
lookatme m: say so 1 ~~ (Int, Str).all; #`(Is < > equal) say so 1 ~~ Int && 1 ~~ Str; 00:45
camelia False
False
thou I did some more revision on the bug I've been fussing with, and reported it. It's interesting, the problem shows up with type names like Pod::Foo and IO::Foo, but not File::Foo or Test::Foo. So it's a class of built-in types only. gist.github.com/softmoth/c8e90a6e4...9c5931e1b2 01:02
Zoffix thou: I've seen tickets for similar (same?) issues: RT#130794 RT#130200 01:07
synopsebot RT#130794 [new]: rt.perl.org/Ticket/Display.html?id=130794 Cannot import a symbol from precomp if its namespace is occupied
Zoffix There might be more
synopsebot RT#130200 [new]: rt.perl.org/Ticket/Display.html?id=130200 [LTA] Error when attempting to use a package that has one of core namespaces in its name
Zoffix lookatme: yes, those are similar in meaning, except that Junctions are a hint to the compiler that it may use more than one thread to perform the calculation 01:09
lookatme Zoffix, thanks
lookatme I wonder in this situation, is Perl6 will do short-circuit evaluation in Junction calculation ? like the logical operator && does ? 01:17
bazzaar drat, wordpress ate my array assignment code line in the comment I added to bduggan's advent post, it should have read my @bricks = <ok split ok ok split ==== >; or even my @bricks=['ok', 'split', 'ok', 'ok', 'split', '====']; 02:24
I can't edit it to clean it up :( 02:25
bazzaar In the words of the Bonzo Dog Doodah Band, 'Sometimes you just can't win ..' :-) 02:30
piojo If I implement .AT-KEY on a container class to allow assignment to empty elements (like @l[1, 3] = 0, 0; or my $elem := %h<x>; $elem = 0;), the right technique is to use a Proxy container for that, right? 03:26
So if the value doesn't exist, .AT-KEY should return a proxy which, if assigned to, will write back into the data structure. 03:27
Zoffix bazzaar, fixed 03:38
Need to use HTML entities: &lt; &gt;
piojo: unsure. In the core it's implemented using a secret mechanism (that will soonish go away) and not a Proxy
Zoffix lookatme: yes, currently it'll shortcurcuit: github.com/rakudo/rakudo/blob/mast...#L190-L197 But you shouldn't rely on it being executed in order or definitely short-circuited. I imagine that code will be more complex when Junctions will actually autothread, since we wouldn't want them to always autothread and there will need to be a thing that figures out when threading is worth 03:42
it, etc.
piojo Zoffix: thanks. If I go ahead with this, I'll try it with Proxy.
lookatme Zoffix, oh, I think all Junction not worth auto thread in most case :) 03:44
Zoffix m: my @a = |((^500_000+3).map: * * 2), 2; say 1 == @a.race(:batch(@a/4)).grep(*.is-prime).elems; say now - INIT now 04:07
camelia True
1.26688215
Zoffix m: my @a = |((^500_000+3).map: * * 2), 2; say 1 == @a.grep(*.is-prime).elems; say now - INIT now
camelia True
2.6217620
Zoffix m: say 2.62/1.266
camelia 2.069510
Zoffix 2x boost for a `one` Junction for a half-mil `*.is-prime` computation :) So somecases are worth it 04:08
lookatme hmm, thanks 04:44
lookatme m: react { whenever supply { emit 123; @ = (< 1 2 3 4 5 6 >.race.map(sub ($x) { emit $x; } ).list) } { .say } } 05:17
camelia 123
An operation first awaited:
in block <unit> at <tmp> line 1

Died with the exception:
A worker in a parallel iteration (hyper or race) initiated here:
in block <unit> at <tmp> line 1

Died at:
emit without…		 05:18
lookatme oh its different thread 05:19
bazzaar Zoffix: thanks for editing the message, I'll try to remember about using the html entities. There's actually still a redundant one line message that ideally could be deleted, but I'm grateful for what you did. 06:28
lizmat clickbaits p6weekly.wordpress.com/2017/12/11/...injectile/ 08:45
yoleaux 11 Dec 2017 17:57Z <AlexDaniel> lizmat: irclog.perlgeek.de/perl6-dev/2017-...i_15562338
lizmat argh, missed that cause of yoleaux absence, I guess 08:46
AlexDaniel lizmat: just edit it in? 08:53
when I was outside perl 6 community, I'd go through several weeklies once a month 08:54
AlexDaniel so I think that some edits a few hours later are totally ok 08:55
lizmat well, it's already 10+ hours later now
AlexDaniel so? :) 08:56
maybe we should make a page with a list of this reports… hmmm
AlexDaniel does it
lizmat AlexDaniel++ 08:57
TEttinger what are quasi and injectile? I didn't see them in the bait 08:58
TEttinger oh ok, 007 09:01
lizmat :-)
AlexDaniel reportable6: list 09:03
reportable6 AlexDaniel, gist.github.com/ad96cf9f9e35012465...ed7791e643
AlexDaniel “You get full refunds if purchasing List.tail doesn't make you happy :-)” bwahahaha 09:10
( twitter.com/LearningPerl6/status/9...4736768007 ) 09:11
lizmat :-) 09:12
parv that sole full-refunds comment is a flippant response, serves only to turn people off. would have been better with (link to) explanation/details. 09:28
moritz humor only turns off people? really? 09:29
it amused at least three people in this channel alone
parv that's great 09:30
moritz parv: anyway, join the conversation on twitter, post a link to the docs that explain how cheap or expensive it is 09:31
parv i would have if i knew about what- or where to find it
moritz parv: sure 09:32
lizmat fwiw, tail(*-N) iterates over the iterable and keeps the last N values it has seen, and starts producing these as soon as the original iterator is exhausted 09:33
so it will only ever keep N values in memory
DrForr_ The last I saw about that he was wishing tail() was actually cdr(), a suggestion that I kinda like but isn't in line with the purpose of the method. 09:39
AlexDaniel … what does it even stand for 09:43
I see that it's a lisp thing, but what's the idea behind naming it “car” and “cdr”? 09:44
DrForr_ Dates all the way back to the PDP-8, when you had an Address Register and Data Register.
(not that modern CPUs don't have separate registers, this was back when these were Big Deals. 09:45
So 'CAR' was the Contents of the Address Register, in other words what the head of the list referenced, and the Data Register was the pointer to the rest of the list.
moritz DrForr_++ # clearest explanation of these names I've seen yet 09:46
so far, the best I've seen was "related to some register names"
DrForr_ I started on the PDP-8 :)
AlexDaniel ahhhh… 09:46
DrForr ) 09:47
AlexDaniel by the way, why not $list[1..*] ? 09:48
AlexDaniel I mean, when we compare it to $list.tail(*-1) 09:48
lizmat $list[*-1] will keep $list in memory and create a slice on that 09:49
which will probably mean everything is in memory twice, at the current state of things
lizmat intends non-single element slices to always be Seqs 09:50
m: my @a = ^10; dd @a[1..*].WHAT # so everything is in memory twice :-( 09:51
camelia List
lizmat which is a pity in the case of a "for @a[1..*]"
which would then reSeq the generated List, which could have been a Seq in the first place 09:52
AlexDaniel mhm tail seems to be at least twice faster, yeah
lizmat afk& 09:54
AlexDaniel releasable6: next
releasable6 AlexDaniel, Next release in 4 days and ≈9 hours. No blockers. 0 out of 97 commits logged
AlexDaniel, Details: gist.github.com/8e45d50f7d60dc2828...6f599aa3c3
Geth doc: 43f06f9781 | (Tom Browder)++ (committed using GitHub Web editor) | doc/Language/pod.pod6
fix typo
 10:44
synopsebot Link: doc.perl6.org/language/pod
Geth doc: e1acf03ec5 | (Tom Browder)++ (committed using GitHub Web editor) | doc/Language/pod.pod6
add words directly from Synopsis 26

Add info on configuration for delimited blocks.		 10:52
scimon Am I missing something obvious? Is there a command line option to get the Pod for an installed module? Like perldoc? 10:53
(Or is there a module for it?)
teatime p6doc ? 10:54
scimon See I thought that was a thing. I don't seem to have it though... I may have to look at my installation again. I was using rakudobrew, then moved to a nice PPA that Johnaton F was updating but it's not been updated since October. That was nice because it linked into the system updater. Hmmmm. 10:56
Found it.
teatime there's packages 11:01
that install a good build, into /opt 11:02
github.com/nxadm/rakudo-pkg/releases
docs.perl6.org/syntax/INTRODUCTION 11:03
timotimo moritz: "I am very happy [...]. I hope you do too!" might want "are" instead of "do"? 11:05
moritz timotimo: thanks, updated 11:07
timotimo good post :)
moritz thanks!
timotimo jeez, the sudoku solver generates a 180 megabyte big profiler html blob 11:11
good thing i'll be writing that frontend to the sql profiler ...
ah, of course, it's deeply recursive 11:15
scimon On some puzzles yes. 11:19
The problem I have with the rakudo-pkg packages is you have to manually update them and I'm lazy. WAY too lazy. 11:20
Having read all the books in the list bar the first and last (and I'm getting the last next week) I probably should review them. moritz++ for a good breakdown there :) 11:22
(Meanwhile my ongoing plan to get paid to write Perl6 is getting traction :) ) 11:23
Meahwhile I need to learn how solr works. On the bright side the Cro::HTTP::Client lib is great for testing API's with. 11:25
AlexDaniel unicodable6: arrow 11:30
unicodable6 AlexDaniel, U+02C2 MODIFIER LETTER LEFT ARROWHEAD [Sk] (˂)
AlexDaniel, U+02C3 MODIFIER LETTER RIGHT ARROWHEAD [Sk] (˃)
AlexDaniel, 617 characters in total: gist.github.com/0fc9f483c5b8cec57f...786e67a612
AlexDaniel not enough! 11:32
I loved ⌁ so much but then I realized that some fonts render it angled 11:33
unicodable6: right arrow 11:35
unicodable6 AlexDaniel, U+02C3 MODIFIER LETTER RIGHT ARROWHEAD [Sk] (˃)
AlexDaniel, U+02F2 MODIFIER LETTER LOW RIGHT ARROWHEAD [Sk] (˲)
AlexDaniel, 217 characters in total: gist.github.com/59f6280504496db739...6c16fb2083 11:36
timotimo performance-wise, the sudoku solver spends rather a lot of time in signature binding and dispatchee selecting, which i believe is caused from junction autothreading 11:39
timotimo the busiest sub by far is find_best_dispatchee, spending 5_635_069 usec, followed by bind_one_param which spends 2_326_574 usec, then infix:<==> 11:43
infix:<==> from numeric (that's the proto i bet) has 7_097_826 usec inclusive time, but only 1_044_802 usec exclusive. add to that int's infix:<==> with 160_611 usec; that leads me to believe that dispatching the right version of infix:<==> over and over via autothread_junction is what causes all the find_best_dispatchee and bind_one_param work 11:45
Voldenet scimon: Cro::HTTP::Client is decent, but I'd not say it's great 11:50
there's seem to be many use cases docs don't cover, like using custom ca file 11:51
scimon Part of my issue in the past has been getting bitten by HTTPS requests and threading. 11:52
(Which now I think about it I don't need for this)
scimon caching the resluts of valid, full and complete speed up the solver... (Scimon... stop poking it). 11:58
timotimo is complete just full + valid? 12:00
also, invalidating with a single call to cell could do a quick check of only the impacted pieces of the puzzle 12:01
scimon It is 12:02
timotimo reset, of course, would have to do a full re-check if you don't want to calculate a diff and such
scimon I was so wedded to the complete junction because it was the first one I thought of.
Bit the bullet.
Deleting it.
timotimo it's beautiful. just rather slow :D
scimon And yeah I might change the cache cearing code. 12:03
Version 1.1.1 will be on github soon.
running on my local box mi6 test went from 14 seconds -> 8 seconds.
timotimo nice. 12:04
scimon Does && short circuit?
timotimo it does
scimon So self.full && self.valid will only check validity once the puzzle is full. 12:05
timotimo m: (say "hi" && 0) && (say "bye" && 1)
camelia 0
1
timotimo oops, heh
m: ((say "hi") && 0) && ((say "bye") && 1)
camelia WARNINGS for <tmp>:
hi
Useless use of constant integer 1 in sink context (line 1)
scimon :D
Anyway. The Valid junction is frankly even more bonkers. That's a nights sleep I'm not getting back. 12:06
timotimo okay, that's actually kind of hilarious
oh, probably just a stdout/stderr ordering issue? 12:07
m: ((note "hi") && 0) && ((note "bye") && 1)
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
hi
timotimo OK, that makes more sense
scimon timotimo : Just pusshed another change to github. Now it does smart cache clearing on a .cell call. 12:17
scimon Wow... so "solving" an empty puzzle with yesterdays code on my machine took 22 seconds. With the updates today (mmm caching) down to 3 seconds. 12:19
Weee.
Right. going to ruslte up lunch and write the creator on my lunch break. 12:20
timotimo i wonder how expensive the repeated "$tx,$ty" concatenations are. compared to, say, having $tx*10+$ty, or having a two-dim array caching these strings
scimon I'm terrible at picking the simplest way of doing things (as I see it) over the most efficient. 12:35
timotimo :) 12:40
wait, if you are bad at that, doesn't that mean that you accidentally do things the most efficient way? :) 12:41
pochi I'm trying to use SetHash to generate a set of uniq objects (Points), how can I tell SetHash to key on something other than object reference? 12:48
timotimo implement a WHICH method for your Point class 12:49
pochi what should it return? 12:52
an int?
timotimo s: Pair, "WHICH", \()
SourceBaby timotimo, Sauce is at github.com/rakudo/rakudo/blob/4fca.../Mu.pm#L26
timotimo we return ObjAt objects from those 12:53
you don't have to (or even should) use nqp:: ops for that, you can just ObjAt.new with a string
s: ObjAt, "new", \(Str:D)
SourceBaby timotimo, Sauce is at github.com/rakudo/rakudo/blob/4fca...Mu.pm#L122
pochi ah
timotimo hm, huh?
s: ObjAt, "new", \(str) 12:54
SourceBaby timotimo, Sauce is at github.com/rakudo/rakudo/blob/4fca...Mu.pm#L122
timotimo that's strange
m: say ObjAt.new("1234")
camelia 1234
timotimo m: say ObjAt.new("1234").perl
camelia ObjAt.new("1234")
timotimo it does work, but sourcery doesn't find it. must be about natives vs objects perhaps?
ilmari m: (foo => 42).WHICH.say 12:55
camelia Pair|Str|foo|Int|42
ilmari m: (foo => 42).WHICH.WHAT.say 12:56
camelia (ObjAt)
pochi yay! I gets uniq values :-) thank you
scimon Righto. Releasing 1.1.1 and going to start work on the creator... new branch I think. 13:03
buggable New CPAN upload: Game-Sudoku-1.1.1.tar.gz by SCIMON cpan.metacpan.org/authors/id/S/SC/...1.1.tar.gz 13:18
timotimo Uncaught RangeError: Maximum call stack size exceeded 13:56
at JSON.parse (<anonymous>)
Altreus Is there some concept that shortens an if-defined? I want to add an id to a URL if id is defined, but i want to include the extra / in that if-defined. $str ~ (defined $id ?? "/$id" !! '') 14:27
Sorry if that's not legit p6
Is there an idiomatic way of defining some "/$id" such that the whole concatenation collapses to no-op if id is not defined? 14:28
timotimo we hvae with and without 14:30
perlpilot Altreus: $str ~ ("/$id" with $id); 14:31
timotimo m: my $id; say "foobar" ~ ("/$id" with $id)
camelia foobar
timotimo m: my $id = 99; say "foobar" ~ ("/$id" with $id)
camelia foobar/99
Altreus Oh em gee
Altreus Arigato old bean 14:32
ilmari my $id; ("/$id" with $id).WHAT.say 14:34
m: my $id; ("/$id" with $id).WHAT.say
camelia (Slip)
Altreus I thought slip but my reading about it did not furnish me with a clear answer 14:37
Also, is there/what is the equivalent of requires in a p5 moose role? 14:38
"I'll use this method but i don't commit to an implementation"
timotimo m: role NeedsMeh { method meh { ... } }; class Oh does NeedsMeh { }
camelia 5===SORRY!5=== Error while compiling <tmp>
Method 'meh' must be implemented by Oh because it is required by roles: NeedsMeh.
at <tmp>:1
timotimo m: role NeedsMeh { method meh { ... } }; class Oh does NeedsMeh { method meh { "i has meh!" } } 14:38
camelia ( no output )
Altreus So ... Is 14:39
is doing that?
timotimo yup
Altreus Oops, sorry, phone keyboard
Thanks again ^^
timotimo YW
Altreus It works 14:42
This is fun
Who dares make programming fun?
perlpilot Altreus: welcome to Perl 6 ;-)
scimon One of the things I'm loving right now is learning new stuff. 14:59
TIL with. :)
AlexDaniel
.oO( who dares make non-perl 6 programming non-fun? )
 15:33
tbrowder how often is docs website updated? 15:45
tbrowder i ask because my last doc change passed travis testing 4 hrs ago but i don’t see the change. 15:46
on docs.perl6.org 15:47
moritz it should be faster than that
docs.perl6.org/build-log/ has the logs
moritz docs.perl6.org/build-log/build-201...+00:00.log looks a bit unhappy 15:49
tbrowder yep, probably my input. i’ll try to remove some formatting 15:50
moritz thanks!
Geth doc: 26c25f6290 | (Tom Browder)++ (committed using GitHub Web editor) | doc/Language/pod.pod6
remove formatting for now
 15:58
synopsebot Link: doc.perl6.org/language/pod
timotimo fun. this web server got owned with a phishing thingie installed on it. it still has the source zipfile in the server's root %) 16:06
"got owned and a ... installed on it" i should say 16:07
moritz most breaches are pretty unsophisticated 16:11
Geth doc: 1c410ac942 | (Tom Browder)++ (committed using GitHub Web editor) | doc/Language/pod.pod6
change table to delimited block form

remove :nested config key		 16:31
synopsebot Link: doc.perl6.org/language/pod
timotimo yeah, i imagine there is a ginormous "long tail" of web servers that were forgotten or something that easily get themselves infected 16:36
mspo shell shock via cgi-bin/ is still around 16:37
probably the easiest hack going
just curl commands to web servers you find sitting around 16:38
timotimo oh, of course 16:40
how fun is it to win these servers back?
geekosaur the _really_ easiest is throwing that at random routers 16:41
the cheapest ones are vxworks, but anyone who needs a bit more installs a linux-based firmwae... and then never updates it
mspo cloud hasn't been great for internet security at-large :) 16:43
geekosaur and, of course, since oyou've taken over a router, you've usually got access to multiple internal machines
'cloud', 'security', only usable with a 'not'
*usable together
(alt: "IoT" = insecurity of things) 16:46
Geth doc: 4f665f7622 | (Tom Browder)++ (committed using GitHub Web editor) | doc/Language/pod.pod6
remove format code D, use C instead

D may not yet be implemented		 16:47
synopsebot Link: doc.perl6.org/language/pod
timotimo "IoT: the S stands for Security" 16:51
tbrowder docs look good now 17:28
thou AlexDaniel, I've responded to your questions on rakudo/rakudo#1300, thank you! 17:51
synopsebot RAKUDO#1300 [open]: github.com/rakudo/rakudo/pull/1300 Untwine Pod
thou AlexDaniel, I'm just not sure how to coordinate communicating about rakudo + roast changes. But the test changes are prepped and ready to go. 17:59
pochi could someone take a look at pastebin.com/PhWD1saG 18:32
I'm getting this error which confuses me: Cannot look up attributes in a Point type object 18:33
(calling is-close from main works, but calling it from BUILD seems to generate that error) 18:34
oh, and I'm using rakudo star 2017.10 18:35
geekosaur during BUILD your object isn't fully constructed yet; you are calling a method on an undefined value
pochi ah, when I try to print $!p1 it is indeed empty 18:37
so, what is the correct method for initializing $.is-short? 18:38
geekosaur also defining your own new is generally a bad idea
you might want to use TWEAK, since it runs after object construction. see docs.perl6.org/language/objects#Ob...nstruction
pochi I see 18:39
what's the purpose of BUILD?
geekosaur mostly, control over when superclass initialization takes place
pochi I use my own new to get my prefered interface for the class btw 18:41
geekosaur well, no. BUILDALL does that, BUILD has lower level control over object construction itself but can't do things with the partially constructed object (it may not have constructed superclasses fully yet iirc? this is up to BUILDALL)
so TWEAK gives you the ability to make adjustments to a built object before it's returned from the constructor
in partiuclar you don;t necessarily have any fields initialized at BUILD time, but do at TWEAK time 18:42
for the full story, there's a page in the docs with more detail or you can inspect the default new and BUILDALL methods
perlpilot pochi: you could make your own method create(int $x, int $y) { Point.new( :$x, :$y ); } instead of making your own new (just a thought)
pochi ah, but I want all classes to have the same constructor 18:43
my new seems to work though, why shouldn't I use that method?
lizmat multi method new(Point:U: int $x, int $y) { self.new(:$x, :$y) } # that should work, no ? 18:44
pochi if I try that I get: Default constructor for 'Point' only takes named arguments 18:47
perlpilot pochi: you should use whatever you want if it makes sense to you. :-) 18:50
pochi at least I have to understand my own code :-) 18:51
perlpilot pochi: though I find it a little strange that you would want to be rid of the "normal" mechanism that's also a little more self-documenting: Point.new( :x(1), :y(2) );
pochi ... and it's great if it works too :-)
pochi well, take the Line class then, I don't want a constructor with p1 and p2, feels unnatural 18:52
also verbose for a Point/Line class
pochi (p1 and p2 should probably be private as well, but then things didn't work) 18:53
perlpilot pochi: point taken ;) 19:06
rindolf Perl's 30th birthday is next monday 19:29
thou I'm excited, I think I've fixed the ugly extra spacing in Pod::To::Text. No more spaces at end of lines or on blank lines in the output.... Simple patch, I'm just running a few more tests on it. 19:40
comborico1611 rindolf, interesting. 19:44
DrForr thou: I'm rewriting the test suite for Pod::To::BlogspotHTML as it happens. 19:45
thou DrForr, nice. I did a little on Pod::To::Markdown conversion to use TreeWalker. I think it might make sense for us both to do something like that independently (either using Pod::TreeWalker itself or making some other visitor implementation as needed), and then we can compare implementations and refactor so ::Text, ::HTML, ::Markdown, ::BlogspotHTML, etc. can DRY as much as possible. 19:47
DrForr Well, it's actually going to be both ::HTML and ::BlogspotHTML because I'm going to get subclassing working properly beforehand. 19:48
thou I think there's a lot of stuff that could even benefit being moved into core, but if we can get everything using a shared Pod::To::Generic, it'd be just about as good. I.e., put all the messy translations of confusing bits that show up in $=pod into a generic handler that all the Pod::To::* utilize. 19:49
BTW I have commit access on Pod::TreeWalker in case we want to continue to use it. 19:50
autarch would likely be happy to add you, too, I assume
DrForr I'm tempted to rewrite it because I already have to do some generic munging of <p/> tags. 19:51
perlpilot had dreams of a Pod::To::LaTeX at one point ... maybe that will be more possible soonish? 19:52
DrForr And IIRC it is adding tags that I need to remove. I want to rewrite it to get rid of hte munging...
perlpilot: That's also on my list.
perlpilot DrForr: nice!
DrForr I'm by no means an expert, but I'd love to be able to throw docs into LaTeX and get better highlighting and typography. 19:54
perlpilot same here. I was mainly dreaming about it for book publishing using Pod. But like many things ... I got side tracked. 19:55
DrForr Well, that was kind of my target too, but the \listing modules out there are kind of awkward. I'd want to use a decent code listig module. 19:56
DrForr The problem is that $..$ is used for math mode, and that *does* get annoying. 19:57
DrForr thou: I imagine autarch would hand me the author bit if I asked :) 20:09
Not that I want it. 20:10
thou :) 20:11
DrForr perlpilot: You'll also want to check out latextemplates.com and the Legrand Orange book. 20:18
[Coke] huh. vi was using 70% of my cpu 21:01
timotimo benchable6: compare HEAD my $a; (^9 X ^9).map( -> ($x, $y) { $a++ }); say $a; ||| my $a; (^9 X ^9).flat.map(-> $x, $y { $a++ }); say $a; 21:01
benchable6 timotimo, starting to benchmark the 1 given commit
timotimo, gist.github.com/99caf769b93eb6ccd2...7bbe278598
geekosaur closed a terminal window with it running? vim likes to go into a tailspin when stdin goes away unexpectedly
(it's *supposed* catch it, save the edit buffer for -r, and exit) 21:02
timotimo benchable6: compare HEAD my $a; (^90 X ^90).map( -> ($x, $y) { $a++ }); say $a; ||| my $a; (^90 X ^90).flat.map(-> $x, $y { $a++ }); say $a;
benchable6 timotimo, starting to benchmark the 1 given commit
timotimo, gist.github.com/25788b01599b86f109...18f7f2a16d
timotimo .tell scimon it might not be worth mutch, but (^9 X ^9).map(-> ($x, $y) { }) is noticably slower than (^9 X ^9).flat.map(-> $x, $y { }): gist.github.com/25788b01599b86f109...18f7f2a16d 21:04
yoleaux timotimo: I'll pass your message to scimon.
mspo [Coke]: coloring a big file? 21:27
Zoffix What precisely does :node($/) do in QAST? 23:37
Like in QAST::Var.new(:node($/), :name<Pair>, :scope<lexical>), 23:38
Zoffix ZofBot: *crickets* 23:43
ZofBot Zoffix, 'What do you mean?' 'Oh, come,' protested Jelliffe; 'there's no need to keep it up with me
thou Zoffix, what's the expected process for simple bug fixes like rakudo/rakudo#1307 ? Is there a regular process that ensures it'll get applied, or do I need to file a bug report or do something else? 23:45
synopsebot RAKUDO#1307 [open]: github.com/rakudo/rakudo/pull/1307 Drop extraneous spaces in code and table blocks
Zoffix thou: no strict process I'm aware of. Core devs usually just merge whenever they see notifications. Personally, I usually ignore codebase areas I'm unfamiliar with (like Pod). You could join #perl6-dev and remind about the PR awaiting to be merged once in a while, in hopes someone who knows what's up would merge :) 23:47
thou OK, thanks. I just didn't want to ignore an important step in the process if there is one. 23:48
MasterDuke thou: i think AlexDaniel has merged the last couple pod related PRs 23:51
thou Thanks 23:52
This one just fixes an annoyance I've had for probably 5 years.

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!