»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
discord6 <RaycatWhoDat> so, uh 00:34
AlexDaniel ya? 00:35
discord6 <RaycatWhoDat> I've never made a language or a compiler before
<RaycatWhoDat> but I think I have to at least make a toy one
<RaycatWhoDat> To that end, should I use NQP or just regular Perl 6
<RaycatWhoDat> or is NQP something I shouldn't be worrying about? 00:37
Xliff RayCatWhoDat: I would do it in Perl6. Convert it to nqp when you have something working.
Take advantage of the higher level constructs while you are working out the kinks.
discord6 <RaycatWhoDat> What's the benefit of converting it to NQP later on? 00:38
Xliff Speed.
discord6 <RaycatWhoDat> On a scale of 1 to zoom, how much speed?
Xliff Probably 3-8x? Depending on what you're doing.
That's on average.
discord6 <RaycatWhoDat> Interesting. Well, let's see if I can't get that going. Lemme check if there's a Lisp dialect using Perl 6. 00:39
Xliff Oh. That's spelled 'nqp'. 00:40
If you are more familiar with Lisp, then I take back everything I've said. :)
discord6 <RaycatWhoDat> o
<RaycatWhoDat> Good talk, then. 😛
Xliff Yeah. I probably should flee now. 00:41
Xliff flees
discord6 <RaycatWhoDat> Ehh, I'll still do what you said 00:41
<RaycatWhoDat> I need to learn Perl 6 anyway
<RaycatWhoDat> And grammars looks sweet
<RaycatWhoDat> look* 00:42
discord6 <Aearnus> learning nqp before Perl 6 is like learning JVM bytecode before learning Java 01:06
<RaycatWhoDat> > NOTE: there's no end-user support for NQP and the behaviour can change without notice. It's a tool for writing Perl 6 compilers, not a low-level module for Perl 6 programmers. 01:08
<RaycatWhoDat> Hmm. Maybe I shouldn't use NQP.
AlexDaniel yeah, you shouldn't 01:17
not for this purpose for sure :) 01:18
or at least, not yet :)
vrurg AlexDaniel: I would be happy if at some point Perl6 can replace NQP completely. Should be possible, I guess. 01:33
AlexDaniel vrurg: I have no idea how we can get there 01:40
and if it even makes sense
vrurg AlexDaniel: neither do I. But it doesn't hurt to dream about. ;) 01:41
It makes sense from the perspective of totally unified environment. It's not always easy to predict what NQP can and what it can't. 01:42
discord6 <RaycatWhoDat> uhhh 01:52
<RaycatWhoDat> Question
<RaycatWhoDat> Belay that. I think I know the answer. 01:53
AlexDaniel what was the question? 01:58
:)
vrurg "May ask you a qustion?" - " Yes!" – "Thank you, I already did!" 01:59
discord6 <RaycatWhoDat> I was trying to figure out what to do once the grammar parses my string. 02:01
<RaycatWhoDat> Is it safe to consider it playing the role of both parser and lexer?
<RaycatWhoDat> Or would it just be considered a parser? 02:02
AlexDaniel yes, and you use Actions to build the Abstract Syntax Tree
discord6 <RaycatWhoDat> dangit 02:04
<RaycatWhoDat> why did I decide to start this tonight
AlexDaniel why not? It's fun
AlexDaniel also it's useful to know how to do this stuff for small things 02:04
discord6 <RaycatWhoDat> True. 02:06
vrurg At least somebody does some really interesting stuff. I'm bored to death polishing makefiles... ;) 02:12
AlexDaniel vrurg: at least we're all grateful for that 02:22
unless you break stuff :D 02:23
vrurg Who breaks? Me? I would never ever! :p
AlexDaniel right right right right
riiiiiight
:)
vrurg is a saint! 02:24
BTW, did you see I made a PR for Hash -> Map coercion with deconting?
Should releive you of backward compat concern. 02:25
AlexDaniel I looked at it, it raised my bloat concern a bit 02:26
but it's probably ok :)
vrurg: speaking of bloat, there's a lot of code in nqp-configure, do we actually need all that?
vrurg AlexDaniel: That's my concern too, but I wanna see what lizmat and jnthn say. Fast and simpler solution would be explicit deconting in Hash::Map method 02:28
AlexDaniel: Most of it. I need patrickb to talk about some cleanups, but he's busy until October. 02:29
AlexDaniel in nqp-configure there's a perl5 file that has 1500 lines, and another one with 1000. Stuff in nqp is much shorter, at 400 freaking lines… :)
vrurg AlexDaniel: remember, it's a merge of two Configure.pl from NQP and Rakudo. Basically, NQP::Config is what you had in common in them. 02:31
AlexDaniel I haven't looked at it closely, so maybe me freaking out is unjustified, but most projects are not of that size and this thing is just for configuring rakudo and nqp…
yeah I understand that it's much better now :)
vrurg NQP::Macros – well, this one is no way to be shorter. Cross-make compatibility requires a lot of macros.
I didn't mean it's better or worse. Just mean that if it gets shorter then no more than 100-200 lines. 02:32
vrurg is thinking about learning cmake and switching to it.
AlexDaniel oh that's gonna make some people scream very loudly :D 02:33
vrurg Why? 02:34
If there're objections – well, it'd spare me a lot of work hours. I'd feel good about it. ;) 02:35
AlexDaniel I don't have any, people just love to hate cmake
vrurg Perhaps it worth a topic in problem-solving then. But, anyway, I would like to relax on fixing a couple of tickets first. 02:37
AlexDaniel sure, cool
vrurg This makefile unification sucket me dry.
*sucked
AlexDaniel vrurg: so what is make.nqp doing? For example, there's Makefile::Grammar, what is that for? 02:38
vrurg AlexDaniel: it is supposed to replace make altogether. But it's useless until we get precompiled moar for as many platforms as possible. Or otherwise there is no way to just run Configure.pl in rakudo and get things done. 02:40
AlexDaniel vrurg: how is precompiled moar going to help? 02:43
vrurg AlexDaniel: how would we run make.nqp to build NQP itself? There is an chicken/egg problem.
There was a discussion with timotimo back in May or June here, on the channel. 02:44
AlexDaniel vrurg: so let's say you have moarvm compiled separately, would you be able to run make.nqp this way?
vrurg Though it was probably related to switching Configure to NQP. Not really sure for now. 02:45
AlexDaniel I thought no because you don't have nqp built yet
right?
vrurg AlexDaniel: wrong. We have stage0 precompiled nqp in nqp. We can use it.
Ok, time to go. Let's continue tomorrow! 02:46
o/
AlexDaniel uh, right… and how does that fly with distros like debian?
you can't have a blob like that, do they create it from scratch somehow? 02:47
robertle: ↑ so what's the policy? I thought you can't have things like this: salsa.debian.org/perl6-team/nqp/bl...nqp.moarvm 02:50
irced i see in the perl6 faq that perl6 runs on a variety of backends. currently MoarVM & JVM, JavaScript in deelopment. what's this about JS? 03:02
AlexDaniel irced: perl6.github.io/6pad/ 03:03
irced takes a look. 03:04
irced interesting, thanks for highlighting 6pad 03:06
irced in the docs for Promise there is sample code as follows for the method in: 03:46
my $proc = Proc::Async.new('perl6', '-e', 'sleep 10; warn "end"');
my $result = await Promise.anyof( my $promise = $proc.start, Promise.in(5).then: { #..snip .then: { $promise.result } 03:47
my $result = await Promise.anyof( my $promise = $proc.start, Promise.in(5).then: { #..snip}).then: { $promise.result }
oh wait, i just got the answer.
robertle AlexDaniel: you are correct, that looks quite wrong :) 05:39
SmokeMachine .tell RaycatWhoDat if you wanna see a example of a simple toy language: github.com/FCO/bernaLang 06:10
tellable6 SmokeMachine, I'll pass your message to RaycatWhoDat
Geth ¦ problem-solving: AlexDaniel assigned to jnthn Issue stage0 is a binary blob, how can we do without it? github.com/perl6/problem-solving/issues/100 06:54
SmokeMachine Xliff: I think having it would be great! :) how is it going? 07:06
tellable6 SmokeMachine, I'll pass your message to Xliff
SmokeMachine irced: would you like to open a PR fixing the `$post1` issue (github.com/FCO/Red/issues/365)? 07:18
abraxxa sena_kun: good morning! 07:22
sena_kun: your fix resulted in the error message (HANDLED) Failed to open file /home/username/perl6/docs/ldap.asn: No such file or directory
sena_kun abraxxa, oh noes... well, that's not the fix itself, rather the update revealed a particular bug that was there for a long time. sigh. you can try to `zef uninstall Cro::LDAP` and then `zef install Cro::LDAP` 07:26
sena_kun needs to think about how to prevent `no such file` error
abraxxa how does it happen? 07:28
No candidates found matching identity: Cro::LDAP 07:29
sena_kun abraxxa, well, currently ASN::META depends on particular files being accessible during module usage... modules that are using it (currently Cro::LDAP) have those files at installation, but not at precompilation, so it fails
abraxxa zef update doesn't help either
sena_kun abraxxa, sorry, forgot it's not in the ecosystem, you need to update && use a local clone again 07:30
abraxxa sena_kun: I forgot! thanks ;) 07:37
sena_kun also, I was looking at the paged control RFC and it is pretty easy to implement, but then I looked at my overall approach to API for controls and thought "It is somehow less than awesome and inconsistent, we need to do better", so it is on my todo list too, probably after $dayjob. :) 07:38
abraxxa sena_kun: great, thanks! 08:31
sena_kun: I pasted the requested output in the github issue
El_Che_ lo 09:00
El_Che sena_kun: if you're added controls, the "relax" (-E relax) is a nice to have 09:01
wamba m: my $turtle = (1, 2 ... *);;my $hare = $turtle.clone;say $turtle.clone.head(2);say $hare.head(2);$turtle.head(2).say 09:51
evalable6 (1 2)
(3 4)
(5 6)
wamba m: my $turtle = (1, 2 ... *); $turtle.say;my $hare = $turtle.clone;say $turtle.clone.head(2);say $hare.head(2);$turtle.head(2).say
evalable6 (...)
(1 2)
(1 2)
(1 2)
wamba How "clone" on "Seq" should be work? 10:02
jnthn It arguably could throw an exception, given there's very little chance of it going well. :) 10:05
tellable6 2019-09-07T13:00:03Z #moarvm <brrt> jnthn presumably with a WinMain, I'd think
jnthn But cloning is shallow in general, so the two Seqs will share the same underlying iterator. 10:06
jnthn Alternatively we could actually define the behavior it currently has. :) 10:08
(I'm quite sure that today it's undefined behavior.)
SmokeMachine Xliff: any luck with the new driver? 10:12
tellable6 SmokeMachine, I'll pass your message to Xliff
SmokeMachine Xliff: was CommonSQL enough? 10:13
tellable6 SmokeMachine, I'll pass your message to Xliff
SmokeMachine sena_kun: what about adding a submodule on `docs/` pointing to the wiki repo? make sense? 10:23
tellable6 SmokeMachine, I'll pass your message to sena_kun
SmokeMachine sena_kun: or maybe make the entire `docs/` be the submodule. That way, when generating the docs (if you push it) it will be on the wiki as well 10:25
tellable6 SmokeMachine, I'll pass your message to sena_kun
pmurias vrurg: rakudo.js build for me, thanks! 10:51
tellable6 2019-09-09T02:42:55Z #perl6-dev <vrurg> pmurias Please, check my latest merge. I made it compile perl6.js and copy it into rakudo.js for back compat purpose. Otherwise things work on my macos and even passes most core tests.
daxim weekly, hsivonen.fi/string-length/ 10:57
weekly: hsivonen.fi/string-length/
notable6 daxim, Noted! (weekly)
number6` hey, everyone. i have a question. why do you have to enclose multiple variables in parens when declaring them, as in my ($x, $y);, but can send multiple arguments to a function without using parens, as in say "the answer is ", 42; 12:47
timotimo sub calls are "list prefix" precedence, i.e. what comes after them is parsed as a list that belongs to the prefix 12:51
i'm not sure why exactly the way "my" works was chosen, but I can imagine having to put parens makes it clear what happens in the case of "my Int $x, $y" (having to write it as "my Int ($x, $y)" 12:52
jnthn Well, `my $x, $y` is actually fine, it just doesn't declare $y, because of precedence 12:53
timotimo right, i parsed the question as presupposing that $y was also meant to be declared
number6` timotimo: jnthn: thank you. 12:55
lizmat perl Configure.pl --gen-moar --gen-nqp --make-install 12:56
results locally for me in:
Missing or wrong version of dependency 'src/Perl6/World.nqp' (from 'gen/moar/CORE.c.setting') 12:57
is this a known issue? or is did I somehow mess up my local install?
timotimo huh, shouldn't be related to the git submodules with the build system stuff, right? 13:00
lizmat is not sure 13:01
timotimo if so, "git submodule update" could help
but i think some setting is active that makes that unnecessary or something? 13:02
lizmat just done that, no action, no error message
timotimo OK
what about a "make clean"? 13:03
lizmat same result 13:08
vrurg lizmat: can you try on with --gen-nqp=master? 13:11
lizmat: I think I know what happens. Do you CORE.c.setting.moarvm in you source root? COREs are now being compiled into blib/ but ModuleLoader may still try loading them from ./ while building. 13:16
lizmat $ ls -ls CORE.c.setting.moarvm 13:17
28576 -rw-r--r-- 1 liz staff 14629864 Sep 9 15:10 CORE.c.setting.moarvm
which implies to me that it is from just now?
aka, it gets placed in the wrong spot? 13:18
vrurg lizmat: try git clean -dfx unless there is no changes you'd like to preserve. And then recompile. 13:19
unless there're changes, actually. 13:20
timotimo not only changes 13:25
every file in there
that's not tracked
i happen to have a *lot* of stuff in my rakudo folder 13:26
SmokeMachine was anyone able to make gh-pages correctly highlight perl6 code? 13:27
lizmat vrurg: well, seems like after that, the build succeeded 13:31
timotimo i hope you didn't lose anything important :S
lizmat there's always time machine :-) 13:32
timotimo good
vrurg timotimo: basically, -dfx is overkill, just to make sure. Should be enough to wipe CORE.*.moarvm from the source root. 13:33
vrurg would as well move perl6*.moarvm from source root too for the purpose of easing cleanups, especially manual ones. 13:34
[Coke] m: say 3 14:11
evalable6 3
[Coke] m: while 1==2 {␤FIRST my $a=3;␤}␤ 14:12
evalable6 WARNINGS for /tmp/X4j1KJhdC3:
Useless use of LOOP_BLOCK_1 symbol in sink context (line 1)
[Coke] ^^ what is the uesless use there?
(golfed down from a larger program exhibiting the same issue)
timotimo m: while False { } 14:13
evalable6
timotimo m: while False { FIRST 1 }
evalable6 WARNINGS for /tmp/EYVNhdKtbO:
Useless use of LOOP_BLOCK_1 symbol in sink context (line 1)
timotimo m: while True { FIRST 1 }
evalable6 (signal SIGHUP) WARNINGS for /tmp/7bqzZ5kbiS:
«timed out after 10 seconds»
[Coke] the original had variables in the conditional.
timotimo ok, so it wouldn't know to dead-code-eliminate 14:14
Xliff m: while 1==2 {␤FIRST my $a=3; $a;}
evalable6 WARNINGS for /tmp/v5SFbEf33F:
Useless use of LOOP_BLOCK_1 symbol in sink context (line 1)
tellable6 2019-09-09T07:06:10Z #perl6 <SmokeMachine> Xliff: I think having it would be great! :) how is it going?
2019-09-09T10:12:16Z #perl6 <SmokeMachine> Xliff: any luck with the new driver?
2019-09-09T10:13:08Z #perl6 <SmokeMachine> Xliff: was CommonSQL enough?
[Coke] m: my $a=0; my $b = 3 while $a < $b {␤FIRST my $a=3;␤$a=$b}␤
evalable6 (exit code 1) 04===SORRY!04=== Error while compiling /tmp/DaPOQpPW8z
Unexpec…
[Coke], Full output: gist.github.com/e06d5cccc6832e582f...13851a2e98
[Coke] m: my $a=0; my $b = 3; while $a < $b {␤FIRST my $a=3;␤$a=$b}␤
Xliff .tell SmokeMachine Busy day, today. I will get back to you with questions. tl;dr -- Didn't work. Got errors. :(
evalable6 (signal SIGHUP) WARNINGS for /tmp/MOeGpb2RKg:
«timed out after 10 seconds»
tellable6 Xliff, I'll pass your message to SmokeMachine
[Coke] m: my $a=0; my $b = 3; while $a < $b {␤FIRST $a=3;␤}␤ 14:15
evalable6 WARNINGS for /tmp/6JQv4OXBLD:
Useless use of LOOP_BLOCK_1 symbol in sink context (line 1)
[Coke] ^^
SmokeMachine Xliff: :(
Xliff Oh! There you are!
SmokeMachine :) o/
Xliff One sec 14:16
timotimo ah dangit, i need to find a code example that exhibits interesting behavior for heap snapshots, but doesn't generate tens of gigabytes of snapshots in the process 14:17
the heapanalyzer itself, when reading a medium-sized snapshot, will generate tens of gigabytes of data, for example 14:18
the confprog compiler just has four snapshots in it that are all very similar
Xliff SmokeMachine: Unfortunately I am in the Proper OS. I guess I could install MS SQL server for Linux....
However yesterdays excursions have convinced me that once is enough. :) 14:19
If I were getting paid, however....
SmokeMachine Xliff: what were the problems you got? 14:24
Xliff Actually, I can't remember. It was a detailed error message that I forgot to save. I'd have to reboot to get it, and it's going to be a busy day, today. 14:25
Probably won't be back on that OS until this afternoon or evening. I'm EDT
SmokeMachine Xliff: no problem! let's see it when you have some time... 14:29
SmokeMachine m: (^10).map({ $_ if $_ %% 2 }).map: &say; 15:05
evalable6 0
2
4
6
8
SmokeMachine m: Supply.from-list(^10).map({ $_ if $_ %% 2 }).tap: &say # wouldn't it make sense to have the same behaviour?
evalable6 0
()
2
()
4
()
6
()
8
()
SmokeMachine m: Supply.from-list(^10).map({ |($_, $_+1, $_+2) }).tap: &say 15:13
evalable6 (0 1 2)
(1 2 3)
(2 3 4)
(3 4 5)
(4 5 6)
(5 6 7)
(6 7 8)
(7 8 9)
(8 9 10)
(9 10 11)
SmokeMachine m: (^10).map({ |($_, $_+1, $_+2) }).map: &say;
evalable6 0
1
2
1
2
3
2
3
4
3
4
5
4
5
6
5
6
7
6
7
8
7
8
9
8
9
10
9
10
11
SmokeMachine shouldn't a Supply.map returning a Slip iterate over it and call `emit()` for each item? 15:16
m: Supply.from-list(^10).map({ $_ if $_ %% 2 }).tap: { .^name.say } 15:17
evalable6 Int
Slip
Int
Slip
Int
Slip
Int
Slip
Int
Slip
SmokeMachine would it work if here (github.com/rakudo/rakudo/blob/3f69....pm6#L323) I change for `[result].map: &emit` ? 15:24
jnthn SmokeMachine: Work maybe, but it'll result in a crazy slowdown 15:31
SmokeMachine: If we want it to work, I'd do it by adding another `multi` candidate `emit(Slip:D)` 15:32
SmokeMachine jnthn: makes sense... and do we want it to work? 15:33
jnthn SmokeMachine: I can't think of any harm that we'd get from it working, and it's probably useful. 15:35
SmokeMachine jnthn: and that makes both `map`s consistent... 15:36
jnthn Well, but one has to be careful making consistency arguments, because for example .flat on a Supply is about flattening the inner Supply values, not any inner List values. 15:37
A List of Supply objects treats the Supply values inside of it as just items, and a Supply emitting List values does similar 15:38
That's why I'm a little wary about letting Slip have any magical behavior here
mayuresh hello. :) 15:41
anyone around?
jmerelo mayuresh: at least one :-)
mayuresh just dropped in to inquire about the status os 'mop' in p6
i meant; "status of 'mop'" :)
jmerelo mayuresh: there's a whole lot of that. Perl 6 includes a metamodel which allows you to build everything, not only objects. Check this out: docs.perl6.org/type/Metamodel::ClassHOW 15:42
mayuresh wow, that's awesome. will get to reading it now. thanks for sharing it. :) 15:43
bye.
jmerelo mayuresh: good luck :-)
SmokeMachine jnthn: imho it would make sense if we have a equivalent of a slip to supplies (as s supply is the equivalent of a list...) 15:45
SmokeMachine I mean a auto-flatten supply... 15:46
jnthn Maybe, but the downside is that it'd not handle the use case you care about in the example (a map with an if) 15:49
SmokeMachine jnthn: Yes, that's why I'd prefer to let the Slip be the Supply's Slip... 15:50
[Coke] "Supply-Side Slip-onomics." 15:54
SmokeMachine [Coke]: Sorry, it's too advanced to my poor English... what does that mean? 15:58
[Coke] an unfunny plan on en.wikipedia.org/wiki/Supply-side_economics 15:59
*play
jmerelo [Coke]: it was mildly funny :-)
jnthn I was amused :P
SmokeMachine Should I open a issue on problem-solve or just try to implement it? 16:01
jnthn SmokeMachine: I think I'd be happy with a PR adding the multi candidate and some tests; if anybody objects, we can go through the full process. 16:04
SmokeMachine jnthn: ok, I'll try to do that! :)
sarna hi, can I somehow express that a function should take either an int or a string with some content? like, sub(param: int | "foo") 16:46
sarna I'm writing some bindings to JS in another language and I've been wondering if that's possible in raku :) 16:47
Elronnd I would make it a multimethod 16:48
sarna I know it can be done in tens of better ways, but it's this particular one I'm curious about :D 16:50
Guest93 sarna: You could also sub ($x where Int | "foo")
sarna Guest93: that works? cool! 16:51
would it work on a class field as well?
Guest93 sarna: yeah, it should 16:54
sarna Guest93: sweet! thanks :)
SmokeMachine m: sub a($b where Int|Str) { say $b.^name }; a 42; a “bla”; a 3.14 16:56
evalable6 (exit code 1) Int
Constraint type check failed in binding to parameter '$b'; expected anonymous constraint to be met but got Rat (3.14)
in sub a at /tmp/4uoPHG6uoP line 1
in block <unit> at /tmp/4uoPHG6uoP line 1

Str
El_Che like Elronnd says, a multimethod is the most elegant (readable and safe) in my view 16:57
SmokeMachine m: sub a($b where Int|Str) { say $b.^name }; a 42; a “bla”
evalable6 Int
Str
SmokeMachine m: sub a($b where Int|Str) { say $b.^name }; a 3.14 16:58
evalable6 (exit code 1) Constraint type check failed in binding to parameter '$b'; expected anonymous constraint to be met but got Rat (3.14)
in sub a at /tmp/53p4cofckJ line 1
in block <unit> at /tmp/53p4cofckJ line 1
Xliff my $a = 'b'; my $c = 'd'; my @e = «$a $c>; @e.gist.say 17:59
m: my $a = 'b'; my $c = 'd'; my @e = «$a $c>; @e.gist.say
evalable6 (exit code 1) 04===SORRY!04=== Error while compiling /tmp/QgLBF2clCR
Cannot …
Xliff, Full output: gist.github.com/e19f19e0916d7db72f...22181f450c
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a $c»; @e.gist.say
evalable6 [b d]
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a "$a $c" $c»; @e.gist.say 18:01
evalable6 [b b d d]
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a "$a $c" $c»; @e.gist.say; @e.elems.say
evalable6 [b b d d]
3
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a "$a $c" $c {$a}{$c}.c»; @e.gist.say; @e.elems.say 18:02
evalable6 [b b d d b d .c]
6
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a "$a $c" $c $a$c.c»; @e.gist.say; @e.elems.say
evalable6 [b b d d b d .c]
6
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a "$a $c" $c "{$a}{$c}".c»; @e.gist.say; @e.elems.say 18:03
evalable6 [b b d d bd .c]
5
Xliff m: my $a = 'b'; my $c = 'd'; my @e = «$a "$a $c" $c "{$a}{$c}.c"»; @e.gist.say; @e.elems.say
evalable6 [b b d d bd.c]
4
Xliff m: my @a = ('a'..'f'); @a.map( "{*}.ext" ).say 18:45
evalable6 (exit code 1) Cannot map a Array using '"*.ext"'
Did a * (Whatever) get absorbed by a list?
in block <unit> at /tmp/vxiFCgwkBN line 1
Xliff m: my @a = ('a'..'f'); @a.map( * ~".ext" ).say
evalable6 (a.ext b.ext c.ext d.ext e.ext f.ext)
patrickb .tell Xliff re DBIish and MSSQL using DBDish::ODBC. I use that in a production system at work 19:02
tellable6 patrickb, I'll pass your message to Xliff 19:03
patrickb Prepared statements don't work. And are not trivial to implement.
tellable6 2019-09-04T01:41:40Z #perl6-dev <vrurg> patrickb Could you, pls, review and test R#3166?
synopsebot R#3166 [closed]: github.com/rakudo/rakudo/pull/3166 [WIP][build] [WIP] [HEADSUP] Extract common parts of backend makefiles into a common template
patrickb .tell Xliff Prepared statements do not work and are not trivial to implement. 19:04
tellable6 patrickb, I'll pass your message to Xliff
patrickb .tell Xliff there is also a PR waiting for merge that fixes some column types github.com/salortiz/DBDish-ODBC/pull/5 19:07
tellable6 patrickb, I'll pass your message to Xliff
patrickb .tell salortiz did you notice github.com/salortiz/DBDish-ODBC/pull/5 19:08
tellable6 patrickb, I haven't seen salortiz around, did you mean sortiz?
patrickb .tell sortiz Did you notice github.com/salortiz/DBDish-ODBC/pull/5 19:09
tellable6 patrickb, I'll pass your message to sortiz
patrickb and with that I'm off again. o/
sortiz .tell patrickb I merge your PR, sorry for the delay. 19:42
tellable6 sortiz, I'll pass your message to patrickb
Kaiepi what was that nqp branch that was supposed to make it easier to debug when something's going awry with precomp? 20:31
Geth doc: 7d6fe10965 | Coke++ | type-graph.txt
UInt is a subset, not a class
 21:21
doc: d2e000d337 | Coke++ | type-graph.txt
have to mark enums as Int
lizmat and another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2019/09/09/...repidated/ 21:34
guifa lizmat++ 21:34
lichtkind lizmat++ 22:44
guifa Ack, refactored my code to use a dynamic variable and when I turned back on a divide and conquer optimization it bombed because dynamic variables only survive a single spin off 22:52
SmokeMachine github.com/FCO/Red/wiki 23:45
wrong link 23:50
fco.github.io/Red/
AlexDaniel lizmat++ # great weekly! 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!