»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
[Coke] waves from... Iowa 00:18
colomon Writing lib/Math/ThreeD/Vec3.pm...done 00:38
Writing lib/Math/ThreeD/Mat44.pm...Segmentation fault (core dumped)
(that's in a parrakudo smoke)
lucas Hello, I made a little toy to help me run my implementations: 02:13
gist.github.com/lucasbuchala/a12a0...9920fc4369
dalek rl6-bench: 8dc1c34 | (Geoffrey Broadwell)++ | minibenchmarks.pl:
Reduce initial SCALE for rc-self-describing-numbers

Rakudo handles this benchmark so slowly that this one test was easily dominating testing time.		 02:58
rl6-bench: 8b08ec7 | (Geoffrey Broadwell)++ | / (4 files):
Add rc-dragon-curve minibenchmark
rl6-bench: 6547ae8 | (Geoffrey Broadwell)++ | / (4 files):
Add rc-9-billion-names minibenchmark
dalek t--IRC: fb2cb3f | (Jarrod Funnell)++ | lib/Net/IRC/ (6 files):
Merge the fixes from my commandhandler branch
 04:52
t--IRC/new-handler: c24efc1 | (Jarrod Funnell)++ | lib/Net/IRC/ (3 files):
Add events for quit/part. Clean up defaulthandler's user tracking
atroxaper Good morning, #perl6 ! 05:16
donaldh o/ atroxaper 05:20
atroxaper donaldh: o/
I heed to use some string-template system. I mean 'Insert to string value by <%= key %>'. I think I can use some regex to find <%= ... %>. But how can I find all that pieces? 05:24
moritz atroxaper: how about using one of the several template engiens found on modules.perl6.org? 05:25
or else, use a global match 05:26
m:g/.../, or $str.match(:g, $regex)
atroxaper Yes! Global search. Thank you. 05:27
I didn't think to find existent template system. But I'm going to look at them. 05:28
atroxaper My panda didn't install module for Rakudo Parrot, but did for MoarVM and JVM? 05:46
I have perl6 like perl6-m 05:48
timotimo check that your $PATH is pointing at the correct perl6 perhaps? 06:21
and the correct panda binary as well; especially try head (which panda)
atroxaper $PATH is pointing to all perl6-* and perl6. perl6 -> perl6-m 06:23
timotimo and the hashbang of the panda command? 06:26
atroxaper #!/usr/bin/env perl6 06:28
timotimo can youexplain your problem a bit more?
atroxaper I have installed Template::Mojo by panda install. And now if I write 'use Template::Mojo' and run per6-p I see that perl6 can't find that module. But perl6-m and -j works well. 06:30
timotimo er ... panda should install it for only one of your perl6s
atroxaper And how can I install it for another? 06:31
timotimo the best approach is probably using rakudobrew, you can "rakudobrew switch" very easily
atroxaper Really? Where can I read about that? 06:32
timotimo hmm, good question; the repository with it is in tadzik/rakudobrew
atroxaper Yes. I found it. Reading ;) 06:33
timotimo we should quite possibly propagate information about rakudobrew closer to perl6.org and rakudo.org 06:34
atroxaper It would be great! 06:35
moritz possibly a line on modules.perl6.org 07:13
and rakudo.org, yes
maybe also mention it in rakudo's README
tadzik blushes 08:07
FROGGS hi tadzik 08:08
atroxaper I'he tried to install Rakudos with rakudobrew and I think it would be better to clone repositories with flag --depth, isn't it? 09:01
FROGGS tadzik: that is for you --^ 09:02
tadzik oh, maybe 09:14
I never saw a problem with not doing that 09:15
moritz it just downloads more stuff
you just can't do much with a shallow clone (a bit better in newer git versions) 09:16
atroxaper tadzik: moritz: I just clone nqp three times :) And i thought that using --depth can allow to install all three backends faster in first time. 09:19
And that change don't require lots modification in rakudobrew. 09:21
atroxaper tadzik: btw how can I uninstall panda? I have installed in by bootstrap. And now I install all by rakudobrew but panda runs from previous path. 09:23
tadzik atroxaper: oh, manually :/ 09:24
remove .rakudobrew/<backend>/panda probably
atroxaper tadzik: no. Panda that I've installed by bootstrap.pl ) 09:25
moritz atroxaper: maybe --reference <path-to-existing-repo> would be even faster
git clone --reference <path> # that one
tadzik ah 09:26
tadzik atroxaper: well, it got installed to wherever your rakudo was, probably 09:26
'which panda' will tell you
moritz it then obtains objects from the reference repo
atroxaper moritz: yes! It would be best! 09:27
tadzik: if I remove panda from that path (which panda) then '> panda' said that can't find panda) Maybe there is some system variable with path to perl6? 09:32
mst WHY WAS I NOT ALREADY IN HERE 09:33
tadzik I don't know a thing about your system variables :)
mst pokes his IRC config with a stick
tadzik poke it with a large trout
mst I already am a large trout, evidently that didn't help
tadzik troo 09:34
atroxaper tadzik: oh... sorry. I commented a part of my PATH variable in bash_profile. But forget about " ". Thank you :) 09:36
tadzik :) You're welcome 09:38
FROGGS mst: you missed all the fun in here :o)
mst FROGGS: I owe Liz and Wendy some cpantesters work 09:39
FROGGS ohh, nice!
mst FROGGS: I did Rakudo::Star on CPAN, now it's time to do it again, properly 09:40
FROGGS that reminds me that I wanted to write/port CPAN::Report::Metabase::Mumble in the very near future
mst FROGGS: also, Wendy stranded me in Schiphol for twenty four hours so I believe she volunteered to play secretary in response; she'll find that out tomorrow :)
FROGGS *g*
mst: so is that about smoking the rakudo installation process? 09:41
mst FROGGS: and also making it installable as a cpan module 09:42
FROGGS ohh, even nicer :o)
mst I did it with rakudo already
just nobody doing release management ever used my work
so this time, I redo it with instructions for stupid people 09:43
(yes, I'm being an asshole, I'm drunk and tired :)
FROGGS :o) 09:43
your accent compensates all that :o) 09:44
(at least according to $wife)
FROGGS hmmm, there is stuff out there I didn't know: metacpan.org/source/MAGNOLIA/Enbld...udostar.pm 09:46
mst I had to remove the old Rakudo::Star because it turns out it had package statements in common with core perl and was breaking the indices ;) 09:49
FROGGS ewww 09:55
atroxaper tadzik++ for rakudobrew 10:27
woolfy I stranded mst on schiphol for 24 hours? I seem to be a very bad person... In that case, I maybe should indeed play secretary for mst... 10:28
Of course I have no idea wht mst means. hen? When? WHEN??? 10:29
dalek t--IRC/new-handler: d413f30 | (Jarrod Funnell)++ | lib/Net/IRC/Event.pm:
Whoops. s/ident/username/
 10:30
woolfy But he will tell me today, so I will be prepared to be ashamed. 10:38
lucas__ Hello channel! 12:39
FROGGS hi lucas__
lucas__ I'm trying to get my head around type constraints. Can you help me determine if the following statment is false or true? "You can only use type constraints in my/state/has declarations and in routine signatures; You can't use it with our and constant declarations and pointy blocks parameters." Did I get it right or wrong?
FROGGS m: -> Int $x { say $x }(42) 12:43
camelia rakudo-moar a7b4cb: OUTPUT«42␤»
FROGGS m: -> Int $x { say $x }("a")
camelia rakudo-moar a7b4cb: OUTPUT«Type check failed in binding $x; expected 'Int' but got 'Str'␤ in block at /tmp/P4VdNRTYAX:1␤␤»
FROGGS lucas__: so it works in pointies
FROGGS m: constant Int x = 42; 12:47
camelia rakudo-moar a7b4cb: OUTPUT«===SORRY!=== Error while compiling /tmp/9R9w25wpCc␤Missing initializer on constant declaration␤at /tmp/9R9w25wpCc:1␤------> constant Int ⏏x = 42;␤ expecting any of:␤ statement list␤ prefix…»
FROGGS okay, that makes sense 12:47
lucas__ FROGGS: it is good it work in pointy blocks :)
FROGGS "You can only use type constraints in my/state/has declarations and in routine signatures (including pointy blocks); You can't use it with our and constant declarations." 12:48
lucas__: pointy blocks are almost identical to subroutines
FROGGS the only difference is that a return falls through 12:48
lucas__ I was thinking more like "constant Int $x = 42;" But I agree it makes a strange sense to type a constant.
FROGGS m: constant $x = 42; 12:49
camelia ( no output )
FROGGS m: constant Int $x = 42;
camelia rakudo-moar a7b4cb: OUTPUT«===SORRY!=== Error while compiling /tmp/ZooPqXgHdm␤Missing initializer on constant declaration␤at /tmp/ZooPqXgHdm:1␤------> constant Int ⏏$x = 42;␤ expecting any of:␤ statement list␤ prefi…»
FROGGS it wants to parse a name (or variable), so a type that looks like a name confuses it
lucas__ yes, S03 says constants can't be used with signatures, because constant declarations already declare names in "type" space. 12:51
I guess only "our" scoped variable remains... 12:52
FROGGS timotimo: blog.yapcna.org/2014/06/23/yapc-is-...n-youtube/ 12:59
psch hi #perl6 13:12
brrt hi psch 13:17
psch hi brrt 13:20
brrt how is it? :-)
psch i'm poking at Actions.nqp to try and figure out how to make tr/// actually work 13:22
fwiu, in smart-match context tr/// has to set its lhs to the result of .trans with the tr/// arguments as par, but return the number of changed or deleted characters 13:23
so to compare the before and after i have to store the original lhs to the smart match somewhere; i think i have figured out how to do that now
brrt ok.. cool :-) 13:24
psch what are you up to? 13:25
brrt i'm working on a bug in moarvm 13:39
basically, 'inline and jit doesn't work together very well yet'
jnthn++ has already suggested the right solution 13:40
psch sounds a bit more involved than what i'm doing :) 13:44
brrt no, not really :-) 13:45
just another language :-)
psch i had had a look at something in moar recently, but got stuck in a spot where FROGGS got stuck as well; that combined with my lack of understand of moar and low confidence in my c knowledge made me put that aside 13:45
brrt what did you want to fix? maybe i can help out 13:46
psch that something was P6int without an explicit nativesize not having the right bitsize
brrt hmmm
psch i.e. on a 32bit machine 'my int $x' instantiates a P6int with 64bit, which is not to spec
brrt what would be the 'right' bitsize according to you?
hmm
psch or rather, it doesn't limit to 32bit
brrt i believe its according to moarvm implementation, yes :-) 13:47
in nqp, my int $x makes a true native int, and those are always 64 bit wide in moarvm, basically because all registers are 64 bits wide
yes, that is a pain point on 32 bits, no, we're very unlikely to ever fix that :-) 13:48
that is to say 13:49
i kind of know how to fix that
but i don't think we should
psch to be fair, the one point where i noticed it being a real issue is with NativeCall, i.e. the test suite tests non-specific int against long, which works on 64bit and fails on 32bit arch 13:50
brrt hmmm
psch but then, interfacing with c code without explicit bit sizes for int is probably a bit not-so-great anyway
brrt ok, i see
hmm, indeed 13:51
i don't actually know how nativecall works, but i kind of imagine it being a pain on 32 bit 13:52
what with the many different ABI's
FROGGS on the other hand you want to make it as easy as possible for the user
FROGGS so we need to make that tests pass on 32bit, without modifying the test 13:52
brrt hmmm 13:53
moritz do we have all of int, int32, int64 available for nativecall?
brrt i don't know about nativecall :-) 13:54
FROGGS moritz: yes
brrt does explicitly passing int32 make nativecall do the right thing? 13:55
i.e.
FROGGS the test is about unpacking a C struct with a long inside using a class that has an Int attribute
err, int
brrt: yes
brrt hmmm
FROGGS only on 32bit of course
brrt hhow... wait,
is long 32 bit on 32 bit platforms?
FROGGS for all to read: github.com/jnthn/zavolaj/issues/37 13:56
brrt: yes, according to Perl -V
brrt o
FROGGS ö
brrt hmm 13:57
i suppose that treating int as int32 depending on platform is correct
FROGGS yes, but trying to do it that way makes it explode 13:59
so, I guess I did something stupid or missed something
that is from my ubuntu 14.04 x86 VM: intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=12345678 14:01
my host: intsize=4, longsize=8, ptrsize=8, doublesize=8, byteorder=12345678
brrt hmmm 14:02
sorry, can't really help you right now 14:03
FROGGS LIVE: www.youtube.com/user/yapcna/featured 14:05
ARRRGGG! Live Streaming ist wegen Rechteproblemen in Deutschland leider nicht möglich.
psch FROGGS: yup :) 14:06
FROGGS >.<
that sucks 14:08
psch well there is proxtube
PerlJam someone should move a mic closer to TimToady or turn up the volume or something. 14:09
psch hm, there's no way for me to leverage the existing levenshtein implementation in World for comparing the tr/// result with the original input, is there? 14:11
FROGGS psch++ # love you! :D
psch FROGGS: you're welcome :)
newbie_Perl6_ Can I get information about Regex? (something in pseudocode, for example) 14:15
token alpha / <[a..z]> / -----> alpha.self = (Regex.CharacterClass, ).List
rule expr / <alpha>+ \+ <alpha>+ / -----> expr.self = (alpha.TokenPlus, '+'.Str, alpha.TokenPlus).List
rule array / '[' <digit>* %% ',' ']' / -----> array.self = ('['.Str, digit.SeperatorPlus(','), ']'.Str).List
moritz newbie_Perl6_: no, they aren't introspectable
cognominal__ too bad we can't see the slides. 14:16
newbie_Perl6_ I think it's useful~~~
newbie_Perl6 how about an introspectable regex? 14:20
moritz newbie_Perl6: if you prototype one, maybe it'll be specced 14:21
PerlJam newbie_Perl6: what you mentioned earlier looks more like the AST than what I would call introspection.
colomon TimToady is on youtube live! 14:23
www.youtube.com/watch?v=mxLIQxSda0E
newbie_Perl6 Hello PerlJam, what's a introspectable regex? some idea? 14:26
newbie_Perl6 Hello moritz. I just think it's a cool thing, but the implemetation is too hard for me. 14:29
hoelzro morning #perl6 14:31
moritz newbie_Perl6: you could try, and ask for help when you get stuck 14:32
PerlJam hoelzro: greetings
moritz newbie_Perl6: I also think it's cool, but it's way down my priorities list
psch hi hoelzro 14:33
lucas__ anyone watching larry live? what kind of vim script is he using? :D
colomon am watching, no idea about the vim
PerlJam TimToady's talk mentions several principles that we've seen before, but I get the feeling that an update (or perhaps an addendum) to the Perl 6 Design Principles document is in order
newbie_Perl6 moritz: I will try. 14:36
PerlJam newbie_Perl6++ trying is good. 14:37
newbie_Perl6 If someone interested in it too, please try it too.
PerlJam, some intereting idea? 14:38
hoelzro so I tried creating a custom RoleHOW to monitor the "closing" of a class, but it seems that specialize/compose are called on declaration, and add_method on the classHOW is called afterwards 14:39
hoelzro (I'm trying to create some postconditions for consumers of my role) 14:39
newbie_Perl6_ sorry, bad internet connection 14:40
hoelzro also, I couldn't figure out how to change the class/role HOW objects from within the same file; I had to create another file and use EXPORTHOW. If someone could guide me on how to do that, that would be great =) 14:44
colomon hmm, what to watch next from YAPC? 14:55
lucas__ colomon: thanks for sharing the link!
colomon you're welcome 14:56
cognominal__ TimToady decided not to unwrap the christmas presents today. These presents are future, or unevenly distributed. 15:02
FROGGS TimToady++ # anyway :o) 15:05
I enjoyed the talk 15:06
cognominal__ yes, it opened a window to his thought process. 15:07
jnthn For regex introspective stuff, my github.com/jnthn/grammar-generative does evil to actually capture the AST produced as they parse. 15:08
moritz .tell newbiew_Perl6 about github.com/jnthn/grammar-generative 15:09
yoleaux moritz: I'll pass your message to newbiew_Perl6.
jnthn I don't know if (a) the code still works, and (b) if it's comprehensible. :)
And (c) if I understand it any more :P
hoelzro: When I want to do stuff at compose time I just replace ClassHOW with something that has a method compose($obj { ...checks/stuff... callsame(); post-compose stuff } 15:12
With that subclassing ClassHOW of course.
hoelzro jnthn: right, but then users of my library would have to use that ClassHOW that I've built 15:13
dalek kudo-star-daily: e8442a6 | coke++ | log/ (13 files):
today (automated commit)
rl6-roast-data: 48aec67 | coke++ | / (5 files):
today (automated commit)
jnthn And...the other way they have to use the RoleHOW you built?
hoelzro jnthn: this is what I have in mind: gist.github.com/hoelzro/d552753137568ac5285c
well, apps consume the App::Subcommander role
so I have control over the latter
so I figured I could detect with a custom RoleHOW when the role is composed into a class 15:14
[Coke] rakudo moar tested clean today; jvm & parrot still have one failing test file.
hoelzro unfortunately, that happens too early for my checking code to make any sense
jnthn Oh.
hoelzro if you need more explanation, I'm happy to give it =) 15:15
hoelzro I would just like duplicate subcommands to be detected at compile time 15:15
jnthn I think I see the problem; need to look at the code to think of a solution...
(the metamodel code
hoelzro ah ha
jnthn )
hoelzro I spent a bit of time looking at it yesterday; I'm sure you'll be able to come up with something 15:16
I think that I may just have to use a ClassHOW and have users subclass App::Subcommander
jnthn Yeah. Not least 'cus I can extend the thing if it isn't up to your use case :P
hoelzro rather than make it a role
jnthn anyway, back to hotel & 15:17
TimToady m: my Int constant x = 42; say x 15:27
camelia rakudo-moar a7b4cb: OUTPUT«42␤»
TimToady m: constant x of Int = 42; say x
camelia rakudo-moar a7b4cb: OUTPUT«===SORRY!=== Error while compiling /tmp/kEdFtpLoDN␤Unexpected named parameter 'SYMBOL' passed␤at /tmp/kEdFtpLoDN:1␤------> ␤»
btyler TimToady++ # learned about z. in vim :)
TimToady std: constant x of Int = 42; say x
camelia std 0f2049c: OUTPUT«ok 00:01 122m␤»
TimToady rakudobug I guess 15:28
btyler: I'm glad someone learned something from it :)
I only covered about 15 of the 56 principles I had available, but I suppose it was unrealistic to think I could get to most of 'em 15:29
TimToady was lucky to just finish reading all the RFCs last night 15:31
colomon TimToady++
TimToady they're quite amazing in spots :) 15:32
both ++ly and --ly
masak good afternoon, #perl6 15:52
atroxaper masak: o/ 15:53
psch hi masak
jnthn masak: afternoon? It's pretty much evening :P 15:55
masak ...four minutes left of the afternoon :P 15:56
TimToady still antenoon :P
PerlJam Still another hour of morning here. :) 15:57
woolfy broke it! 15:58
TimToady ayup, woolfy++
testing-to-destruction++ 15:59
Ven is watching woolfy++ live 16:00
TimToady lunch & 16:01
dalek rl6-bench: 5a159b3 | (Geoffrey Broadwell)++ | minibenchmarks.pl:
Add curated tags for minibenchmarks
 16:04
Ven TimToady++'s talk was a bit more for p5ers 16:15
troydm hey all! I've just installed rakudo wth parrot and it takes 4-5 seconds just to start perl6 executable, why is it so slow? 16:16
colomon troydm: errr…. good question? I just tried it on my Linux box, and it was definitely under 1 second. 16:17
I expect JVM to be that slow, but not Parrot
troydm lpaste.net/530134410415046656 16:19
colomon troydm: thats .465 seconds "real" for me. what sort of machine are you running on? 16:20
Ven troydm: can you perl6 -v, please :P ?
troydm This is perl6 version 2014.04 built on parrot 6.1.0 revision 0
gtg be back in 30 mins 16:21
thou TimToady++ # Inspiring an incredibly fun community 16:35
atroxaper While using role a can define method with { ... }. But can I define method with ( ... ) and { ... } also? 16:45
jnthn As in, with a signature? 16:46
atroxaper I'm going to pass arguments my introspection to some method. I just want to force customer of role to implement some method with some arguments.
by introspection*
jnthn Oh 16:47
It doesn't try and enforce a signature match
At present, anyway.
masak TimToady++ # not giving up through X years of second system syndrome
atroxaper jnthn: ok. I will find another way maybe. 16:48
troydm so I'm back, guys why is perl6 parrot so slow on startup? 17:08
psch troydm: as colomon asked, what kind of machine are you running on? 17:10
timotimo perl6 parrot should start up fairly quickly, the times where it took a second are long gone
troydm psch: virtual machine on virtualbox 17:21
troydm psch: it takes 8 seconds to load perl6 on my virtual machine 17:21
psch i have a similar setup, running a debian vm on a win7 host; i get ~1.5s real for -e'say "Hello"' on parrot perl6 17:22
troydm and other languages/interpreters work fast
I've compiled rakudo on this machine
timotimo even when the cache has warmed up?
troydm yes 17:23
psch for reference, a laptop i have here runs moar -e'say "Hello"' in about 2.7 real; parrot is about 3 times slower i seem to recall timotimo++ said in his talk 17:29
that laptop is a old 2.2ghz single core though
timotimo my talk? 17:34
psch yeah, you showed benchmarks at gpn, didnt you 17:35
i watched the first half hour yesterday before going to bed
psch maybe that's why i misremembered it, if i did that :P 17:35
PerlJam Did doy just say Rust is similar to Perl 6?!? ;) 17:54
moritz if you look hard enough, you find some similarities 17:56
like, both are still in flux 17:57
both aren't widely adopted yet
... and that's about where the similarities end :-)
colomon lizmat++ is talking on my computer! 18:00
okay, lizmat is breaking up on my computer. feed is very choppy 18:02
PerlJam colomon: It's probably because I switched to watching lizmat ;) 18:03
(But the feed is smooth for me)
moritz sounds like your computers are connected.
colomon stop sucking down all the bits!@!!!!! 18:04
moritz will work from $home tomorrow, with public transport on strike, and family sick
colomon moritz: :( 18:05
moritz colomon: at least my employer doesn't complain, that's worth something
colomon sick family is never fun, though.
PerlJam (moritz's employer)++
moritz colomon: aye
colomon lizmat has stopped breaking up. yay! 18:06
PerlJam What's with all of the mirrors in all of the rooms? Are they trying to make the rooms look bigger?
moritz or afraid of vampires? 18:07
timotimo psch: i didn't know you knew german :)
PerlJam It's Florida ... who ever heard of vampires visiting Florida? 18:08
psch timotimo: german is, in fact, my first language :) 18:12
timotimo ah! 18:13
colomon CLDR, eh? 18:17
colomon is watching patch++ now
itz watchs timtoady on www.perl6.org/archive/rfc/28.html 18:19
itz hmmm no lizmat sound 18:24
PerlJam itz: In that talk he also called out one of my RFCs www.perl6.org/archive/rfc/114.html as rejected :)
itz ah sound now 18:27
masak PerlJam: wow, you submitted four of them. 18:28
colomon PerlJam++ 18:29
masak and cognominal++ two.
PerlJam masak: I did? I only remember a couple. 18:30
masak well, it depends on what exactly "maintainer" means on www.perl6.org/archive/rfc/by-group.html
PerlJam Ah ... I apparently wanted to gie ALLCAPS subroutine names back to the end-users by proposing a way to signify special subs. 18:33
masak special subs, as in phasers? 18:43
PerlJam aye 18:44
lizmat good *, #perl6! 19:22
do we actually have an idea how to implement "goto" in rakudo / moar ?
psch hi lizmat
lizmat psch o/
masak hm, I remember discussing `goto` with someone recently. 19:23
was it you, FROGGS?
FROGGS masak: possible 19:24
goto shouldn't be that hard to implement in moar me thinks 19:25
it is like unwinding after an exception happened to the spot right after the exception happened 19:26
lizmat so it's just a SMOP ? 19:26
FROGGS so the goto label stores its "pc" (p-something counter), and then we just unwind to it using nqp::unwind perhaps 19:27
lizmat *program* ?
FROGGS "perhaps" :o)
moritz I think the PC stands for "politically correct" register :-) 19:29
lizmat :-)
FROGGS hehe
lizmat moritz o/
FROGGS "Perl 6 has P.C. built-in" - that fits :o) 19:30
masak I have an idea that I can't quite express yet. 19:34
it may turn into one of my YAPC::EU talk proposals. 19:35
the idea is something like "a macro toolkit for building control flow has an API that looks stackless, regardless of actual underlying implementation"
FROGGS what a nice short catching title :P 19:36
moritz s/has/as/ ?
masak moritz: no, 'has' 19:37
masak as in, the idea is that a toolkit (invariably?) *has* such an API. 19:38
lizmat s/has/which has/ ?
masak you keep wanting my idea to be a noun phrase! :P
[Coke] is nearly at Worthington, Minnesota. 19:38
lizmat choo choo! 19:39
lizmat assumes [Coke] is on a train 19:39
fab1an Hi there. I am trying to connect to an sqlite db. Using DBIish.connect('SQLite') command i need to add an database name. Does anyone knows how? 19:40
moritz fab1an: didn't we have that topic, like, yesterday?
fab1an yap. i am working on an example for the readme. 19:41
moritz wates far too much time on dhmholley.co.uk/civclicker.html building a zombie-only empire
moritz fab1an: did you get it working from looking at the test code? 19:42
[Coke] lizmat: nope. the wife is driving us across country in a rented car. 19:43
lizmat ah, your support human (as Chip Salzenberg explained to me today)
:-)
fab1an unfortunately no. copy pasting the connection steps to the repl does not work, 19:45
[Coke] hee. 19:46
[Coke] Hey, tell chip I said hi! 19:46
moritz sm: use DBIish; my $dbh = DBIish.connect( 'SQLite', :database<camelia.sqlite3> ); say $dbh.prepare('SELECT abs(4)').fetchrow.perl 19:47
camelia star-m 2014.04: OUTPUT«Cannot locate native library 'libsqlite3.so'␤ in method postcircumfix:<( )> at /home/p6eval/star-2014.04/languages/perl6/lib/NativeCall.pm6:127␤ in method connect at /home/p6eval/star-2014.04/languages/perl6/lib/DBDish/SQLite.pm6:221␤ in method connec…»
moritz fab1an: how does it not work? 19:48
sm: use DBIish; my $dbh = DBIish.connect( 'SQLite', :database<camelia.sqlite3> ); say $dbh.prepare('SELECT abs(4)').fetchrow.perl 19:49
camelia star-m 2014.04: OUTPUT«fetchrow_array without prior execute␤ in method fetchrow at /home/p6eval/star-2014.04/languages/perl6/lib/DBDish/SQLite.pm6:136␤ in block at /tmp/XCrF0Z6Nit:1␤␤»
moritz sm: use DBIish; my $dbh = DBIish.connect( 'SQLite', :database<camelia.sqlite3> ); my $sth = $dbh.prepare('SELECT abs(4)'); $sth.execute; say $sth.fetchrow.perl 19:50
camelia star-m 2014.04: OUTPUT«Array.new("4")␤»
moritz fab1an: working example produced right there for you 19:50
fab1an i am trying it... and it works. :-) 19:51
what is this construct: :database<camelia.sqlite3>? 19:52
masak fab1an: it's a pair. 19:52
fab1an: where the key is a string.
er, value* 19:53
though the key is a string, too :)
fab1an: it means 'database' => 'camelia.sqlite3'
moritz short form of database => 'camelia.sqlite3'
[Coke] uploads one of the worthington shots, though this is not the best one:
feather.perl6.nl/~coke/worthington.jpg
moritz masak: without the quotes around the key (that wouldn't be a named arg)
fab1an ahh ok.
i will add an example script with a short description to the readme. i think this will help other trying to use DBIish. 19:55
moritz ++fab1an
[Coke] crap, anyone II might of visited perl wise on the west coast is probably on the EAST coast. :) 19:56
masak moritz: ah, right. 20:00
[Coke] gah. *might have 20:01
FROGGS lizmat: about the goto SMOP.... goto is just that ugly that it is not even on my TODO :o) 20:03
lizmat ah, ok 20:04
lizmat makes mental note
FROGGS is still fighting Archive::Tar... 20:06
which also has at least one goto
lizmat can you get around it (no punt intended) 20:07
*pun
FROGGS I guess a loop and a redo will do 20:08
brrt whats wrong with goto?
FROGGS yes, it is only one label and the goto's come afterwards in the same sub 20:09
brrt: you often can't see the label, so you cannot see where it jumps to
lizmat brrt: it needs to go looking up the stack for it at runtime 20:10
at least in p5 :-)
brrt would be trolling if he suggested to get a bigger screen (or smaller letters or better glasses)
what
up the... what
lizmat call stack 20:11
brrt i'm kidding, though, but, look up on the stack? how even
FROGGS in our case the label is just a lexical, like we redo a loop atm
brrt ahaa
the very late binding thing :-)
psch wrt "can't see the label": zf* when on the label in vim should let you see it... :)
FROGGS the label would be in the call chain... hmmm 20:12
brrt i'm still struggling to imagine a case wherein the goto label would be dynamically bound 20:13
oh, i know one case
exceptions
psch anyway, tr/// now just needs actual code to compute the distance instead of just doing &infix:<ne> on the original lhs and the result and escape sequences 20:15
the latter is what i've been trying the last few hours, and it's enough for today
so, good night #perl6 o/
lizmat good night Psyche^_ 20:15
.oO( not fast enough :-)
 20:16
lucas__ p6: sub foo is export(:a, :b, :c) {}; say "alive"; 20:26
camelia niecza v24-109-g48a8de3: OUTPUT«===SORRY!===␤␤Export tags NYI at /tmp/tmpfile line 1:␤------> sub foo is export(:a, :b, :c) ⏏{}; say "alive";␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line …»
..rakudo-{parrot,jvm,moar} a7b4cb: OUTPUT«alive␤»
lucas__ p6: sub foo is export(:a :b :c) {}; say "alive";
camelia niecza v24-109-g48a8de3: OUTPUT«===SORRY!===␤␤Multi colonpair syntax not yet understood at /tmp/tmpfile line 1:␤------> sub foo is export(:a :b :c⏏) {}; say "alive";␤␤Export tags NYI at /tmp/tmpfile line 1:␤------> sub foo is e…»
..rakudo-{parrot,jvm,moar} a7b4cb: OUTPUT«alive␤»
lucas__ commas are optional in export lists?
or maybe I didn't get pair syntax right 20:27
std: sub foo is export(:a, :b, :c) {} 20:28
camelia std 0f2049c: OUTPUT«ok 00:01 127m␤»
lucas__ std: sub foo is export(:a :b :c) {}
camelia std 0f2049c: OUTPUT«ok 00:01 125m␤»
masak 'night, #perl6 21:08
FROGGS gnight masak 21:09
Kirkkaf Hello, can someone provide some guidance on whether I should begin my journey with Perl 5 or 6? I am require to learn Perl to further my career but have never really done any serious programming. 23:05
japhb_ Kirkkaf, What do you mean when you say you are required "to learn Perl to further my career"? Are you working in a group that does a lot of Perl already? If so, you probably want to learn the same one they are using, which is most likely a Perl 5. 23:09
If you were just told "Learn a Perl, it's good for your brain", then your choice is more open. We tend to believe Perl 6 is pretty good for your brain, but it is not bulletproof, hardened, and ready for large-scale high-volume deployments yet. (We are still working on a few final features, as well as heavily working on optimization, bug squashing, and shaving off the sharp bits of edge cases.) 23:11
nickf_mt So, quick question... Looking to get started playing with Perl 6... I note the perl6 book on github was last updated in 2012... Would this still be a useful resource to start with, or should I look at more current online tutorials/refs? 23:13
Kirkkaf japhb_, my company is writing alot of Perl but the version was not specified so I guess I will have to get the details but are the two very different in design? 23:17
japhb_ Kirkkaf, Yes, Perl 6 is a completely new language, more of a sister language than a version increment. 23:17
Perl 6 includes a large number of features that couldn't reasonably be retrofitted into Perl 5 due to backwards-compatibility constraints in the Perl 5 community. 23:19
So Perl 6 takes advantage of the compatibility break to include some amazing new features.
japhb_ nickf_mt, What kinds of things are you interested in learning? 23:21
japhb_ Note that I've not read the GitHub Perl 6 book, as I joined the community before it was written. 23:22
nickf_mt hrm... didn't have a lot of specifics in mind, just looking to play with a new language. I guess I'll probably be wanting to interface with databases, use regex if I use it at work down the line 23:25
maybe output excel files. (working through the cpa exam atm, starting as a tax accountant in september)
nickf_mt Any idea how much the language itself has changed in the last 2 years? As opposed to ironing out implementation/performance issues 23:29
japhb_ nickf_mt, You'll definitely want to learn about Grammars. jnthn++ has a good talk on this: jnthn.net/papers/2014-fosdem-perl6-today.pdf 23:33
nickf_mt excellent, thank you. downloading :) 23:34
japhb_ As for language changes, there have been changes in the concurrency design, an expansion of Rakudo to multiple backends that had some knock-on effects as we learned more, a fleshing out (and implementation) of the more advanced module ecosystem design, and so on. 23:35
The core language hasn't changed much, just gotten smoother, less buggy, faster, and so on. 23:36
Oh, there have also been improvements in I/O, especially with the creation of the async I/O model.
(I'm sure there are more that I am forgetting off the top of my head.)
nickf_mt ok... so sounds like I should plan on looking elsewhere once I get past the very basics, unless they update that book :) 23:38
given my large block of freetime to really dig in will be August though, I shan't count on that happening 23:39
Oh, sounds like the rest of the family's home. Thanks for the info, japhb . See y'all next time I have a little downtime! 23:41
japhb_ nickf_mt, The channel is quite helpful too, and most active from about 8 AM to midnight in the European timezones (others are available in Asia and Americas, but fewer of us at any given time).
nickf_mt, Enjoy!
nickf_mt alrighty, thanks again! 23:42

Powered by Cro and the Raku® Programming Language.

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

Thank you!