🦋 Welcome to Raku! raku.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: colabti.org/irclogger/irclogger_log/raku
Set by ChanServ on 14 October 2019.
SmokeMachine kawaii: github.com/FCO/Red/blob/master/t/4...formater.t 00:00
kawaii SmokeMachine: I will test this tomorrow morning, thank you! :D 00:29
SmokeMachine kawaii: thanks! and if is someone interested to know all experimental features on Red: fco.github.io/Red/tutorials/experimental 00:34
guifa .tell lizmat OOPS, as per seems the usual, I’m a terrible ecosystem citizen and forgot to add Intl::Timezone to the ecosystem file list. My bad =/ 01:05
tellable6 guifa, I'll pass your message to lizmat
Geth_ ecosystem: alabamenhu++ created pull request #520:
Adding Intl::Usertimezone to the list		 01:09
rir 'p6: my $sc = [ [0, 1], ]; $s.elems.say;' 02:51
ShimmerFairy p6: my $sc = [ [0, 1], ]; $s.elems.say; # doesn't work with the surrounding quotes 02:55
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$s' is not declared
at <tmp>:1
------> 3my $sc = [ [0, 1], ]; 7⏏5$s.elems.say; # doesn't work with the su
rir p6: my $sc = [ [0,1],]; 02:56
camelia ( no output )
rir p6: $sc.elems.say; 02:57
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$sc' is not declared
at <tmp>:1
------> 3<BOL>7⏏5$sc.elems.say;
ShimmerFairy p6: my $sc = [ [0, 1], ]; $sc.elems.say;
camelia 1
ShimmerFairy You type $s instead of $sc in that first example
rir Thanks, I am getting the desired construct but is there another way to write such an array as a literal? 02:58
ShimmerFairy m: my @sc = [0, 1],; say @sc.perl 03:00
camelia [[0, 1],]
ShimmerFairy That's a bit shorter, though you may find the trailing comma out in the open a bit weird. 03:01
m: my $sc = $[0,1]; say $sc.perl;
camelia $[0, 1]
ShimmerFairy m: my $sc = $[[0,1]]; say $sc.perl;
camelia $[0, 1]
rir The concern is that .elems == 1; 03:02
ShimmerFairy oh, in that case 03:03
rir m: my $sc = $[ [0,1] ]; $sc.elems.say; 03:04
camelia 2
ShimmerFairy m: my @sc = 0,1; say @sc.elems;
camelia 2
rir m: my $sc = $[ [0,1] , ]; $sc.elems.say;
camelia 1
ShimmerFairy m: my $sc = [0,1]; say $sc.elems; 03:05
camelia 2
ShimmerFairy The problem is that the trailing comma in [ [0,1], ] tells Raku you explicitly want a one-element list. 03:06
m: my $sc = [[0,1]]; say $sc.elems; 03:07
camelia 2
rir That is what I want! But I cannot find it documented; I.e.: [[ 0, 1] , ] 03:08
ShimmerFairy This demonstrates trailing comma at the beginning: docs.raku.org/language/list 03:09
rir m: say reduce &infix:<->, (10,); 03:14
camelia 10
rir m: say reduce &infix:<->, (10);
camelia 10
rir Thanks ShimmerFairy. I would not have gotten that a ] is the same as a ) . 03:20
ShimmerFairy m: say [1,2].perl; say (1,2).perl; 03:23
camelia [1, 2]
(1, 2)
ShimmerFairy m: say [1,2].WHAT; say (1,2).WHAT;
camelia (Array)
(List)
ShimmerFairy rir: They do create different objects, but the way they function from a syntax perspective is the same. (An Array is a List you can change by adding or removing elements) 03:24
(docs.raku.org will, of course, better explain the differences between the two, but in general Array is a mutable List) 03:27
Geth_ ecosystem: db373c2331 | L'Alabameñu++ (committed using GitHub Web editor) | META.list
Adding Intl::Usertimezone to the list
 07:09
ecosystem: a05ce75293 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | META.list
Merge pull request #520 from alabamenhu/patch-7

Adding Intl::Usertimezone to the list Thanks!
cpan-raku New module released to CPAN! Red (0.1.21) by 03FCO
El_Che_ a new release? i am on holiday but I will try to start the build thia evening 09:26
Altai-man_ Yes. Sorry for the delay. 09:27
El_Che_ .tell Altai-man no sorries, thank you for the release 10:25
tellable6 El_Che_, I'll pass your message to Altai-man_
El_Che_ travis is building the linux packages now 10:26
Altreus jnthn: in fact it was to *ensure* termination - I wanted to know about errors from the promise, so I had to somehow resolve it 10:58
holyghost My Escape key is broken on my main develop machine (a portable) so I had to switch from vi/vim to xemacs :o) 11:15
jnthn Altreus: That's also a very valid thing to do, yes :) 11:18
holyghost my main editor on macs was aquamacs with gnus as a news reader, I still remember the keys
Altreus :) 11:21
El_Che remap the keys? 11:44
caps locks is a good candidate
raku-bridge <tmtvl> I think vim has a Control+something-or-other bind to leave insert mode, but I've been using Emacs for long enough that the Vim bindings have left my mind. 11:49
holyghost El_Che : I am slowing down like that and I also have to reprogram the leds ;-) 11:53
holyghost "When I am on Windows, I use xemacs" :-) 11:57
codesections Control + [ leaves insert mode in Vim. (in terminal vim, it's not really that it's *bound* to the same function as Esc, it's that it *is* Esc – or, rather, that Esc sends a control-[ key sequence 12:29
rypervenche I was just typing out the ctrl+[ option (also ctrl+c, although it also does something a bit differently) and then saw codesections's reply >< 12:33
holyghost ok thx 12:36
problem is I have to change my typing, but a bit of emacsen won't hurt 12:41
I'd like to program an editor in scheme, I started on one in objc in XCode but lost interest 12:42
I got "Learning GNU Emacs" (O'Reilly) as a present from my profs 12:43
goto-char is probbly the best/mot used function
*most
That way I can hack lisp on several OSes 12:44
too bad slime.el is broken these days
slime was the best lisp mode
kawaii I have a submodule like `My::Example::Thing` which contains a variable, i.e. `my $test = 'some thing'`, but if I `use My::Example::Thing;` in my script and then try to call the variable I get `... not defined`, is there a proper way to expose it? I've tried using both `our` and `$GLOBAL::`. 12:45
Altai-man_ kawaii, setting it to `our` should just work. Also `is export`. 12:46
tellable6 2020-07-21T10:25:49Z #raku <El_Che_> Altai-man no sorries, thank you for the release
ShimmerFairy I believe 'our' would expose it as $Name::of::the::Module::test 12:48
kawaii Altai-man_: I tried `my $thing is export = ...` but got errors about expecting constraint 12:49
Altai-man_ kawaii, `our $foo is export = 42`? 12:50
samebchase- holyghost: you can try sly as well
kawaii Altai-man_: `wasn't defined` :( 12:51
holyghost samebchase- : ok 12:52
I'm bit confused on many or less colors though, lost my .emacs file long ago
I had a nice black and blue system
I'll code up some custom-faces later on 12:53
ShimmerFairy I happen to like the dark solarized look (as in, looks like a Solaris computer) for my emacs, there's one or two themes out there for that.
holyghost ShimmerFairy : right, CDE (KDE from sun) had ushca colour scheme 12:55
*such a
El_Che rakudo-pkgs for 2020.07 are built and in the repos 13:46
Altai-man_ El_Che++ 13:47
holyghost For the RSC, lizmat, JJMerelo and others : I would like to suggest a milestones file with raku goals, put online somewhere, so everyone gets the synergy and can coordinate 14:43
This way we see the future 14:44
Then if I may say so, if raku gets modularized, so APIs in the code, it would be easier to swap out things
That way you can build several rakus, e.g. by changing the network module 14:45
guifa o/ 14:45
holyghost perl5 has a system ecosystem for that, for example HTTP::Tiny 14:46
just my 2 cents
holyghost We see the future, s I said :-) 14:46
*s
*as
You get more people focused like that :-) 14:47
security through obscurity isn't all that, FWIW 14:49
Now I have to watch out that I don't repeat myself, but I will when RSC is in place 14:50
So perhaps a modular raku but some project goals such as masak's Algoloid Macro, online 14:53
To help each other out
By the modular system, you comprehend that it's for everyone's benefit 14:54
for raku it'd be a major rewrite though, which isn't always a good idea, OOP can be put there also 14:55
but that's against most C hackers who can do it in C instead of e.g. elf C++
holyghost a OOP raku system can then be further developed managed e.g. by the same people above everyone 14:57
major nd minor systems
*and
macos X is written in C++, whereas BSD nd linux are not (but in C) 14:58
the last are more monolythic
with C++ I mean pre-11,14,17 for a decent C++ compiler such as c++ on BSDs 15:00
or with "gcc -std=c89" for C++ on the command line 15:01
[Coke] holyghost: I am having trouble following what you're talking about. 15:02
holyghost the future :-) 15:03
I am elaborating a milestones web page 15:04
[Coke] ok. hard to tell which points go together, which are separate, and what, the C/C++ is about. 15:05
holyghost I'll try to aid more when the RSC is in place, I'm just talking for a raku coding channel
holyghost [Coke] : well, we'd have a more flexible compiler/interpreter in an OOP language 15:08
then again C has its major benefits, I am not against a C system
it's just talk for if we could do a OOP version of the perl6/raku system
Which I am not going to do on myself
but if there's interest, there'd be more code and mor hacking 15:09
It's somewhat personal, but I want some distraction from always the same C raku system 15:10
I have started unterpaganperl6 which uses formulas and is written in C++ for example
a formula is just backtrackable for syntax 15:11
you just reparse if needed
We cn go to commercial UNIXEN/Linuxen also 15:12
anything with a C/C++ compiler works
as gcc/g++ stands for
although c++ is more clean (as a compiler command) 15:13
We always need more I guess :-)
Most things are not some casual write or rewrite 15:14
C and C++ (or some other mjor OOP language) are just simple things for milestones file 15:17
timotimo how does "C and C++" exactly relate to raku? 15:18
holyghost I'm not saying that it's the only way to go
timotimo : it's more modular in C++, I'm not saying that the developers should turn towards C++ 15:19
C is fine by me as it cn be used to program OOP such as C++
I'm no OO zealot myself BTW 15:21
In C++ it's easier to have modulr compiler
but in C you can do the same thing 15:22
Everyone has put its time into raku in C, so we shouldn't throw it away I mean 15:23
patrickb c: '' ~~ m:i/ ^ [a | bc * | ''] $/
timotimo "it" is more modular "in" c++?
committable6 patrickb, ¦'': «Cannot find this revision (did you mean “all”?)»
tadzik you can write OO in C just fine, that's what Linux does, or Gtk 15:24
timotimo both rakudo and nqp are very big parts written in nqp, which is object-oriented
tadzik you don't need language features for it (though they can certainly be helpful)
patrickb m: '' ~~ m:i/ ^ [a | bc * | ''] $/
camelia ( no output )
holyghost sure, as I said, you can program OO in C
patrickb m: say 'matched' if '' ~~ m:i/ ^ [a | bc * | ''] $/ 15:25
camelia ( no output )
patrickb m: say 'matched' if '' ~~ m:i/ ^ [a | b * | ''] $/
camelia matched
holyghost I know Gtk widget construction system as you say
patrickb Is the above a bug?
timotimo glib in general
holyghost elaborating on milestones then again, for our synergy 15:26
rypervenche patrickb: Why do you think it would be a bug?
timotimo patrickb: i think it's supposed to say "null regex not allwoed"
rypervenche Oh
timotimo m: say "m" if '' ~~ m:i/^ [a | b | ] $/ 15:27
camelia 5===SORRY!5=== Error while compiling <tmp>
Null regex not allowed
at <tmp>:1
------> 3say "m" if '' ~~ m:i/^ [a | b | 7⏏5] $/
expecting any of:
infix stopper
patrickb m: say 'matched' if '' ~~ m:i/ ^ [a | b * | ''] $/; say ', '; say 'matched' if '' ~~ m:i/ ^ [a | bc * | ''] $/;
camelia matched
,
holyghost again, more modular raku should be better for more hackers 15:28
patrickb The only difference is the "b" vs "bc". I don't get why that makes it match / not match the output.
holyghost hold on, I'll recapitulate when the RSC is in place
timotimo patrickb: b* can match "", bc* must at least have a b to match 15:30
patrickb oh
timotimo since * binds tightly to the c in the second
patrickb dumb
timotimo :)
Geth_ doc: codesections++ created pull request #3528:
Remove documentation for (nonexistent) `allowed` method
timotimo holyghost: how is raku not modular? 15:31
kawaii I have an object which I'm calling a method on for use in a string i.e. `description => "{$user1.nick} has given {$user2.nick} a reputation point!"`, however it's possible for `.nick` to return `Any` (nothing) if the user has no nickname, there's another method I can use `.username` instead though as a fallback, is there a clean way of doing this? :) 15:32
holyghost timotimo : modules can be replaced 15:33
swapped in and out
thus more programmatic systems
timotimo yeah? 15:34
holyghost sure
timotimo how is that not yet the case?
holyghost I am no raku developer, but if it's in already, fine by me
timotimo i think it would be worthwhile if you checked stuff out before writing pages full of irc messages with suggestions that don't apply? 15:35
holyghost a module is piece of code which can be swapped 15:36
as I said before 15:37
several developers can work on a module, wheres others do not
that way you integrate shit
Where I come from (inkscape and worldforge) you write OO for that 15:38
timotimo rakudo is already OO on its inside
holyghost Yes, e.g. a struct is a public class 15:41
that's the simplest idea possible
jnthn You seem to be thinking rakudo is written in C, but it isn't (well, maybe 0.5% is, and even that tiny bit's time is short...) 15:42
The only significant component written in C is MoarVM
Which is the runtime, not the compiler
timotimo in rakudo, Perl6::Compiler is a class, Perl6::Grammar is a class, Perl6::Optimizer is a class, all the nodes in the AST are classes, they get roles mixed in 15:43
holyghost I'm a lost soul on that
timotimo the Perl6::Grammar is derived from HLL::Grammar and also has roles mixed in
jnthn The RakuAST approach is even more object oriented :)
holyghost ok :-) 15:44
jnthn : in what language is the rest written ?
which languages ?
timotimo a big portion in nqp, the rest in raku 15:45
holyghost Anyway, I'm just a package writer but if things can be swapped out and in, I rest more easily :-) 15:46
holyghost lynxes
right so it's not quite perl, parrot VM
timotimo parrot VM is in the past 15:47
holyghost sees the syntax 15:48
a parrot grammar engine :-) 15:50
CORBA lost :-)
[Coke] no
parrot isn't a thing anymore.
holyghost I know that
ok, so C++ is eventually out of the question and raku has rights over the nqp language 15:51
It's perl all the way 15:52
[Coke] no, it's not perl either. :)
holyghost :-)
I am happy it's has OOP features, which is logical for today's compilers/interpreters 15:53
holyghost Thing is, modularnes is good in several ways, as I said 15:53
it's a good focus
Petr37 Nice day 😎 Can i install Raku on windows 10 iot core? 15:54
patrickb Petr37: Did you try?
holyghost thx jnthn
Petr37 patrickb, not yet
patrickb Petr37, building rakudo could be a challenge as the iot core is very limited wrt what's installed. 15:55
patrickb Petr37, but I'd expect it to run fine. 15:56
patrickb Petr37: So maybe just go for the precompiled releases. 15:56
holyghost So perl6 uses perl7 :-)
patrickb Petr37: Do know that rakudo can use quite a lot of memory, so depending on the device you want to run on this might be a limiting factor.
Petr37 raspberry 4 8 gb ram 15:57
patrickb Petr37: Then you shouldn't have any problem :-) 15:58
Petr37 patrickb, iot core have many limitions
timotimo holyghost: no 15:59
holyghost So I guess perl7 and nqp direct perl6 in the future
patrickb m: say 'm' if '' ~~ m:i/ ^ [a | ''] $ /; say 'n' if '' ~~ m/ ^ [a | ''] $ /; 16:02
camelia n
timotimo oh interesting
s the :i makes it behave differently
patrickb ^ Is me dumb again?
timotimo that looks more like a bug 16:03
jnthn That looks like a bug
patrickb Should the correct output be '' or 'mn'?
jnthn mn 16:04
Since '' matches
MasterDuke bisectable6: say 'm' if '' ~~ m:i/ ^ [a | ''] $ /; say 'n' if '' ~~ m/ ^ [a | ''] $ /;
bisectable6 MasterDuke, Will bisect the whole range automagically because no endpoints were provided, hang tight
MasterDuke, Output on all releases: gist.github.com/4eb79bec0a2382796b...f1f0a7b23f
MasterDuke, Bisecting by output (old=2017.02 new=2017.03) because on both starting points the exit code is 0
patrickb will rakudobug
bisectable6 MasterDuke, bisect log: gist.github.com/c0e2b9cbfeb44e7ab1...f7f91b3194 16:05
MasterDuke, (2017-03-10) github.com/rakudo/rakudo/commit/22...9854e33739
MasterDuke, Output on all releases and bisected commits: gist.github.com/8de1da70b3cce877c5...b2f248be7e
patrickb bisectable6: say 'm' if '' ~~ m:i/''/; say 'n' if '' ~~ m/''/; 16:14
bisectable6 patrickb, Will bisect the whole range automagically because no endpoints were provided, hang tight
patrickb, Output on all releases: gist.github.com/935125f2a0422665b4...72a9f8bbd9
patrickb, Bisecting by output (old=2017.02 new=2017.03) because on both starting points the exit code is 0
patrickb, bisect log: gist.github.com/ca23e459299fed58f6...29f592e921 16:15
patrickb, (2017-03-10) github.com/rakudo/rakudo/commit/22...9854e33739
patrickb, Output on all releases and bisected commits: gist.github.com/39012039e947288855...3dc7846c0b
patrickb rakudobugged: github.com/rakudo/rakudo/issues/3815 16:16
timotimo what's the command to get a bump's stuff linked? 16:17
rev: github.com/rakudo/rakudo/commit/22...9854e33739
timotimo Geth_: ver github.com/rakudo/rakudo/commit/22...9854e33739 16:20
Geth_ timotimo, version bump brought in these changes: github.com/perl6/nqp/compare/2017....-g4ceee06d
holyghost lizmat, JJMerelo and other RSC members in the future : I will recpitulate for a minor/major discussion on modularization and a projectgoals/milestones file 16:23
patrickb github.com/MoarVM/MoarVM/compare/2...-g58457845 16:24
github.com/MoarVM/MoarVM/commit/58...c91194f40a <- this is probably the offending commit
holyghost RSC : in the future I meant
a RSC bot would be nice BTW 16:25
(RSC == Raku Steering Council) 16:26
The "Big Head" from 3rd rock of the sun :-) 16:27
jnthn patrickb: Hm, wonder if it's github.com/MoarVM/MoarVM/commit/58...130f8cR615
patrickb Found that out already ;-) 16:28
Oh! You pointed to a specific line!
No, I didn't see that yet! 16:29
AlexDaniel` holyghost: modularization? 16:30
holyghost AlexDaniel : see bove discussion, to swap things out and in 16:33
OOP wise
*above lol
AlexDaniel` what do you mean by that?
ah okay 16:34
holyghost To make more rakus :-)
not by patches but by more system libs/packages modularized
e.g. HTTP::Tiny in perl5 is int the perl5 tarball itself
but it's raku code 16:35
I'm not talking about raku packages
I'm taling about modifying the whole system piecewise
then getting more programmatic systems
so it has to be elevated, the raku system into modules I mean 16:36
it's better as it's adavnced OOP techniques which make the kernel 16:37
it's not in fully yet, but as nqp is a perloid it should do, I'm not sure about objects in nqp though 16:38
which enhances
blessed hashes I thought
I dunno :-)
timotimo no 16:41
holyghost java was programmed in java bck in the day now it's gcj
timotimo objects in nqp are very very close to objects in rakudo 16:41
not blessed hashes
holyghost ok
timotimo gcj has been abandoned some years ago
holyghost openjdk, sunjdk 16:42
timotimo "It was part of the GNU Compiler Collection for over ten years but as of 2017 it is no longer maintained and will not be part of future releases"
gcj doesn't exist any more
openjdk and sun jdk don't use gcj, and didn't use gcj
holyghost I know
holyghost blackdown.org, back in the day had the first sun java license which became the openjdk 16:44
the first license after windows and solaris jdk from sun itself of course :-) 16:45
Altreus big giant head 16:50
holyghost lol
holyghost modularization and milestones.txt, that's all I know :-) 16:54
or maybe milestones.php on github :-)
anyway, I'd better elevate with something else 16:57
jdoege I see what was done for the "install" process for Raku and I think it is much, much improved. (ra)Kudos to all involved. 17:13
tadzik :) 17:38
holyghost jdoege : I hd nothing to do with it, but we're doing our best :-) 17:40
El_Che++ 17:42
holyghost a wiki would also be nice, e.g. wiki.raku.org, as on c2.com (smalltalk nd OO people) 18:13
There's even one for romhacking.net
on c2.com there's much more differentition thn on most wikis, it was the first wikiwikiweb ever made 18:14
tech talk by the wiki users on subjects
then we can comment on features of raku 18:15
holyghost a wiki with logins 18:27
holyghost a META server can hold the ips of raku among other things 18:40
and IPs of game servers :-)
perl6 MUDs 18:43
I'm sorry, raku MUDs
holyghost maybe a nice project for someone, implement some finger daemon in raku or some other language, so one can read status messages and e.g. a milestones file (using telnet for example or http) 18:46
raku ports :-) 18:47
6969
Petr37 anyone use Raku in commercial project s? 19:11
moritz I know nine does 19:14
holyghost It's GPL2 and Artistic 2 licensed, so you can I guess
kawaii Yes, I am too
Petr37 thanks. i want use Raku in production (commercial) 19:16
holyghost It's a good programmatorical system 19:17
Petr37 holyghost, very good system )) 19:18
holyghost I concur 19:20
oddp Are some of these commercial project performance sensitive and require some dirty tricks or reaching down to c here and there? Has anyone written some articles using raku in large projects? 19:22
lizmat oddp: if there are "dirty tricks", I would assume it's mostly reaching out to NQP code in Raku to make hot code paths perform better 19:31
*not* reaching down to C (as is mostly the case with Perl and XS)
although some people might actually use NativeCall for their bespoke C code
but in any case, that would still be a lot cleaner interface 19:32
wrt articles about large projects: I would love to see and promote one!
holyghost we should become a better system for e.g. large projects, nice programatorical elaborated 19:34
maybe I said too much :-) 19:35
oddp Alright, thanks, guess I'll be doing a nqp deep dive then. Pretty much ignored it till now. 19:40
Liz, looking at some of the performance optimizations in recent raku commits, are there plans to revert some nqp usage back once raw raku code closes the gap more and more each release? 19:41
[Coke] I think that's a long term goal, yes. 19:43
gfldex oddp: Raku was designed to be able to self host. But that's a long way off.
oddp Okie dokie, good to know. 19:45
lizmat oddp: yes, that's the goal 19:48
gfldex .oO( You need a lot of shoestring for a 30 year bootstrap. ) 19:50
lizmat is looking forward to using RakuAST to have sprintf generate code, rather than running a grammar for every call to sprintf 19:52
[Coke] www.onlinegdb.com/Hy2lA2NxP ... wow 20:02
raku-bridge <theangryepicbanana> Just putting this here in case anyone is interested (since Raku is available on repl.it at repl.it/languages/raku): blog.repl.it/langjam 20:03
raku-bridge <theangryepicbanana> since raku is pretty good for making languages (in case you haven't noticed ;) ) 20:04
holyghost lol [Coke] 20:08
repl.italy :-) 20:09
guifa lizmat: that’s basically the same as me for all of the international formatting
lizmat ++guifa 20:10
holyghost lizmat, guifa, gfldex : if I may say so, you better watch out somewhat for a glibc/libc thing 20:12
guifa Thankfully mine generally uses the same dozen or so formats, so I can cache them a little bit, but even then, best I can do is a list of codes (including string literals) that get stringified. But being able to produce a code object without EVAL would be sweet 20:13
holyghost concurs
glibc destroyed kFreeBSD for example 20:14
although I'm not saying there should be native code for system calls
There ws Debian/BSD back in the day but glibc destroyed kFreeBSD which was ian and debby's thing 20:15
kFreeBSD is hosted on debian machines 20:16
glibc is a macro bitch whereas BSD libc is stable
libc implements system calls if you didn't know 20:17
if you can bypass glibc (not libc) with system calls in nqp you're ok, but not otherwise 20:19
so the long term project goal, as liz said is on the way 20:20
holyghost Any chicken can install debian, was in the install manual of debian earlier 20:21
So for major perl systems, we have MacOS X libc, BSD libc (including windows) and linux glibc 20:25
if you use the kernel interface you're locked in, so we _Will_ have to choose I guess
Larry wants BSD, Windows, Linux and MacOS X software, go figure 20:26
as raku/perl6 I mean
they're programatoriclly incompatible 20:27
General Public Virus (The New Hacker Dictionary), GPL 20:28
holyghost I went from linux to BSD for that, kFreeBSD was dead long ago, as somewhat but I don't need any more 20:30
holyghost mea culpa, Debby 20:31
holyghost Again, watch out for kernel /usr/include/X code 20:32
as Liz's sprintf
holyghost We better program on a sparc, mips64 and powerpc architecture :-) lossage again 20:36
we don't have GNU libc on commercial and BSD UNIXes though
for debian : apt-get dist-upgrade 20:37
a genius stroke
as Windows is based (with patents) on FreeBSD, it's a good monolithic kernel, I use NetBSD though as it's more stable and has more architectures 20:41
I was a severe debian user, back in the day
MacOS X has a better kernel as it's OOP programmed
anyway, I am exagerating :-) 20:42
on a NetBSD you can compile a clean gcc cross-compiler for m68-coff in a giffy
BSD kernels compile without warnings, whereas kernel.org (blah) do not 20:43
m68k-coff is for previous than elf amigas or appleII for example 20:46
you can code your own OS in it
coff is a microcode/binary format 20:47
there is also M$ coff
holyghost holyghost++ from Debby from debian and Ken :-) 20:48
glibc _can_ destroy perl6/raku 20:50
so watch out Liz and JJ for kernel interfaces beyond glibc/libc
GNU/google is a bitch for all of us
especially the BSD and MacOS (X) folks s it's called now
*as 20:51
lizmat MacOS is going to get interesting in more than one way for me personally:
holyghost K&R includes UNIX not glibc BTW (K&R is Kernighan and Dennis Ritchie's ANSI C programming language book
ok liz 20:52
lizmat 1. the way Apple is going forward, my next notebook is *not* going to be a Mac
2. with a Risc processor at the heart of a next generation of hardware, Raku's JIT will be useless
holyghost CISC vs. RISC 20:53
UNIX can goad along though, as its libc is for commercial and common UNIXen and raku can cope with that though 20:54
holyghost I'm not saying MacOS is big endian 20:54
vrurg lizmat: the 2 is truly worrisome one. I hope that by the time my macbook gets to old we'll have ARM JITting.
*too old 20:55
lizmat yeah, it would be a nice project, worthy of TRF funding
holyghost IRIX and SunOS are monolithic kernels but macs have a good kernel 20:56
it hopefully stays so
lizmat could also be nice to get Raku running better on Raspberries and the like
vrurg Otherwise I like the move. I never liked and fully trusted CISC. Even compared to DEC's PDP command set it was hellishly messy. 20:57
lunch time... 20:58
holyghost I am also pro RISC processors but CISC vs. RISC is a holy war 20:59
mea culpa
ShimmerFairy CISC only made sense in an era where making life easier for assembler programmers was important, but thanks to the x86 it'll likely never go away. 21:00
holyghost You can install several OSes on PPC and mips though 21:01
ShimmerFairy : right you are
motorola/freescale does not produce anything against intel anymore, mactel whereas PPC macs 21:02
e.g. IBM
we _can_ use these processors for phones though, it'd be a nice talk 21:04
FWIW windows mobile is better than android somewhat
and android _IS_ USENIX
as it includes TCP/IP stack, G4 is DSL and so on 21:06
all from USENIX
4G I meant
ethernet layes -> UNIX 21:07
*layer
from Palo Alto, Xeroc Parc all that, Jobs started with NeXT from the same guys 21:08
mad it into MacOS for the appleII
guifa I really have too much fun making cover images for github: github.com/alabamenhu/UserTimezone...s/logo.png
holyghost ok, doesn't link in seamonkey 21:12
* in my seamonkey 21:13
anyway, use a better libc, so nqp does not die
kFreeBSD is hosted on debian server BTW, if you can get its BSD libc for it, I'd thank you for it 21:15
*servers
by paying I mean :-)
holyghost use a good kernel interface, it helps out nqp/perl6/raku 21:17
debian servers host the iso and usb images I mean
man pkg_acc if you want to try out, Net or Open BSD 21:21
*pkg_add
man is a terminal command for manual 21:22
you can read it then
holyghost man pkg for FreeBSDs 21:26
holyghost glibc is GNU libc and libc is ANSI C UNIX libc, just for the unterwizards 21:28
"When I am on windows, I mostly use XEmacs" (among other things) :o) 21:34
it'd better be for the editor folks :-)
oh no! it's a linux program but never worry (on NetBSD as it's open source) 21:35
rjeli hi, i know very little about raku, but i'm writing a toy symbolic CAS that has to match left recursive patterns. wondering if raku grammars can support parsing non-text, i.e. an AST? 21:37
so i would need to translate a "pattern" ast to a grammar that parses another ast of expressions
holyghost thanks for withstanding me. 21:38
rjeli : have a rule system :-)
timotimo rjeli: it's not fully implemented at the moment, but it was speculated in the speculations 21:39
rjeli ok so is it like, 1. possible with a close reading of the docs, 2. possible with a close understanding of the language and some "interesting" hacks, or 3. not really possible (just write my own parsing algorithm) 21:41
timotimo you could spend some time implementing it into rakudo itself, but you'd also theoretically be able to use regular string matching tgether with assertions with code in them 21:42
rjeli ok, thanks 21:43
dumping to a canonical text representation is gross but im starting to think it's the easiest way of doing it :) :/ :) 21:45
lizmat rjeli++ 21:46
timotimo i meant something a bit cheekier 21:47
like, have essentially an empty string "tape" where the only thing it serves as is something where the cursor position can be advanced in
holyghost lol rjeli 21:48
rjeli oh my
holyghost :-)
timotimo though that's probably only sensible for a flat representation rather than a tree
rjeli thats.. not a bad idea
holyghost << 0
guifa rjeli: I am workingt o create a system that could do something like that 21:54
All I have is a proposal ATM though.
I’ve spent most of the time so far working on the binary version of it 21:55
rjeli: see gist.github.com/alabamenhu/2fec7a8...4a2ae2f04d
holyghost guifa : make a modular system component with it
modularization galore :-) 22:06
guifa holyghost: my ultimate goal is integrate them in some particularly tight way with Raku, similarly to how regex is (since they’re basically regex for binary and objects). 22:15
holyghost bummer 22:17
I don't follow but do the regex thing :-) 22:18
holyghost is bit drunk
or dlsym, dlopen 22:19
man, I didn't follow that at all :-) 22:22
holyghost guifa : are you integrating binary systems ? You should go OSS *lol* 22:24
we needs it :-)
We should go commercial for larger projects :-) 22:26
raku-activestate.com *lol* 22:27
raku-bridge <theangryepicbanana> binary grammars would be cool
lizmat getting activestate involved with Raku would not actually be a bad idea 22:28
but they seem entrenched in Perl 5|7
raku-bridge <theangryepicbanana> reminds me of Red/Rebol's PARSE dialect, which are similar to Raku grammars, but can also parse binary data and actual values
holyghost we needs it :-)
raku-bridge <theangryepicbanana> (both of which I think have been proposed for Raku iirc) 22:29
holyghost I'd like to
lizmat I've said before that I have half a mind re-implementaing the regex engine in Raku
holyghost sure liz
lizmat but only half a mind
holyghost we needs them binaries
activestate.com ?
raku-bridge <theangryepicbanana> wasn't raku's regex engine already implemented a second time? iirc it was originally in parrot
jnthn Whenever I think about it I wonder how many minds it will need... :)
holyghost lol 22:30
stupid bots :-)
raku-bridge <theangryepicbanana> (wait is the discord bridge broken or something?)
holyghost roflmao
It's controlled I tell you 22:31
raku-bridge <theangryepicbanana> I'll take that as "partially"
holyghost they're just AIs
lizmat botbotbotbotbotbotbotbotbotbotbotbotbot o boy
raku-bridge <theangryepicbanana> lol
holyghost fuck off ! :-)
lizmat I've been accused of being a bot on Twitter more than once :-)
holyghost like that :-) 22:32
bananas
raku-bridge <theangryepicbanana> really though, I think raku could take some ideas from red/rebol's Parse thing
holyghost It's difficult to tell the AI from a real person, liz
the AI is more I :-)
raku-bridge <theangryepicbanana> I would agree 22:33
holyghost ok, I need to sleep but I'm going to hack at 5am
we needs it :-)
raku-bridge <theangryepicbanana> ye
holyghost raku-bridge : tell me of sheherazades 22:34
*Sheherazades
raku-bridge <theangryepicbanana> idk what that is
lizmat
.oO( I'm a person, not a number :-)
holyghost you're stupid !
you're not liz
raku-bridge <theangryepicbanana> fwiw my name is theangryepicbanana here on discord 22:34
<theangryepicbanana> I thought the bot displayed my name lol 22:35
holyghost discordia is against marxism BTW, raku-bridge
raku-bridge <theangryepicbanana> discord is easier to use on a phone though
holyghost right you are, is it a mobile ?
timotimo probably a rotary phone 22:36
holyghost oh you want to tell us that you have your own interface ?
raku-bridge <theangryepicbanana> way
<theangryepicbanana> *wat
holyghost watts
raku-bridge <theangryepicbanana> I mean I can send a screenshot if you want
holyghost sure then I'll hve my lynx program download it :-)
raku-bridge <theangryepicbanana> cdn.discordapp.com/attachments/633...image0.png
holyghost lol 22:37
raku-bridge <theangryepicbanana> lol imagine not having embeds
holyghost roflmao
no maos inhere please ?
*in here
raku-bridge <theangryepicbanana> nice try. ios
holyghost just parse !
I'm having trouble with the status of your head 22:38
install festival voxes at once !
raku-bridge <theangryepicbanana> yeah I'm kinda lost too
<theangryepicbanana> lol
holyghost thanks :-)
raku-bridge : we got out of control, how do you parse ? 22:39
holyghost is going to grab another beer
raku-bridge <theangryepicbanana> wdym parse? like, interact with IRC?
holyghost hail raku-bridge ! :-) 22:40
for example
raku-bridge <theangryepicbanana> it's a bot someone made I think
holyghost ?x &|- raku-bridge
it does not parse though
raku-bridge <theangryepicbanana> oh you mean red/rebol parsing?
<theangryepicbanana> actually I'll just get on irc lol 22:41
holyghost do that then :-)
I'd really appreciate it (ask-p *lol*)
I mean askp
lizmat sleep& 22:42
holyghost then have a mailbox system for a VM
gn lizmat !
raku-bridge <theangryepicbanana> I'm not liz 22:42
holyghost right, here you are
raku-bridge <theangryepicbanana> lol
theangryepicbana ok there
holyghost grand knights they are 22:43
gkn
theangryepicbana lol I'm too much of a zoomer to use IRC regularly
holyghost sure
you're username is blogged
your name is blocked
theangryepicbana yeah it's too long for most sites I
holyghost lol 22:44
theangryepicbana *I'm on
holyghost sure, what's with the bot, it's written in what language ?
theangryepicbana most likely raku
holyghost coolnes
theangryepicbana using the shuppet module
holyghost lol
theangryepicbana anyways as for what I was saying before, are you at all familiar with the Rebol or Red languages?
holyghost I'm writing scheme bots ATM
no 22:45
theangryepicbana that's what the Parse thing is from
ah ok
holyghost is it any good ?
right
theangryepicbana yeah, it's pretty easy to use
and super easy for making DSLs
holyghost sure, I'm mostly in for AI and adaptive systems in programming
I don't have the time to look it up though
so it's networked ? 22:46
theangryepicbana well I'll link you then rebol.com/docs/core23/rebolcore-15.html
not really
like, domain-specific languages
holyghost ok
thing is, it does not link here, bummer 22:47
theangryepicbana or "mini languages" within an existing language
oof
copy+paste it maybe...?
holyghost I did
theangryepicbana oh
holyghost seamonkey
I'm a bit slammed by the KGB droids :-)
theangryepicbana I'm just using whatever client came up first (which happens to be kiwi)
holyghost now my beer getting 22:48
kiwi is all right
theangryepicbana well as for the link thing, you could just search up "rebol parse" and it's like the first result online
holyghost ok 22:49
right banana : so you did actually wrote the bot in a theorem prover language at first sight 22:51
theangryepicbana uh no lol
holyghost no ?
theangryepicbana no someone else made the bot
holyghost ah ok, something else than raku I guess 22:52
problem is you do too much stuff in one language
?
theangryepicbana basically there's a channel in the raku discord server that interacts with this channel
holyghost I understand, then you just hijack and paste
theangryepicbana not quite?
holyghost I am writing scheme IRC bots ATM
theangryepicbana are you at all familiar with discord? 22:53
holyghost lol
holyghost no 22:53
theangryepicbana mhm
holyghost you just popen ?
theangryepicbana it's similar to normal IRC, but has more features
holyghost ok
theangryepicbana no, there's a normal application
holyghost is reading the rebol parse stuff
theangryepicbana discord.com/new 22:54
holyghost It seems all right, but I'm no fan of haskell posts 22:56
HLL parsing is ok though
theangryepicbana lol it's not haskell
holyghost right
theangryepicbana it's more similar to a parser generator or PEG parser similar to raku, but more dynamic
holyghost just read angryepicbanana *lol*
coolnes 22:57
PEG ?
oddp Erm, did I unknowingly join #dadaism? ;)
holyghost dynamic in the way of upper HLL features ?
pegleg
theangryepicbana PEG stands for "parsing expression grammar" 22:58
holyghost sure oddp :-)
right
theangryepicbana it's an idea that Raku grammars use
holyghost problem is I'm more into older systems but rms told me to implement a langauage such as in slib into m68k
ok, I'm don't know much about that grammar 22:59
theangryepicbana ah ok
there's always time to learn something new I guess 23:00
holyghost grammars are mostly used within our system
sure
fact is, you have to have the right grammar compiler e.g. yacc but that's old
holyghost My promotor Theo always told me to look into grammar things, languages 23:01
but I'm more of a AI fan
which you cannot do right these except for scheme and LISP
AFAIK
I know HLL features and grammars though 23:02
theangryepicbana well, you can always use AI modules from Python in Raku by using Inline::Python
holyghost nice
I didn't know there's a python inline
holyghost TBH, we had a double java VM which had futures 23:02
it had inline java but I told them to program in the new language 23:03
it was called ambienttalk
theangryepicbana I've heard of that before
holyghost see on VUB pages, vub.ac.be
it's not stupid but there are trolls in it
It's google 23:04
I told them to produce something with it instead of just complaining about the trolls
so they have end user software
google Belgium
I told them to write a new AWT with it and so on 23:05
without inlining
I'm not too sure about those features for languages, personally
inlin I mean
s/inlin/inline/
theangryepicbana hmm 23:06
yeah that kinda stuff is out of my area lol
holyghost what'd you think about that
ok
It's a double VM
*drink*
WAM I'm smoking 23:08
holyghost theangrybanana : how did you code the surfacing system of the bot ? 23:11
It's just a redroid ?
theangryepicbana like I said earlier, I did not make the bot. I don't remember who did, but they're probably around here somewhere
holyghost theangryepicbanana I mean
ok sorry
theangryepicbana no problem, just wanted to remind you 23:12
holyghost thing is, I cannot code shit anymore as my scheme code won't work
I was elevated but than again not
anyway, rebol is nice
I just system hack it again and again
theangryepicbana cool 23:13
holyghost is a bit too drunk
anyway, if it's not scheme I don't want to learn it 23:15
bracket wise
theangryepicbanana : we're lost in shadowlands :-) 23:16
an Iron Crown Enterprises card game (CCG) 23:17
they're the old Tolkien artists card game
theangryepicbana "anyway, if it's not scheme I don't want to learn it" so... lisp languages?
holyghost indeed
for AIs, theorem provers, prolog/progol and adaptive systems 23:18
theangryepicbana mhm. scheme is not the only lisp language btw
holyghost I'm writing a theorem prover for delimiters in java tokenizers
theangryepicbana ok 23:19
holyghost java has a StringTokenizer class since java 1 or 2 23:19
it's not fast though
we i13s for it :-)
I'm also doing a prolog system for it at the same time 23:20
or prgol I don't remember
s/prgold/progol
s/prgol/progol
*we need 23:21
you can shift 32/64 bit code for it, but I wanted something heavier
I wrote a intelligent window manager code for it for gnustep 23:22
theangryepicbana gnustep? so... objective-c?
holyghost it's online somewhere, I don't remember the login on github
indeed
On Xliff's server (mine) I have a objc from gcc for psp 23:23
objc is way cool for it
psp == Playstation Portable 23:24
I don't remember what I did for the intelligent window system though
theangryepicbana yeah objective-c is cool, but it's interesting that other people still actually use it for stuff
holyghost sure
It's NexT 23:25
your macos 7.x.y :-)
7.5.5
and earlier
rms paid a version for it for GNU
libobjc2 from gcc is bull
on macs it's nicer 23:26
holyghost I have a mountain lion 10.8 system where I write some things on, it's Xliff's server only client except for my ibook G4 23:27
theangryepicbana ideally, you'd want to use clang objc
holyghost indeed, it does not grovel
llvm and such
theangryepicbanana : what do you like most, macs (OS X I mean) or BSDs ? 23:28
theangryepicbana I've never used BSD, so probably Mac OS
holyghost sure, *smoking* 23:29
you cannot do rebol on old macs though 23:32
so I guess you like more PPC systems ?
it'd be nice to have such things on IRIX or SGI machines 23:33
which are now sony, psp , R9000 been
theangryepicbana "you cannot do rebol on old macs though" I do believe that there are versions for PPC macs
yes
*so yes
idk what IRIX or SGI are though 23:34
holyghost *fetch a drink, smoking* 23:35
holyghost right, troubles I guess 23:38
anyway, rebol is a nice language, but I'd reimplement it in scheme, just for gigs 23:39
like my tokenizer code
in scheme
it's just HLL functionality 23:48
holyghost in C for example 23:54

Powered by Cro and the Raku® Programming Language.

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

Thank you!