»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:03 vlixes joined 00:05 spider-mario left 00:08 jerome_ left 00:14 bsb left 00:16 risou left 00:18 risou joined 00:20 eviltwin_b joined 00:22 allbery_b joined 00:23 geekosaur left 00:25 jerome_ joined 00:26 eviltwin_b left 00:28 risou left, ashleydev left 00:29 risou joined
lichtkind _sri: core of perl 6 is rather slim (before prelude) but just the attitude see everything as parsing grammars isnt metn to be minimal like tcl :) 00:34
00:36 ashleydev joined
_sri lichtkind: looking through rakudo it appears to still be years away from becoming truly usable, just trying to find the silver lining :/ 00:40
lichtkind _sri: i understand
00:41 risou left
lichtkind _sri: i wasnt even able to write a simple rakudo patch 00:43
00:43 risou joined 00:44 eviltwin_b joined, allbery_b left 00:56 Khisanth joined
sorear good * #perly6 01:00
01:03 thelazydeveloper left 01:07 eviltwin_b left 01:09 eviltwin_b joined 01:16 scott_ joined, scott_ is now known as Guest67706 01:20 bsb joined, tokuhirom joined 01:25 orafu left, geekosaur joined, orafu joined 01:27 eviltwin_b left 01:29 eviltwin_b joined 01:32 geekosaur left
dalek kudo/nom: 7eae9b2 | pmichaud++ | src/ (2 files):
Fix [|] and [&] to provide flat junctions (RT #112696, masak++).
 01:43
blets: 631e78c | (Herbert Breunung)++ | docs/appendix-a-index.txt:
better desciption of A and polish comparator ops
 01:47
01:51 whiteknight left 01:54 allbery_b joined 01:56 eviltwin_b left 01:57 eviltwin_b joined 02:00 allbery_b left 02:05 allbery_b joined
dalek kudo/nom: 2d2584a | pmichaud++ | src/core/metaops.pm:
Implement triangle forms for junction reducers: [\|] and [\&]
 02:06
02:07 eviltwin_b left 02:08 tyatpi joined
dalek blets: d43436b | (Herbert Breunung)++ | docs/appendix-a-index.txt:
enforce new format rule (links to glossary are italic) and (put markdown formating of links inside the square brackets )
 02:26
blets: bd655c2 | (Herbert Breunung)++ | docs/ (2 files):
densify report table
02:26 allbery_b left 02:28 ponbiki joined, geekosaur joined 02:30 ponbiki left 02:31 ponbiki joined, zb joined
dalek blets: b1997ef | (Herbert Breunung)++ | README.md:
add cowens to authors list
 02:32
02:39 vlixes left
dalek blets: d1abda7 | (Herbert Breunung)++ | docs/appendix-a-index.txt:
categorization has to be linked to B if possible
 02:43
02:43 orafu left 02:45 orafu joined, fgomez left 02:53 tyatpi left 03:01 Lorn joined 03:06 tyatpi joined 03:08 eviltwin_b joined 03:11 geekosaur left 03:17 wutawutawhat joined
wutawutawhat how close are you guys to 1.0? 03:17
dalek blets: a8fc6d5 | (Herbert Breunung)++ | docs/appendix- (2 files):
repair links in the contextualizer table
 03:18
lichtkind wutawutawhat: you mean that question serious?
wutawutawhat: actually nobody knows, its very complex progress is made dayle and case it which Perl 6 is usable all growing all the time 03:19
wutawutawhat: but finished product needs some years more since has a depths like no othe rlanguage im aware of 03:20
03:21 allbery_b joined
sorear wutawutawhat: Numbers mean nothing. I released version 17.0 three days ago. 03:21
03:23 eviltwin_b left 03:24 lichtkind left
sorear (for generous values of "I") 03:27
colomon masak++ 03:30
03:38 eviltwin_b joined
_sri s/1.0/stable/ 03:39
my bet would be on 5-10 years 03:40
03:41 allbery_b left 03:42 allbery_b joined
PerlJam _sri: does that include CP6AN ? 03:43
_sri i don't consider CP6AN that important, it will happen on its own once the base system is stable
in fact, the whole embedding perl5 nonsense is completely unnecessary imo 03:44
03:45 eviltwin_b left 03:46 REPLeffect joined
PerlJam you don't think that a perl 6 implementation should run perl 5 programs? or something else? 03:46
_sri prove you have a stable perl6 core with enough documentation for everyone to pick up the language and everything will be ported in no time 03:47
just look at the excitement about the p5-mop effort, the whole community is desperately waiting for progress 03:48
03:48 raiph joined
_sri but i guess perl6 is an all-or-nothing project... so who knows, maybe it will never reach that point 03:49
sisar phenny, tell tadzik, "perl6 version 2012.04.1-20-g0059aa6 built on parrot 4.3.0 revision RELEASE_4_3_0-12-gf983396" (32-bit Ubuntu). I get the warnings whenever I try to install any module. 03:51
phenny sisar: I'll pass that on when tadzik is around.
sisar: 03 May 17:44Z <lichtkind> tell sisar "just have a look section C is much larger than average so the C heading as a backling to the to looses its functionality if you have to scroll too much, so I added some C (second letter a-m)"
sisar: 03 May 17:44Z <lichtkind> tell sisar "and Cn for second letter N-z"
sisar: 03 May 20:05Z <tadzik> tell sisar which Rakudo is that? I don't get any warnings in Panda, with or without line numbers
PerlJam as long as we have an interested and talented group of implementors, perl 6 will eventually mature into "stable"
sisar phenny, tell lichtkind, ah, thanks for explaining that. 03:52
phenny sisar: I'll pass that on when lichtkind is around.
_sri PerlJam: i wouldn't be too sure about that, the focus seems to be on experimenting with exciting new features, not stabilizing the project 03:54
looking through the commit history i'm mostly just thinking "oh, that's a fun looking feature, but i have no idea what i would ever use it for" 03:55
PerlJam _sri: What "exciting new features" have people been experimenting with?
sisar phenny, tell tadzik, maybe its time to start version nos for panda ?
phenny sisar: I'll pass that on when tadzik is around.
sisar afk 03:56
_sri PerlJam: magical documentation comes to mind 03:57
comments that result in documentation based on their context 03:58
sure that's fun... but damn... is that why i have to wait 10 years for perl6?
imo stuff like that should be kicked from the language core and added later on with incremental updates 03:59
PerlJam I've never really thought of that feature as the main focus of Rakudo development. It's an implementation detail that found a champion. 04:01
_sri yea, there's a complete lack of focus
PerlJam Things like nailing down lists, laziness, parcels, etc. have been the main focus in my eyes. 04:02
_sri like, why is there no canonical todo list that defines exactly what needs to be done for perl6 to be considered stable? 04:03
benabik _sri: It's a fallacy to assume because someone implemented feature B, it slowed down feature A. 04:04
_sri benabik: you can't say the opposite either though
04:05 benabik left
PerlJam _sri: would you contribute more if there were such a list? 04:05
_sri PerlJam: almost certainly
04:06 benabik joined
PerlJam I don't think anyone but Larry has a feel for the general shape of Perl 6. (what it *must* have to be Perl 6.0.0) 04:07
_sri PerlJam: but i'm getting the impression that the goal is not so much a stable perl6 but researching language design
PerlJam The only thing that irks me slightly about any of the Perl 6 implementations (but particularly Rakudo) is that they are still too slow
04:08 eviltwin_b joined
benabik _sri: I can't say the opposite? What's the opposite of what I said? 04:09
_sri PerlJam: define a smaller core language, release a stable but slow version, and people will make it faster
PerlJam _sri: who decides what makes the cut for that smaller core? 04:10
_sri PerlJam: that's the big question :)
benabik _sri: A list of things to do: github.com/rakudo/rakudo/blob/nom/docs/ROADMAP
04:11 allbery_b left
benabik What needs to be done to have a stable "Perl 6" is to have a complete implementation. Most of the changes to the spec are done in response to someone trying to make it work. 04:12
_sri that's very important information
PerlJam benabik++ 04:13
_sri so perl6 is still in the design phase
PerlJam I had forgotten about the ROADMAP since I haven't had my Perl-6-implementor hat on in so long
benabik perl6 is not using the waterfall development model. 04:14
_sri how can there be a roadmap if those features are not completely designed?
benabik design and implementation happen in cycles
PerlJam _sri: implementations inform the design and vice versa
_sri that's very frustrating 04:15
this is exactly wat gives the impression that perl6 will never be "stable" 04:16
PerlJam This is exactly how Agile works only Perl 6 doesn't have the pressure of deadlines and budgets.
_sri even more reason to cut the losses and settle for a smaller language 04:17
that's what i meant with all-or-nothing project 04:18
benabik _sri: We have a smaller language. We call it Perl 6. It works. Today. People use it. 04:19
_sri that's just not true
PerlJam It is true.
04:19 allbery_b joined 04:20 eviltwin_b left
PerlJam Maybe you object to the fact that we can probably name all of the people who regularly use perl 6 so it's not "generally used" like python or php or ruby or perl or haskell or ... 04:21
04:22 alester joined
PerlJam Still ... we have some early adopters. Just not the critical mass needed to cross the chasm yet 04:22
_sri what is the deprecation policy on stable rakudo releases then?
where is the canonical language reference i can use to ensure i only depend on supported features? 04:23
PerlJam _sri: the Synopses are the language reference 04:24
_sri so i can use everything that's in the synopses? 04:25
PerlJam everything that works :)
_sri that's what i mean
benabik _sri: Welcome to a small open source project. If you want strict depreciation policies, this is not the best place to be.
_sri i just want what perl5 provides, a little more modern 04:26
04:26 MContagious joined, wutawutawhat left
PerlJam _sri: what do you want the current perl 5s don't have? 04:26
s/the/that/ 04:27
_sri you just can't say perl6 is ready, it would be silly to make people bet their businesses on a language that's still being developed
PerlJam "make people"?
People make their own choices
_sri PerlJam: deprecation policies and reliable references of supported languages features
PerlJam _sri: perl 5 has those things. What "little more modern" things do you want from perl 5? 04:28
_sri "stable" release i can rely on
PerlJam _sri: maybe Perl 5 already has everything you need and you really have no need of Perl 6
_sri PerlJam: type system, exceptions and a few smaller things 04:29
yes, i'm mostly happy with perl5, but it will never have a type system or good exceptions
PerlJam perl5 has multiple type systems on CPAN ;) 04:30
_sri i'm not mentioning sensible oo-system since p5-mop can solve that
PerlJam: right ;p
PerlJam Perhaps perl 5 can grow a type system like it's growing p5-mop 04:31
04:31 eviltwin_b joined
_sri that seems very unlikely 04:31
PerlJam p5-mop seemed highly unlikely just a few years ago. 04:32
04:34 allbery_b left, geekosaur joined 04:37 eviltwin_b left 04:39 ponbiki left 04:54 rurban left 04:55 rurban joined 05:01 birdwindupbird joined 05:04 MContagious left 05:09 sisar left 05:15 eviltwin_b joined 05:16 eviltwin_b left, eviltwin_b joined 05:17 geekosaur left 05:28 eviltwin_b left, geekosaur joined
moritz \o 05:32
05:36 alester left 05:38 geekosaur left 05:39 geekosaur joined 05:47 kaleem joined 05:57 raiph left 06:00 woosley joined 06:03 wolfman2000 left
dalek Iish: 66f1aca | moritz++ | / (2 files):
[Pg] basic grammar-based tokenization of SQL
 06:07
06:23 tokuhirom left, tokuhirom joined, woosley left 06:26 plobsing left, geekosaur left 06:28 geekosaur joined 06:32 wtw joined 06:44 eviltwin_b joined, fglock left, MContagious joined 06:45 Guest67706 left 06:47 geekosaur left 06:53 scott_ joined, scott_ is now known as Guest96768
tadzik May the 4th be with you, #perl6 06:56
phenny tadzik: 03:51Z <sisar> tell tadzik "perl6 version 2012.04.1-20-g0059aa6 built on parrot 4.3.0 revision RELEASE_4_3_0-12-gf983396" (32-bit Ubuntu). I get the warnings whenever I try to install any module.
tadzik: 03:55Z <sisar> tell tadzik maybe its time to start version nos for panda ?
07:00 brrt joined 07:03 pmurias joined 07:04 geekosaur joined 07:06 eviltwin_b left 07:07 woosley joined
pmurias _sri: re "just want what perl5 provides, a little more modern" there have been a few implementations that aimed at providing a usable subset of perl6 v6.pm, perlito6 (some by translation to perl5) but people (here) weren't interested 07:07
07:11 eviltwin_b joined
moritz I guess that subset implementations have a different target audience than full iplementations 07:11
07:14 geekosaur left 07:15 xinming_ joined, Entonian joined 07:18 xinming left
dalek Iish: c4f91e1 | moritz++ | lib/DBDish/Pg.pm6:
[Pg] small refactor

move statement preparation out of submethod BUILD to get better error handling use grammar-based tokenizer for rewriting placeholders		 07:19
07:21 fhelmberger joined 07:22 fglock joined 07:27 plobsing joined 07:30 woosley left, sporous left
dalek Iish: 251b9bf | moritz++ | t/99-common.pl6:
[t] Hi, this is your son's school. We're having some computer trouble.
 07:31
07:32 sporous joined
dalek nda: d00a058 | tadzik++ | lib/Panda/ (2 files):
Avoid some uninitialized variable warnings, sisar++
 07:36
07:36 sporous left, sporous joined
tadzik phenny: tell sisar d00a058586 should fix the warnings, let me know if everything's allright 07:37
phenny tadzik: I'll pass that on when sisar is around.
tadzik phenny: tell sisar I'm all for versioning panda, but I wanted to do some bin/panda refactors first 07:38
phenny tadzik: I'll pass that on when sisar is around.
moritz seems the fix for the warnings line number was good enough to be useful :-) 07:40
tadzik indeed :)
which commit is that? I don't think I noticed it
07:41 fgomez joined
moritz 1972721e4675a8a65e0b6d671e02fb7ce6f0cf53 07:41
yesterday 07:42
tadzik ha, nice one
moritz++
07:44 fgomez left, fgomez joined
dalek Iish: ebda4c2 | moritz++ | lib/DBDish/mysql.pm6:
[mysql] did your really name your son robert'); drop table students;--
 07:44
tadzik karma students; 07:45
aloha students; has karma of 0.
07:54 allbery_b joined 07:57 eviltwin_b left 08:09 REPLeffect left 08:14 geekosaur joined 08:17 allbery_b left, kresike joined 08:21 eviltwin_b joined 08:22 REPLeffect joined 08:24 allbery_b joined, geekosaur left 08:27 eviltwin_b left 08:30 eviltwin_b joined 08:33 allbery_b left, geekosaur joined 08:35 eviltwin_b left
dalek rlito: 4d56b5d | (Flavio S. Glock)++ | / (6 files):
Perlito5 - add $ENV{PERL5LIB} to @INC
 08:46
08:50 eviltwin_b joined, geekosaur left 08:52 liuchong joined 08:56 tokuhirom left 09:05 thelazydeveloper joined 09:09 mucker joined
dalek rlito: dc1a0ec | (Flavio S. Glock)++ | TODO-perlito5:
Perlito5 - TODO update
 09:11
09:18 cogno joined 09:19 geekosaur joined 09:20 sreejit joined
dalek Iish: 2a0dfe4 | moritz++ | README:
revamp README. Chose a 2-clause BSD license
 09:22
09:22 eviltwin_b left, daxim joined, bbkr joined 09:27 plobsing left 09:29 cogno left 09:34 eviltwin_b joined, geekosaur left 09:36 eviltwin_b left 09:38 geekosaur joined 09:40 bbkr left, bbkr joined 09:41 BlueT_ left, BlueT_ joined
jnthn oh shark o/ 09:41
fglock jnthn: hi! 09:42
moritz lolshark! 09:43
daxim needs moar lazzz0rs on their friggin heads 09:44
moritz jnthn: any LHFs in rakudo that you are aware of?
jnthn: I mean, low for me, not for you :-)
mucker fglock: when I use rakudo against it gives an error. Is that expected ? 09:46
against src6
dalek ar: f4960ac | moritz++ | / (2 files):
ship DBIish instead of MiniDBI
 09:47
fglock mucker: I haven't tried for a while, so it is kind of expected - I'll try to reproduce, but it will take a while to setup things here
mucker: last time I tried I got a parrot error 09:48
(installing rakudo star) 09:49
mucker cd src6;export PERL5LIB=lib 09:50
perl6 util/perlito6.pl
Unable to parse blockoid, couldn't find final '}' at line 106
zsh: exit 1 perl6 util/perlito6.pl
the line 106 looks pretty reasonble to me 09:51
moritz the real error might be before that
jnthn moritz: Well, if it's test passes you want, maybe look at some of: S02-types/keyhash.t, S02-types/keyweight.t, S03-sequence/limit-arity-2-or-more.t 09:52
fglock mucker: I think you mean PERL6LIB instead of PERL5LIB
jnthn moritz: For something slightly more adventurous, working on import lists would be worthwhile.
moritz: And should be somewhat in reach.
moritz jnthn: oh, import lists sound nice 09:53
mucker fglock: yeah PERL6LIB, same error
jnthn moritz: Just curious - did you add $?FILE and $?LINE?
moritz r: say $?FILE, $?LINE
p6eval rakudo 2d2584: OUTPUT«/tmp/S9hzqkbpVF1␤»
jnthn moritz: Noticed we don't run S02-magicals/file_line.t yet
moritz jnthn: did you forget to git pull? 09:54
jnthn: it's in t/spectest.data
jnthn huh, just pulled...
oh, so it is. 09:55
Sorry. :)
moritz++
moritz no problem :-)
jnthn: before I approach import lists, there is a TODO comment in traits.pm about EXPORT_SYMBOL 09:56
# TODO: Make this much less cheaty. That'll probably need the
# full-blown serialization, though.
jnthn yeah, it's those $*W calls
moritz jnthn: what kind of non-cheatyness did you have in mind?
jnthn moritz: The reason it makes the $*W calls is becase before bs we had to explicitly record the package creations as events so we'd get deserialization code generated. 09:57
moritz: Now there's no reason to do that. 09:58
moritz jnthn: and now that's not true anymore, it should just copy stuff into packages?
jnthn Yeah
For example:
$*W.pkg_compose($install_in);
That can just become
fglock mucker: why are you running it with rakudo? (is is self-hosting)
jnthn $install_in.HOW.compose($install_in);
fglock mucker: it should work though
09:58 mucker left
jnthn Easy way is just to look at what the $*W methods in question do :) 09:59
09:59 am0c joined
moritz jnthn: on the p6 level, how are exports supposed to be organized? 10:00
jnthn: I gather that there's a list (or a hash?) of tags in a (which one?) package somehow
jnthn moritz: S11 is LTA. But to a first approximation... 10:01
10:01 mucker joined
jnthn UNIT has a lexical package EXPORT. 10:01
The tags are below this.
Everything that's ever exported goes into ALL
Things that are just tagged "is export" go into DEFAULT
Do the default import list is in EXPORT::DEFAULT 10:02
The full one in EXPORT::ALL, etc.
moritz and for a Foo tag, we create a EXPORT::Foo... what? List?
jnthn Package
moritz ok
jnthn sub foo is export(:KiwiFruit) { }
10:02 eoh| joined
jnthn Installs a EXPORT::KiwiFruit::<&foo> 10:03
moritz and then 'use Test;' at compile time stubs in a 'my &ok', and at some later point, does &ok ::= Test::EXPORT::<&ok> ?
jnthn use does a need and an import
moritz and need does the stubbing
and import the binding
jnthn No 10:04
need is purely loading the module.
See line 758 in Grammar.pm.
moritz ok
jnthn It calls $*W.load_module - this does the loading of the module and triggers GLOBAL merging.
Then it calls do_import
Which gets passed the arglist
That in turn calls $*W.import 10:05
The import method in world stubs and binds.
That is, all the binding is done right away at compile time.
moritz allright
10:05 sivoais left
jnthn Thus why... 10:05
r: use Test; BEGIN { plan 42 }
p6eval rakudo 2d2584: OUTPUT«1..42␤# Looks like you planned 42 tests, but ran 0␤»
arnsholt r: role W[$s] { method q { say $s; nextsame } }; ([but] class {}, <hacker Perl6 another Just>.map: W[*]).q
jnthn Works.
p6eval rakudo 2d2584: OUTPUT«Method 'count' not found for invocant of class 'W'␤ in <anon> at src/gen/Metamodel.pm:1176␤ in method reify at src/gen/CORE.setting:4753␤ in method reify at src/gen/CORE.setting:4678␤ in method reify at src/gen/CORE.setting:4678␤ in method gimme at src/gen/COR…
10:05 eoh left
moritz arnsholt: * doesn't autocurry inside a postcircumfix 10:06
jnthn arnsholt: That won't...what moritz said
arnsholt Oh, right. Dang =)
jnthn r: role W[$s] { method q { say $s; nextsame } }; ([but] class {}, <hacker Perl6 another Just>.map: {W[$_]}).q 10:07
p6eval rakudo 2d2584: OUTPUT«Just␤another␤Perl6␤hacker␤»
moritz jnthn: so if I want to implement import SomeClass; that would just be a call to do_import?
arnsholt There we go. Thanks =)
moritz people sometimes want to write class A { sub f is export {... } }; import A; f()
arnsholt It's still a bit too long for a sigline though 10:08
jnthn moritz: That's where things start getting murkier. The "is export" trait is also meant to install things in the export table of all pakcages we're nested in, finally cascading up to the top level UNIT::EXPORT
10:08 eviltwin_b joined, geekosaur left
jnthn moritz: EXPORT_SYMBOL does not do that yet. 10:08
moritz jnthn: is there a convenient way to walk the outer chain of a package? 10:09
arnsholt r: role W[$s] { method q { say $s; nextsame } }; ([but] class {}, W[$_] for <hacker Perl6 another Just>).q
p6eval rakudo 2d2584: OUTPUT«Method 'q' not found for invocant of class 'List'␤ in block <anon> at /tmp/V6zOZ1ONNe:1␤␤»
jnthn moritz: Not at the moment, but it should be easy to create/maintain one. 10:11
10:11 Guest96768 left
moritz jnthn: allright. Now I have stuff to think about and hack on :-) 10:11
jnthn moritz: Just push onto it pkg_create_mo and pop from it in pkg_compose, in World. :)
Yay :) 10:12
Oooh, pmichaud++ fixed a bug while I was sleeping too :) 10:14
10:16 geekosaur joined 10:19 eviltwin_b left, am0c left
moritz does $*W.add_object adds it to the list of objects to be serialized? 10:21
10:27 plobsing joined
moritz r: use Test; use Test; 10:27
p6eval rakudo 2d2584: OUTPUT«===SORRY!===␤Cannot import symbol '&plan', since it already exists in the lexpad␤»
moritz r: use Test; say &ok.package.^name 10:28
p6eval rakudo 2d2584: OUTPUT«Test␤»
sjn that message should read "Cannot import symbol '&plan' from package 'Test', ...(etc)" 10:30
moritz sjn: that's what I'm trying to do :-) 10:31
sjn awesome \o/
10:34 stepnem left
jnthn suspects an extra arg to $*W.import(...) may be the best way 10:34
Since the error most usefully will have what the user wrote in their use/import statement 10:35
&ok.package.^name approaches would lead to a less useful error in the case of, say, re-exporting.
moritz right
sjn what's $*W, btw? 10:36
10:37 geekosaur left, eviltwin_b joined
moritz the World, ie the place where all the declarations live 10:37
sjn sounds scary
and big
jnthn sjn: Grammar: the syntax. Actions: the execution/semantics. World: the declarations. 10:38
It's just separation of concerns :) 10:39
moritz sjn: it's scary and big, but letting all that stuff floating around without abstraction would be much more scary and big :-) 10:40
jnthn Yes. We've already tried that. :) 10:41
estrai how can one access $*W? I tried this C<< perl6 -e '$*W.perl.say' >> and it said "Dynamic variable \$*W not found" 10:44
tadzik it's in the compiler itself, not a program being ran 10:45
jnthn Right. Programs aren't meant to access $*W. It's a compiler implementation detail. 10:48
(Though a quite important one to know about if you're going to work on Rakudo.)
10:51 Veeru joined 10:52 Veeru left 10:57 sreejit left
fglock mucker: I can reproduce the blockoid error now; I don't think it will work in rakudo though, because of the "Redeclaration of symbol" errors (perlito depends on open classes to work) 11:02
the would take quite some time to rewrite
mucker ok :)
glad to help out if its simple enough 11:03
11:03 JimmyZ joined 11:04 geekosaur joined, sisar joined, brrt left 11:05 Guest50704 left
jnthn r: class Foo { }; augment class Foo { } 11:05
p6eval rakudo 2d2584: OUTPUT«===SORRY!===␤augment not allowed without 'use MONKEY_TYPING'␤at /tmp/d4edpnkZir:1␤»
jnthn r: use MONKEY_TYPING; class Foo { }; augment class Foo { }
p6eval rakudo 2d2584: ( no output )
jnthn fglock: Rakudo supports open classes like ^ 11:06
sisar tadzik++ 11:07
phenny sisar: 07:37Z <tadzik> tell sisar d00a058586 should fix the warnings, let me know if everything's allright
sisar: 07:38Z <tadzik> tell sisar I'm all for versioning panda, but I wanted to do some bin/panda refactors first
11:07 eviltwin_b left 11:08 geekosaur left 11:12 eoh| left, eoh joined
JimmyZ is there a way to forbid augment class? 11:14
fglock r: method emit_perl6 { @.or_list.>>emit_perl6() }
p6eval rakudo 2d2584: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 2␤»
fglock this is the blockoid error
jnthn JimmyZ: It's forbidden by default; you have to use the pragma to allow it. 11:15
flussence: Did you mean @.or_list>>.emit_perl6() ?
er, sorry
fglock: ^^
11:15 pernatiy joined
jnthn r: method emit_perl6 { @.or_list>>.emit_perl6() } 11:15
p6eval rakudo 2d2584: OUTPUT«Useless declaration of a has-scoped method in mainline␤»
sisar tadzik, yup! No more warnings. 11:16
tadzik \o/
sisar tadzik, how do i remove a module ? 11:17
fglock jnthn: I don't know which class will be loaded first; can I just "augment" a class before declaring?
sisar tadzik: (using panda) 11:18
fglock r: use MONKEY_TYPING; augment class Foo { }; augment class Foo { }
p6eval rakudo 2d2584: OUTPUT«===SORRY!===␤You tried to agument class Foo, but it does not exist␤at /tmp/h27o_OCwam:1␤»
jnthn fglock: Don't think so, however...
r: use MONKEY_TYPING; class Foo { ... }; augment class Foo { } 11:19
p6eval rakudo 2d2584: ( no output )
11:19 stepnem joined
jnthn If you can stub all the classes with the ... syntax first, you can augment them beyond there. 11:19
fglock ah, nice
11:20 xenu joined, xenu is now known as Guest29268
dalek rlito: 6650db6 | (Flavio S. Glock)++ | TODO-perlito6:
Perlito6 - add TODO: rakudo fixes suggested by jnthn++
 11:23
11:26 kaare__ joined, tyatpi left 11:27 geekosaur joined
tadzik sisar: that's Not Yet Implemented :) But I do have plans for that 11:35
11:35 Entonian left
tadzik want to hack on it a bit? Shouldn't be too hard 11:36
11:36 thelazydeveloper left, thelazydeveloper joined 11:38 Entonian joined, ponbiki joined
timotimo i'm looking to write a little tool that gives me an overview over git and hg repositories on my two computers as well as bitbucket and github. is there a module that helps me handle ssh in perl6 or do i have to do everything with "shell execution quotes" and the ssh commandline tool? 11:40
11:41 geekosaur left, eviltwin_b joined
moritz I believe we have stuff like File::Find 11:42
tadzik We don't have anything about ssh 11:43
timotimo OK 11:48
wondering if i should do it in python (which does have those modules that would be helpful) or in perl6 (for practice)
in perl6 it would probably end up as a very glorified shell script :) 11:49
11:50 geekosaur joined 11:53 eviltwin_b left
tadzik use NativeCall, write Net::SSH :) 11:53
JimmyZ Net::expect 11:54
11:56 lichtkind joined
dalek kudo/nom: d337e35 | moritz++ | src/Perl6/ (2 files):
carry around the name of the to-be-imported module, and use it in an error message
 11:57
11:58 sporous left
dalek rlito: f490f1a | (Flavio S. Glock)++ | / (21 files):
Perlito6 - method hyper syntax use >>. instead of .>>
 11:59
12:00 sporous joined 12:01 skids left
colomon tadzik: I'm running emmentaler on niecza ... 12:07
tadzik colomon: \o/
colomon tadzik: mind you, that's literal, no idea how well it works yet
but it's busy trying 12:08
moritz colomon: want to conspire with me to get DBIish backends working on niecza?
colomon ooooo
12:09 geekosaur left, eviltwin_b joined
colomon moritz: would love love love love to, but my tuits may be a bit scarce for a few days.... 12:09
moritz colomon: I primarily need (1) some pointers on how to install modules on mono
C# based modules, that is 12:10
(2) do a niecza-specific search paths
(3) some pointers on how to marshall arguments from/to .NET land
colomon moritz: ah, seems to me the actual issues may be a bit meta
moritz colomon: I'm fine with doing most of the work myself
12:11 mucker left
colomon moritz: that is to say, 1) there's no "official" way to run modules in niecza 2) there's no official way to install both rakudo and niecza modules on the same system 12:12
My current method is to use ~/.niecza for niecza modules 12:13
moritz colomon: it's not just about installing. Something like a NIECZALIB env variable would already help
colomon and support it "manually" by adding -I/Users/colomon/.niecza/lib to my niecza command line
yes, that's true. 12:14
hmmm
12:14 jaldhar left
moritz I've recently added RAKUDOLIB to rakudo for just that purpose 12:15
colomon moritz++ 12:16
12:16 BlueT_ left, BlueT_ joined
colomon moritz: gotta go nom breakfast, but I'll take a peek at the niecza source when I get back 12:16
12:18 geekosaur joined
moritz colomon: but when I want to install like sqlite bindings for C#, how do I do that? 12:20
12:20 eviltwin_b left
colomon moritz: mono handles all that...I don't quite remember the details, but it was pretty straightforward 12:25
moritz colomon: any keywords I can search for? 12:26
colomon: "mono install package" just shows me lots of installation instructions for mono
same with other phrases I tried :/
colomon "global assembly cache" 12:27
www.mono-project.com/Assemblies_and_the_GAC
12:28 JimmyZ left, JimmyZ joined
colomon though system.data.sqlite recommends not using it, for reasons I do not understand 12:29
12:29 MContagious left
moritz colomon: but does that mean I first have to install stuff manually, and then run gacutil to install it? 12:30
or is there some 'install it from monoPAN right away' command?
colomon there's no monoPAN so far as I know
so yeah, that's how I've done it -- build and then install using gacutil 12:31
hmmm, there's other suggestions on that page 12:33
colomon is very close to a complete mono / .NET neophyte 12:34
12:34 spider-mario joined 12:36 thelazydeveloper left
moritz meh, system.data.sqlite comes with a .bat file for building :/ 12:36
colomon tadzik: apparently niecza works better than rakudo! www.harmonyware.com/perl/emmentaler.html ;)
moritz :-) 12:37
12:37 ponbiki left
moritz oh, it seems that mono already comes with Mono.Data.Sqlite 12:38
colomon mono++ 12:40
tadzik colomon: yay, awesome! :)
colomon that should provide a good start for you, I hope
jnthn NativeCall works on Niecza? :)
colomon jnthn: Everything works on Niecza!
including modules that don't work period! 12:41
jnthn Wow!
colomon it's no doubt powered by nanotechnology!
or s-circuits!
or even transistors! 12:42
12:42 thelazydeveloper joined
tadzik *gasp* 12:42
moritz rod logic! 12:44
[Coke] here's a perl 6 irc bot project: write a bot that talks to the people who drop in to say "is perl6 production ready yet?" 12:46
moritz [Coke]++ # nice idea 12:47
ooh, aloha is silent!
[Coke] "What do you mean, $ready?" 12:48
moritz: bacek++ fixed her yesterday.
more like, "spayed"
moritz \o/
flussence
.oO( spayesian filtering )
 12:49
12:50 spaceships joined
colomon tadzik: any chance you can post a really short projects.json for testing purposes? I'm trying to edit down the one I have, but it's giving my editor fits.... 12:53
tadzik colomon: not right now, I'm at $work. But you should be able to 'json_pp < projects.json' and then remove everything but Acme::Meow, or such 12:58
colomon tadzik: no worries, I just figured something out
tadzik okay
colomon tadzik++
13:01 PacoAir joined
colomon oh! 13:03
my problem is that you're depending on Panda.stage, which didn't exist in the version of Panda I forked to make panda-niecza
tadzik ha 13:04
colomon no wonder the results are so good!
tadzik we really shouldn't have separate panda-niecza
once you merge the upstream, could you paste somewhere the diff from my tree to yours?
colomon sure, but we'll still have all the organizational issues to figure out. 13:06
colomon has no clear idea how to merge the upstream...
colomon just found the instructions on github. 13:07
tadzik hm, neither have I :)
colomon help.github.com/fork-a-repo/
tadzik . o O ( panda upstream )
gfldex is there an emacs mode that actually works? 13:08
moritz if no emacs mode would work, people wouldn't use emacs, no?
gfldex i'm quite sure there are plenty of folk that have modes and use emacs 13:09
jnthn I'm guessing gfldex's question has an implicit "well for Perl 6" at the end. ;)
colomon tadzik: conflicts everywhere!
gfldex i have the one from pugs repro and a simple "where 0 < * < 12" breakes it
tadzik colomon: ragefac.es/43 13:11
gfldex: did you try the one listed on perl6.org?
gfldex tadzik: i just did, same result 13:13
pmichaud_ good morning, #perl6
moritz good am, pm_
tadzik good afternoon, pmichaud 13:14
pmichaud "stable" / documentation --- I wonder if it's worthwhile for us to come up with a tool/documentation set that simply copies the parts of the synopsis that Rakudo implements.
13:14 geekosaur left
pmichaud then when someone says "where's the documentation for Rakudo", we point them to that 13:14
when we add a feature, we incorporate that section of the synopsis into the "Rakudo set" 13:15
13:15 geekosaur joined
jnthn hi, pmichaud o/ 13:15
Hmm. Not a bad idea. :)
pmichaud if we want to know what pieces aren't implemented, we do a diff between the rakudo set and the synopses
jnthn wonders how much effort it'd be to keep it sync'd up :)
moritz I'm kinda skeptical
[Coke] make the tool like roast where multiple implementations can mark something as "WFM"
moritz 1) maintenance effort
pmichaud if we incorporate sections by reference as opposed to actually copying text, that could work 13:16
moritz 2) you often need parts of the specs that are poorly implemented to even understand the other stuff
pmichaud I'm not sure that #2 is much of an issue; perhaps I'm wrong there though.
and, of course, we can always say "refer to the master synopses for the full detail set" 13:17
moritz and 3) the spec isn't meant as user docs
pmurias there used to be a version of the spec with links to test results
pmichaud the spec isn't meant as user docs, but it's what we have right now.
13:17 geekosaur left, eviltwin_b joined
pmichaud I don't know when "user docs" will arrive, nor of a plan to deliver them (beyond the "Using Perl 6" book which seems stalled) 13:17
I mean, I have a fair bit of skepticism about this approach also, but it might be worth a shot to see what happens 13:18
it might help out in quite a few user-community issues 13:19
and if the tool is built correctly, it'd be possible for niecza++ and other future p6 implementations to use as well
so yes, similar to roast is done (Coke++)
*to how 13:20
tadzik or maybe we could use the feature matrix instead of roast 13:21
pmichaud the feature matrix is good in that it gives a high-level picture of what is done and isn't done 13:22
but somewhere we need a more detailed description that says "this is exactly what version X of Rakudo supports"
In looking at the synopsis text, I think I might have an idea of how to do a first-pass of what I'm thinking of 13:23
maybe I'll prototype a tool a bit later
might even do it in p6 :-)
pmurias pmichaud: do you remember the smartlinks tool (which seems to have bitrotted) that used to show which parts of the synopsis pass their respective tests 13:24
pmichaud pmurias: yes 13:25
moritz pmurias: it seems to be fine, because it is what generates the HTML version of perlcabal.org/syn/
pmurias: it's just not connected to a smoker right now
pmichaud pmurias: but that tool seems to go the opposite direction of what I'm thinking
there probably is a way to unify synopses + roast tests + implementation feature matrix into a single system, but I don't see the path for that yet. 13:26
13:27 geekosaur joined
arnsholt The first two at least might be amenable to some kind of literate-ish approach, I suppose 13:28
This is the spec, and these are the tests for this bit of spec
pmichaud arnsholt: thus far it's been really difficult to organize roast directly according to the spec (at least, to any greater degree than it is organized now) 13:29
13:30 eviltwin_b left
JimmyZ thinks the speed still is a block 13:30
moritz jnthn: I don't think anybody disagrees with that
erm, meant JimmyZ
sorry, badtab 13:31
pmichaud tab--
colomon tadzik: okay, this merge is sooooo not happening today
tadzik :)
13:31 geekosaur left, eviltwin_b joined
[Coke] # 05/03/2012 - rakudo++ ; niecza (94.32%); pugs (42.23%) 13:34
arnsholt pmichaud: Yeah, that's the hard part
[Coke] no real change in the past few days
[Coke] updates that so it shows how many the winner is passing... 13:35
# 05/04/2012 - rakudo++ (21610) ; niecza (94.32%); pugs (42.23%)
pmichaud [Coke]: are those stats logged somewhere?
ooc? 13:36
[Coke] pmichaud: gist.github.com/1476841#file_perl6_pass_rates 13:37
sjn pmichaud: reimbursement should arrive soon, btw :)
[Coke] that gist is setup to run daily at noonish my time.
sjn pmichaud: sent you the funds yesterday
13:37 JimmyZ_ joined
sjn tadzik: still need the receipts from you 13:38
(and fsergot)
pmichaud sjn: receipts were mailed out yesterday morning
sjn: thanks much for everything
tadzik sjn: right 13:39
13:39 JimmyZ left, kaleem left, JimmyZ_ is now known as JimmyZ
tadzik sjn: still haven't figured how to get non-hungarian wizzair invoice, will take a closer look today 13:39
PerlJam greetings #perl6!
13:40 kaleem joined 13:41 kaleem left
huf tadzik: why hungarian?! 13:41
tadzik huf: hungarian airline
PerlJam tadzik: google translate :) 13:42
jnthn
.oO( Hungarian notation )
 13:43
huf wizzair is hungarian?
hoo
tadzik aye
13:43 bluescreen10 joined
huf well, i'm not surprised then :) it's probably as bad as the country :) 13:44
13:45 sivoais joined
daxim don't be a jerk 13:45
13:45 geekosaur joined
huf what, i live here, i *know* what it's like :() 13:46
13:48 eviltwin_b left 13:51 geekosaur left, eviltwin_b joined 13:53 eviltwin_b is now known as geekosaur 14:03 mikemol joined 14:04 alester joined 14:09 eviltwin_b joined, geekosaur left 14:10 eviltwin_b is now known as geekosaur 14:17 benabik left 14:19 eviltwin_b joined, geekosaur left, benabik joined 14:20 eviltwin_b is now known as geekosaur 14:24 tokuhirom joined 14:26 thou joined 14:27 plobsing left 14:34 wtw left, PacoAir left 14:38 kresike left 14:39 Guest29268 left 14:40 lestrrat left, lestrrat joined
dalek blets: 48da665 | (Herbert Breunung)++ | docs/appendix- (2 files):
backlink smartmatch table
 14:42
14:44 xenu joined, xenu is now known as Guest14743 14:51 raiph joined 14:53 kresike joined 15:02 skids joined 15:06 birdwindupbird left 15:07 NamelessTee joined
dalek ast: f9c119d | moritz++ | S14-roles/parameterized-type.t:
RT #101426, roles as default argument to role parameters
 15:13
moritz that was a typical masakbug :-)
15:14 tokuhirom left
masak hey hey hey 15:17
yes, that one looks like mine :P 15:18
masak admits to being fond of nesting things 15:19
15:19 estrabd joined
jnthn Yo dawg, I heard you like roles... 15:20
15:25 liuchong left 15:26 plobsing joined 15:31 pernatiy left 15:33 kresike left 15:36 pernatiy joined, eviltwin_b joined, geekosaur left, eviltwin_b is now known as geekosaur 15:39 pernatiy left 15:40 pernatiy joined 15:46 mspaulding joined 15:47 mdupont joined 15:51 Patterner left 15:54 Psyche^ joined, geekosaur left, eviltwin_b joined 15:55 eviltwin_b is now known as geekosaur 15:57 JimmyZ left 15:58 eviltwin_b joined
dalek blets: 8348dbd | raiph++ | README.md:
fix typo
 15:58
blets: 02da222 | (It's secret to everybody)++ | README.md:
Merge pull request #4 from raiph/master

fix typo
15:58 geekosaur left, eviltwin_b is now known as geekosaur 16:01 havenn joined
dalek blets: 157eadf | (Herbert Breunung)++ | docs/appendix- (3 files):
fixed formating in G and links context ops
 16:02
blets: f13437e | (Herbert Breunung)++ | README.md:
Merge branch 'master' of github.com:perl6/tablets
16:06 geekosaur left 16:08 geekosaur joined
dalek blets: 4aa68fb | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
harmonize Operator Associativity table with tablets syntax rules
 16:11
16:18 Vlavv` joined 16:19 Vlavv_ left
dalek ast: 73407b7 | moritz++ | S06-multi/proto.t:
RT #111454, our proto visible from the outside
 16:27
16:27 fhelmberger left
moritz wow, the spec is pretty silent on Foo:: being the stash of Foo 16:29
masak yes.
moritz ack only finds two instances of \w::\s
and both refer to built-in ones (COMPILING:: and CALLER::)
jnthn Does it not talk about Foo::<$x> and the line? 16:31
*like
?
That pretty much implies it.
moritz well, it could be that Foo::<..> is a special form
jnthn True.
But it isn't, per STD, which is semi-spec. :) 16:32
16:32 Guest14743 left, Psyche^ is now known as Patterner
TimToady S02:2895 16:32
and scanning for \w::\s when most code is in C<>
doesn't work too good 16:33
16:34 daxim left, fglock_ joined 16:39 geekosaur left 16:42 pernatiy left
dalek ast: d70ee4c | moritz++ | S02-packages/package-lookup.t:
explicit tests for Package:: lookup
 16:42
kudo/nom: 637197e | moritz++ | t/spectest.data:
run new test file
 16:43
16:45 adu joined
dalek kudo/nom: ef715f0 | moritz++ | t/spectest.data:
[t/spectest.data] bring back in asciibetical order

also delete reference to outdated S07 test		 16:46
16:48 PacoAir joined
dalek ecs: c964fd8 | moritz++ | S02-bits.pod:
[S02] use a better example for process globals

  %*PID is not mentioned or explained anywhere else		 16:51
16:54 geekosaur joined 16:56 wooden left 16:58 geekosaur left, eviltwin_b joined 16:59 eviltwin_b is now known as geekosaur 17:01 eviltwin_b joined, geekosaur left 17:02 eviltwin_b is now known as geekosaur
lichtkind moritz: are you sure you last spec commit wasnt a typo? 17:08
phenny lichtkind: 03:52Z <sisar> tell lichtkind ah, thanks for explaining that.
17:08 xenu joined, xenu is now known as Guest55778 17:10 Chillance joined
moritz lichtkind: which part of it? 17:10
oh, it already uses $*UID above 17:11
lichtkind++
dalek ecs: b57a05b | moritz++ | S02-bits.pod:
[S02] do not use the same variable twice as example, moritz-- lichtkind++
 17:12
17:17 eviltwin_b joined, geekosaur left 17:18 eviltwin_b is now known as geekosaur 17:20 noam left, noam joined 17:21 noam left, noam joined 17:23 geekosaur left
lichtkind moritz: also thought about that pull request on specs but im not sure what to make of it 17:23
17:24 orphu joined 17:25 adu left
moritz neither 17:26
17:27 noam left
moritz r: say ::A.new; class A { method Str { 'Hello World' } }; 17:27
p6eval rakudo ef715f: OUTPUT«===SORRY!===␤Could not locate compile-time value for symbol A␤»
17:27 plobsing left, noam joined
moritz it seems that rakudo doesn't support ::Foo to refer to undeclared typenames at all 17:27
17:28 geekosaur joined
lichtkind moritz: yeah but thats a bug request not pull request :) 17:30
moritz well, I don't even know what the desired behavior is, so I don't know if that's a bug 17:31
dalek blets: eb3b5fb | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
double linkfix in Contextualizers op table
 17:33
17:37 birdwindupbird joined
jnthn r: say ::('A').new; class A { method Str { 'Hello World' } }; 17:38
p6eval rakudo ef715f: OUTPUT«A.new()␤»
jnthn Can always be explicit. :)
r: say ::<A>.new; class A { method Str { 'Hello World' } };
p6eval rakudo ef715f: OUTPUT«A.new()␤»
jnthn Or like that.
jnthn is happy enough with ::A meaning "introduce a type variable" in contexts where it makes sense to do so, and not meaning something in others, fwiw. 17:39
masak +1 17:40
17:40 Entonian left
jnthn If we can settle on that, we can even give the above an awesomer error saying just that. :) 17:40
17:43 geekosaur left, Entonian joined
dalek ast: d395085 | moritz++ | S04- (2 files):
rakudo unfudges
 17:47
moritz autounfudge++ 17:48
17:52 geekosaur joined
dalek blets: fa87451 | (Herbert Breunung)++ | docs/report.pl:
report also heading number
 17:54
colomon could really use given / when in his current bit of C++ code 17:55
dalek ast: a65f9a2 | moritz++ | S02-names/caller.t:
fudge caller.t for rakudo
 17:58
kudo/nom: bc520af | moritz++ | t/spectest.data:
run caller.t test
18:00 geekosaur left, eviltwin_b joined 18:01 eviltwin_b is now known as geekosaur 18:03 Entonian left
moritz r: say 'f fo foo fooo' ~~ m:2nd/fo+/ 18:03
p6eval rakudo ef715f: OUTPUT«=> <foo>␤␤»
moritz r: say 'f fo foo fooo' ~~ m:2th/fo+/
p6eval rakudo ef715f: OUTPUT«=> <fo>␤␤»
18:03 Celelibi joined
Celelibi hi 18:04
jnthn r: my sub modify { $CALLER::foo++ }; my $foo is dynamic = 42; modify; say $foo
p6eval rakudo ef715f: OUTPUT«43␤»
moritz hello Celelibi
jnthn Thought so...
Celelibi Is there an almost complete implementation of perl6?
jnthn takes a closer look at why #?rakudo todo 'CALLER + rw'
moritz Celelibi: there are two, for some values of "almost"
Celelibi: both rakudo and niecza are nice to use at this point 18:05
Celelibi humhum... what is paroot? 18:06
parrot*
moritz Celelibi: a virtual machine, on which rakudo runs
Celelibi ok
masak hi, Celelibi 18:07
Celelibi hi masak
BTW, perl6 looks pretty amazing. :)
dalek blets: ee78585 | (Herbert Breunung)++ | docs/report.pl:
use smarter formating
moritz Celelibi: glad you think so. I agree :-) 18:08
jnthn moritz: lives_ok { modify() }, ...
moritz jnthn: eeks.
jnthn moritz: That ain't gonna work because now the CALLER is that closure, not the place where my $foo is declared.
masak Celelibi: Perl 6 is amazing. most people who come in here are amazed. :)
moritz jnthn: right
masak Celelibi: can I demo some things to you?
Celelibi :)
I have a question before. 18:09
masak s/can/may # before jnthn++ corrects me :P
Celelibi: ask it.
jnthn rakudo: say "I can AND may be demonstrated"
p6eval rakudo ef715f: OUTPUT«I can AND may be demonstrated␤»
Celelibi are the operators used with the [] reduce meta-operator forced to be endomorphisms?
masak nope.
the ones used with the = meta-operator are, though. 18:10
for some value of "endo" and "morphism" :P
Celelibi: are you a Haskellite?
Celelibi nop.
moritz r: say 'f fo foo fooo' ~~ m:2rd/fo+/
p6eval rakudo ef715f: OUTPUT«=> <foo>␤␤»
jnthn moritz: I'm dinnering now, feel free to fix, or I can diddle it later :) 18:11
Celelibi The only functionnal language I used is O'Caml.
18:11 eviltwin_b joined, geekosaur left
masak Celelibi: oh, that makes sense too. 18:11
Celelibi (and the functionnal features of perl5)
18:11 eviltwin_b is now known as geekosaur
masak I was wondering why someone would think of endomorphisms when hearing about meta-ops. 18:11
Celelibi Just the math word that describe best what I have in mind. :) 18:12
18:12 geekosaur left
moritz r: say 'f fo foo foo' ~~ m:3rd/ <ident> / 18:13
p6eval rakudo bc520a: OUTPUT«=> <foo>␤ ident => <foo>␤␤»
Celelibi If I want (for example) to make a tree out of a list (using some AVL or so) can I use the meta-operator [] ?
dalek blets: 7952320 | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
Contextualizers are prefix, not infix
moritz r: say ([=>] 1, 2, 3, 4 ).perl 18:14
p6eval rakudo bc520a: OUTPUT«1 => 2 => 3 => 4␤»
moritz meh, Pair .perl fail
18:14 geekosaur joined, estrabd left
masak moritz: istn't that right, though? 18:14
moritz r: say ((a => 1) => 2).perl 18:15
p6eval rakudo bc520a: OUTPUT«"a" => 1 => 2␤»
masak moritz: I mean, that's what you'd expect if [] honors direction of associativity.
moritz r: say (a => (1 => 2)).perl
p6eval rakudo bc520a: OUTPUT«"a" => 1 => 2␤»
masak oh, that is fail indeed.
moritz masak: .perl doesn't know anything about associativity
masak moritz: rightright, but that doesn't mean that the first output was fail.
18:15 adu joined
masak because if all reduce does (conceptually) is replace the commas with => ops, that's not a bad representation of the result. 18:16
18:18 eoh left, eoh joined
Celelibi what does the => operator here ? 18:18
18:18 eoh left
moritz Celelibi: construct a Pair 18:18
masak Celelibi: it's a Pair construc... what moritz++ said.
dalek ast: 672ef55 | moritz++ | S05-modifier/counted.t:
remove lots of wrong :x() tests
 18:19
18:19 mtk left
colomon moritz++ 18:20
moritz those tests were just Oh So Wrong
masak \o/
dalek kudo/nom: 8fff4e7 | moritz++ | t/spectest.data:
run counted.t test file
Celelibi All these bots lines are quite annoying...
tadzik 'evening
moritz they tested that :x(3) matched, but expected $/ to hold only one match
tadzik Celelibi: they are also very helpful to developers. YOu are free to ignore them if you mind them :) 18:21
moritz Celelibi: /ignore dalek should help
colomon n: say ([=>] 1, 2, 3, 4).perl
p6eval niecza v17-2-gb8fcf85: OUTPUT«1 => 2 => 3 => 4␤»
masak moritz: reading the first few ones. what was wrong about those?
ok($data ~~ m:1th/fo+/, 'Match 1th'); 18:22
colomon n: say ([R=>] 1, 2, 3, 4).perl
p6eval niecza v17-2-gb8fcf85: OUTPUT«4 => 3 => 2 => 1␤»
masak looks kosher to me.
moritz masak: there's no :th pair
masak: there's just :nth
masak you mean, there's no :th pair *any more*?
there certainly used to be one.
:st :nd :rd and :th 18:23
moritz oh wait
masak and they all mapped back to :nth.
when did that change?
moritz moritz--
I just grepped for examples, and found none
masak and if it did, can we remove the silly rule about :<digit>pair
moritz second moritz-grep fail
masak because that's the only reason it's there, for :2nd and its ilk.
sounds like a revert is upcoming ;) 18:24
moritz a parital one
the :x tests were still wrong
masak aye.
that's because no-one except pmichaud understands :x :P
18:24 mtk joined
masak I don't understand it, and I've re-implemented PGE from scratch once. 18:24
moritz akshually I understand it, and I have implemented it in rakudo 18:25
masak oh, good. then there's two of you ;)
moritz want an explanation?
tadzik seen sisar 18:26
aloha sisar was last seen in #perl6 7 hours 8 mins ago saying "tadzik: (using panda)".
18:26 plobsing joined
sisar tadzik, yes ? 18:26
tadzik oh, I feared that "wanna hack?" scared you :P
so, wanna hack on Panda?
sisar yeah, actually i did look at the source code of panda. 18:27
masak moritz: just tell me this, please: how does :x relate to :nth and :g ?
sisar looked kinda understandable
tadzik I have sort of a plan for uninstalling panda modules, which should be quite easy to bring to dayling
s/ng/ght/
I'm glad to hear that :) 18:28
moritz masak: all of :x, :nth and :g make the regex engine search for more than one match
tadzik so the first thing we need is to keep track of the files we install
sisar ok, go on
masak moritz: go on.
moritz masak: both :nth and :x filter the matches afterwards, and possibly fail the whole thing
masak oh!
tadzik so I guess Installer.pm has to take care of a file with all the files it installs, and to what modules they belong
masak so, what's :nth(2) :x(3) mean? 18:29
moritz masak: a match that always fails
tadzik I'm all for some dead-simple format here, something along Module::Name\nfile\nfile\n\nAnother::Module... and so on
masak moritz: why?
moritz masak: because the :nth(2) filter only leaves the second match, and then there can be no 3 matches
tadzik once we know that, we can think about adding an uninstall method to Panda::Installer, which will remove files and the database entry
masak moritz: so you're saying, :nth filers first, and then :x checks for multiplicity. 18:30
filters*
moritz masak: correct
tadzik my dream, and the reason to have both 'installed' and 'installed-dep' in the .panda/state file is that we should be able to remove unnecessary dependencies after removing the module that pulled them in
moritz masak: though :nths is lazy
masak moritz: why is that behavior... oh.
tadzik in a sortof mark and sweep manner, if you know what I mean
moritz so :nth(1, 3 ... *) :x(5) can match, and stops when 5 filtered matches are found 18:31
18:31 geekosaur left, eviltwin_b joined
tadzik a thing to keep in mind is that we probably want to skip updating the database if the modules are installed to some temporary location; imagine a debian packager who does panda install --destdir=/temporary/location 18:31
masak moritz: this makes sense.
18:31 eviltwin_b is now known as geekosaur
tadzik sisar: does that make sense so far? 18:31
masak moritz: and :g never disallows a match, right? only :x does? 18:32
moritz: and I guess :nth can, too.
18:32 ponbiki joined
moritz masak: correct 18:32
18:32 yohan joined
masak moritz: does :x(*) mean :g? 18:32
sisar tadzik: yeah i get it, but... i'm a bit tired today. I'll think about what you said, and ask you more questions tomorrow. Ok? 18:33
tadzik sisar: sure
yohan hi i heard on perl6 today, is it a better perl5? or a separate language for different uses?
moritz masak: I think so. Not sure if we support :x(*)
tadzik I'm going for a dogwalk right now, and will probably do some panda hacking later, but not related to module installation
sisar tadzik: great. Tomorrow some panda hacking !
tadzik \o/
18:33 mtk left
masak tadzik: dog walk? where are you? :) 18:33
sisar good night #perl6
sisar zzz... 18:34
tadzik masak: in Izabelin, near the forest :)
moritz yohan: better for some uses, worse for others
tadzik sisar: good knight. Dream of pandas :)
masak tadzik: oh, just like I thought!
tadzik: wave hello to Pando from me!
er, Panda*
18:34 mtk joined
tadzik masak: I was actually considering getting a dog recently, a real one, not the toy I have now ;) 18:34
yohan moritz: where should/shouldn't it be used?
sorear good * #perl6
masak knows the name, just can't spell
tadzik but I'm far to seldom in my flat to care about it properly
hello sorear
masak sorear! \o/ 18:35
colomon o/ 18:37
dogs++
moritz yohan: sorry, have to leave for dinner; I hope somebody else answers you 18:38
yohan moritz: oh sure enjoy your dinner 18:39
masak yohan: let me try and assist you. 18:40
yohan masak: that would be great 18:41
masak yohan: we can't answer "yes, it's a better Perl 5", because that's not the political party line.
and in fact, I wonder if that answer isn't giving the wrong impression anyway. 18:42
Perl 6 isn't what you get if you take Perl 5 and make it better.
Perl 6 is what you get if you take Perl 5, extract the Perl from the accidents, then purify it for 12 years, and think really hard about how to do things better.
that said, I think Perl 5 is great language. I use it more or less every day. 18:43
yohan masak: when do you usually use perl6?
masak right now, in the first half of 2012, we tend to say this: learn, use, and love Perl 5. you'll never regret being good at it. but if you're an adventorous spirit with a curious heart and a tolerance for pain, do try Perl 6 out. we think it's great. 18:44
sorear yohan: Perl 6 is our attempt to redo Perl in a less awkward way
masak yohan: I usually use Perl 6 when I have plenty of time, don't mind hittin a few snags, and want to explore some ideas with my favorite syntax and semantics.
hitting*
yohan sorear: do you find perl5 awkward? 18:45
masak some parts of it undoubtedly are.
note that that doesn't make Perl 5 useless by any means.
just less-than-perfect sometimes. :)
Perl 6 is less of "Perl 5 plus one" and more of "Perl, again, in new ways". 18:46
spider-mario and it’s not comparable with python 2 vs. 3
adu I'm new to perl, and I'd rather start with perl6 than perl5, are there tutorials for people like me?
masak adu: yes.
spider-mario it’s another level of reengineering :)
masak adu: hold on, getting you links. 18:47
adu I'm familiar with a few dozen other languages, if that helps
lichtkind python 3 brought less new stuff then 5.10 :)
spider-mario :D
masak adu: strangelyconsistent.org/blog/june-1...and-output
yohan as i found by googling perl6 project is more than 10 years old! why is it taking too long to come to live?
masak adu: perl6.org/documentation/
yohan: because people keep expecting it to be as good as Perl 5, and Perl 5 is 25 years old ;) 18:48
well, that lineage is.
sorear yohan: larry and pm got cancer in the middle of those ten year
18:48 havenn left
masak sorear: pm didn't. family of pm's did. 18:48
sorear right, why do I keep making that mistake
adu I've heard perl6 has amazing grammar sublanguage
yohan larry has cancer? oh :(
masak adu: that is true. 18:49
adu: wanna see some?
18:49 kaleem joined
adu masak: yeah 18:49
masak adu: github.com/moritz/json/blob/master...Grammar.pm
pmichaud as far as pm concerned, "family of pm has cancer" is almost equivalent to "pm has cancer", at least in terms of pm tuit supply :)
18:49 kaleem left
masak adu: this little file parses JSON. 18:49
masak hugs pmichaud 18:50
adu masak: is TOP builtin?
18:50 brrt joined
masak adu: no, it's just the default top rule. 18:50
yohan it feels like the perl6 is something like php6.. everybody talks about it but it never becomes a reality..
sorear yohan: AFAIK, Larry is in decent shape now
adu or is it just capitalized so people know which module export to use
pmichaud perl6 is reality now
masak adu: but everything with capital letters is special somehow. 18:51
pmichaud it kind of bugs me that we hear "perl6 never becomes reality" when it's been shipping releases for 4+ years
sorear yohan: has php6 been making monthly preview releases for several eyars now?
adu masak: so you don't have to use JSON::Tiny::Grammar::TOP, just JSON::Tiny::Grammar?
masak yohan: what pmichaud said. it's kind of depressing to us working on Perl 6 to hear people think that it isn't a reality.
adu: right. it dispatches to TOP automatically.
colomon adu: github.com/colomon/ABC/blob/master...Grammar.pm
adu I know it's a reality, I'm not that new 18:52
masak adu: see github.com/moritz/json/blob/master...iny.pm#L50 for how it's done.
yohan forgive me for saying that, but i said it is not a reality because i never saw it being used anywhere!
gfldex folk like to have opinions much more then they like to have a clue
trolls are best to be ignored
pmichaud there was a time when linux "wasn't being used anywhere" 18:53
masak yohan: I've been using Perl 6 to drive my blog for going on two years.
adu I started using parrot about 3 years ago, back when "make install" gave an anoying warnings
pmichaud or sometimes "anywhere" really means "anywhere outside of <insert bubble here>"
I'm not intending to pick on yohan++ here -- I saw this same conversation come up in the overnight logs 18:54
yohan mmm.. i am just trying to convince myself to use it.. 18:55
pmichaud people who say "perl6 doesn't exist" are just denying reality. a far more helpful comment is "okay, perl6 exists, but I can't use it yet because <insert reason here>"
and "<insert reason here>" can be "not enough other people are using it yet"
masak pmichaud: outsiders' discourse always seems binary and black/white to me.
pmichaud masak: yeah, I think people are too busy trying to simplify things into binary yes/no sorts of things 18:56
masak so naturally newbies will have the same thinkos before we corrupt^Wacclimatize them.
pmichaud well, I continue to here it from people that I wouldn't consider to be "newbie" in the Perl community (or even in the context of Perl 6 stuff)
s/here/hear
masak pmichaud: nodnod
yohan i like this community you belive in perl6 very much :) 18:57
adu if that's so, then C# doesn't exist
pmichaud anyway, more errands here
bbl
sorear wait, c# doesn't exist? 18:58
I've been using it regularly for years
yohan when is perl6 going to be production ready? **i think the answer will be.. "when its ready" ;)
spaceships i heard it would be done by christmas
sorear yohan: 'define production'
18:59 geekosaur left, eviltwin_b joined, eviltwin_b is now known as geekosaur
sorear I think we're a long way away from "you could write the next Facebook in Perl6 instead of PHP and it would be fine" 18:59
yohan sorear: when will i be able to use it on my mission critical servers like i do with perl5
masak pmichaud: mst has made me surprised/disappointed in that way a few times. see strangelyconsistent.org/blog/idiomatic-perl-6
sorear yohan: perl6 is already quite *reliable* 19:00
yohan: if your mission critical servers have a low request rate, you may already be able to do that
yohan sorear: do you use it in your own career?
masak ahahaha
sorear I don't have a career yet. Do you know where I could find one?
masak waves to sorear from Sweden 19:01
adu masak: this JSON example is awsome, I've written parsers in Haskell, Python, and Go, and I already see it being easier in Perl6
[Coke] hugme: hug pmichaud
hugme hugs pmichaud
masak adu: there you go, then.
brrt sorear: western europe, mostly
masak adu: if you really want your noodle blown, have a look at STD.pm6
brrt skilled computer programmers are in huge demand right now 19:02
masak yup.
(we're hiring)
yohan sorear: oh i thought you have a day job sorry
brrt masak: who is 'we'?
gfldex yohan: i play an internet spaceship game as an internet spaceship producer. I use perl6 to do the numbers. For my internet spaceship production, Perl 6 is quite ready.
masak brrt: edument.se/
[Coke] (hiring) as are we.
masak brrt: a Perl 6-friendly consulting/teaching startup in southern Sweden. 19:03
brrt seems fun.. i just speak so little of swedish :-p
masak gfldex: that's funny, I have a friend who uses Perl 6 for his Eve statistics ;) 19:04
moritz brrt: that's not too hard to learn, it's a Germanic language too
dalek kudo/nom: cf40f24 | moritz++ | src/core/Str.pm:
[Str.match] add :th as alias for :nth
jnthn You...have to learn Swedish to work at Edument? Nobody told me! :P
gfldex masak: nice, could you ask him if gfldex ever shot him?
masak brrt: in Sweden we speak English quite well, and that's the lingua franca in programming anyway.
dalek ast: 24aff96 | moritz++ | S05-modifier/counted.t:
restore :th tests, masak++
adu hahaha there are 26 precedence levels
masak gfldex: I'll do that.
adu: there used to be fewer ;)
moritz masak: there also used to be more :-) 19:05
masak we've done it all, baby.
19:06 thou left, kaare__ left
sorear adu: do you see those equal signs? you can add your own 19:07
Woodi gfldex: you do T[1234] or you on other side of Launcher ? :)
gfldex Woodi: sorry i dont get you there 19:08
Woodi do you use PLEXes or sell manufacture them ? :)
sorear n: sub infix:<@>($x,$y) is tighter<+> { $y * 5 + 3 }; say 1 @ 2 + 3; say 1 @ 2 * 3 19:09
brrt masak: if you're still hiring whenever i'm finished studying i'll keep it in mind
p6eval niecza v17-2-gb8fcf85: OUTPUT«Potential difficulties:␤ $x is declared but not used at /tmp/5zKkXkHiWj line 1:␤------> sub infix:<@>(⏏$x,$y) is tighter<+> { $y * 5 + 3 }; say␤␤16␤33␤»
gfldex i did not pay CCP for more then 3 years
masak brrt: sounds good.
sorear adu: INFINITE PRECEDENCE LEVELS
19:09 bluescreen10 left
sorear CCP? CCCP? 19:09
adu sorear: wow 19:10
Woodi time is worth something...
masak sorear: CCP makes Eve Online.
sorear: Icelandic company.
Woodi anyway do we have tail calls recursion somewhere in compilers ?
gfldex Woodi: there is a difference between income and profit. I prefere profits.
masak and it's not CCCP, it's СССР. 19:11
:P
Woodi I just hate skilling :)
gfldex Woodi: and since i have crossed 100MSP on two chars quite some time ago, I can turn my time into ISK fairly efficient. 19:12
dalek ast: f70f5c1 | moritz++ | S02-names/caller.t:
fix caller.t

it used lives_ok and dies_ok on stuff that inspected CALLER. jnthn++
jnthn moritz: heh, I'd just opened that in my editor :) 19:13
moritz++
Woodi pls tell what you use Perl6 for ?
sorear Anything, so long as it doesn't need to be fast.
masak code.
Woodi o, as we are about spaceships.... www.luminous-landscape.com/articleI...ge_021.jpg
jnthn mostly uses Perl 6 to implement Perl 6, and to write shiny things to show off at conferences. ;) 19:14
Woodi UI is quite button heavy...
moritz Woodi: I use it for small day-to-day tasks, usually oneliners
Woodi right, but I was asking gfldex for specifics...
tadzik gosh, forest is dark this hour 19:15
yohan jnthn: do you find perl6 more magical than perl5? :) 19:16
dalek blets: 48aa63a | (Herbert Breunung)++ | docs/ (6 files):
adding a FAQ page
19:16 orphu left
lichtkind yohan: different kind of magic :) 19:16
yohan i used to show off with my perl5 one liners :) 19:17
19:17 adu left
jnthn yohan: Depends what you mean by "magical". In a "wow, I can do X in a very neat way" then yes. In a "huh that's semantically spooky" sense, not really...Perl 6 is pretty consistent. :) 19:17
dalek blets: b908633 | (Herbert Breunung)++ | docs/_nav- (2 files):
linkfix for nav header
 19:18
yohan jnthn: so its kinda perl trying to get a pythonic flavour i suppose :) 19:19
masak not really :)
Python is an OK language, but philosophically it puts itself opposite to Perl in many ways. 19:20
to the extent that such differences matter at all, Perl 6 isn't much closer to Python than Perl 5 is.
tadzik wow, we got like 4 new people this evening, if I count correctly
where is this favourite face of mine...
19:20 havenn joined
tadzik ragefac.es/154 19:21
yohan that means perl6 is getting popular
19:21 havenn left
yohan tadzik: its strange how 9gag has affected all the world.. 19:21
19:21 bluescreen10 joined
tadzik yohan: oh, I'm pretty sure this one is way older than 9gag 19:21
dalek blets: 72a97f3 | (Herbert Breunung)++ | docs/appendix-f- (2 files):
forgot to rename appendix F page
 19:22
yohan which means you know of it as well :)
tadzik I've been using it for longer than 9gag exists. Trendy sites come and go, memes are usually old
well, maybe not longer than it exists, as I checked now, but you know the drill 19:23
yohan: and re popularity, yes, I've noticed that too
yohan tadzik: i understand :)
i spend more time on dwarf fortress :) 19:24
tadzik so, what brought you to Perl 6?
yohan i heard of it today while talking to a friend .. i never knew there is another perl project in parallel with perl5 19:25
masak tadzik: I have a feeling the Rakudo Star release is the ultimate cause of the influx. 19:26
tadzik you think so?
masak I think so. 19:27
tadzik it didn't get many downloads as I look at it
19:27 noam left, noam joined
yohan oh rakudo star is available in the fedora repositories cool :) 19:28
19:28 wolfman2000 joined
yohan where can a find a quick tutorial? 19:28
masak tadzik: well, it can still get publicity.
yohan: perl6.org/
yohan oh now i see butterflies! 19:29
colomon fedora's R* probably isn't the most recent, is it?
_sri it would get more downloads if there was a reason to use perl6 for anything else than language research
[Coke] I picture a p6 logo with camelias floating around some poor confused programmer's head. 19:30
_sri: help us out! write a module that you think people will use.
brrt the using perl 6 book is pretty good
_sri [Coke]: i will, as soon as there's a "stable" release with deprecation policy and reference of supported features
masak it could be better. :) and more complete.
yohan hahahaha [Coke] ! 19:31
colomon: 0.0.2011.07_3.6.0-1.fc16
pmichaud _sri: "stable" is meaningless.
_sri pmichaud: then perl6 is not usable yet
pmichaud a reference of supported features: perl6.org/compilers/features
colomon yohan: yeah, the most recent rakudo star is 2012.04
pmichaud I'm working on an even more detailed reference now. 19:32
yohan colomon: oh ! who is the maintainer? he is too lazy :)
_sri you keep asking for big awesome projects to use perl6 but don't provide a stable platform, that's not gonna work
pmichaud _sri: I don't recall asking for big awesome projects to use perl6
_sri: I don't know where you're getting that from.
_sri not you specifically
masak _sri: you keep saying Perl 6 isn't stable but don't give more detailed examples of what you mean. that's not gonna work. 19:33
moritz pmichaud: btw, have you seen gist.github.com/2346494?
pmichaud _sri: what is the "you" you're referring to, then?
_sri masak: deprecation policy and reference of features affected by it
brrt masak: you wrote the perl 6 book?
masak _sri: you're just yet another outsider thinking you know how to "fix" the Perl 6 effort.
pmichaud _sri: I'll be happy to put together a deprecation policy.
masak brrt: I have authored parts of it. 19:34
pmichaud _sri: at this stage, I'll even be happy to follow it
yohan what perl6 book?!
are there published perl6 books?!
masak yohan: github.com/perl6/book
pmichaud _sri: would you be willing to draft a deprecation policy for us, ooc?
masak yohan: not published, yet.
pmichaud or at least tell us what you expect such a policy to have?
_sri masak: alright, then i'll just shut up
moritz yohan: it's published in the sense that we had releases
masak _sri: no no, listen to pmichaud instead :) 19:35
colomon there is a published perl 6 book, but it was VERY premature
masak _sri: he's much more diplomatic than I am.
brrt i read the poker example
(subtypes)
yohan moritz: cool this is great
brrt and i was wondering if i should be scared or in awe
pmichaud I'm not joking or being sarcastic -- it would be a great help to have an "outsider's" view of what the deprecation policy should look like.
masak brrt: both :)
tadzik brrt: the border is thin :)
pmichaud also, I think that Rakudo is indeed "stable" enough to be able to follow a sane deprecation policy
yohan and to build the book you will need perl5 .. ironic :) 19:36
pmichaud so, if lack of a deprecation policy is a barrier for some, let's get one.
masak pmichaud: for a great deal of features, Rakudo is stable enough.
brrt perl5 obviously comes with a lot more modules
like, insanely many
masak we don't have threading/events/concurrency yet.
I don't imagine macros are stable in the spec yet.
pmichaud I do wish to restate my earlier comment that perl 1.0, 2.0, 3.0, and even 4.0 (to some extent) did not have modules.
19:37 eviltwin_b joined, geekosaur left
masak nod. 19:37
pmichaud Modules didn't arise with Perl's birth. Also, Perl was considered "stable" long before it had modules. :-)
19:37 eviltwin_b is now known as geekosaur
pmichaud gotta pick up kids from school -- bl 19:37
*bbl
_sri: I'm quite serious that any input you could give on a deprecation policy would be a big help. 19:38
bb;
bbl
19:39 thou joined
yohan i remember a PHP Girl talking about """'Perl's Unicorn support'"""" :) i am still looking for her 19:42
moritz :-) 19:43
masak aww
yohan www.youtube.com/watch?v=oeghXnAGJV4...re=related
19:43 MayDaniel joined
dalek blets: a446cfe | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
restructure appendix B a bit
 19:43
19:44 ponbiki left
flussence random thought: to be a *good* language, we need to be at least better than Perl 5. To be a *popular* language, we only need to beat PHP at its own game. 19:46
masak uh-oh, Perl v. Python... :)
yohan masak: ;)
timotimo what, perl5 is the lower bound for what's "good languages"? 19:47
masak flussence: to be a *cute* language, we need to have nicer idioms than Ruby. :)
timotimo i don't think perl5 is that bad
masak timotimo: it isn't generally, but for Perl 6, I'd say it is.
timotimo that's an acceptable metric
19:47 sosey joined
sosey hi 19:48
masak sosey: hi!
what is this, September? :P
moritz it's the month after a Star release, it seems :-)
sosey @masak: sosey is my surname, i dont know what it means
19:48 LylePerl joined
yohan All material in this repository is licensed under a CC-by-nc-sa license: <creativecommons.org/licenses/by-nc-sa/2.5/> (attribution, noncommercial, share-alike), unless explicitly stated otherwise. 19:48
19:49 havenn joined
yohan what does that exactly mean? 19:49
moritz yohan: what it says
yohan: or put differently, which part is unclear to you? 19:50
yohan cc-by-nc-sa :)))
Celelibi Is there a Data::Dumper module in perl6 ?
jnthn Celelibi: Built in; just use the .perl method
moritz r: say (1, "foo").perl
p6eval rakudo cf40f2: OUTPUT«(1, "foo")␤»
jnthn r: [1, { a => 2, b => 3}].perl.say
p6eval rakudo cf40f2: OUTPUT«[1, {"a" => 2, "b" => 3}]␤»
Celelibi ok :)
masak yohan: cc-by-nc-sa is just a shorthand for "Creative Commons, Attribution, Non-Commercial, Share-Alike" :)
19:50 sosey left
Celelibi then, what happened to the range operator ? 19:51
yohan ah ok i hate these shortnames
Celelibi r: say (1 .. 3).perl
moritz Celelibi: it now constructs a Range object
p6eval rakudo cf40f2: OUTPUT«1..3␤»
Celelibi ok
moritz r: .say for 1..3
p6eval rakudo cf40f2: OUTPUT«1␤2␤3␤»
Celelibi what's this .say ? 19:52
moritz it's short for $_.say
yohan time to go back to forth.. goodbye
tadzik bye o/
Celelibi you mean every object has a say method ? 19:53
19:53 yohan left, snearch joined
masak the Python guy is so mean. :) 19:53
Celelibi is everything object in perl6 ?
masak Celelibi: yes. even things which are not objects, are objects. 19:54
tadzik objects are objects too
Celelibi ok.
19:54 mikemol left
jnthn Native types are not objects, but those will autobox if you try to use them as an object. 19:54
r: my int $x = 42; $x.say
p6eval rakudo cf40f2: OUTPUT«42␤»
Celelibi Then... Where can I find documentation about the native types wrappers ?
masak Celelibi: S12. 19:55
jnthn Well, guess S02 and S09 specs various aspects of native types too.
masak troo
pmichaud there are at least a few areas in which Perl 6 is "better than" Perl 5. :-) 19:56
oops, scroll error
masak still worth pointing out ;)
pmichaud I was responding to
19:46 <flussence> random thought: to be a *good* language, we need to be at least better than Perl 5. To be a *popular* language, we only need to beat PHP at its own game.
masak there are at least a few areas where Perl 6 beats PHP at its own game, too. but that's more of an all-or-nothing matter, I guess. 19:57
what we need is a PHP-killer Rakudo distro.
but we know that.
tadzik I don't think anything could kill PHP
jnthn tadzik: What if Chuck Norris and PHP got in a fight? 19:58
moritz tadzik: you're right. Even the PHP folks have tried very hard, and have failed :-)
tadzik jnthn: Chuck will probably say "huh, it's dead already" and move on ;)
jnthn ;)
Celelibi Actually I wonder what remains of perl5 in perl6? 19:59
moritz or "I don't fight with children"
brrt i so badly want to chime in with something mean about php
just everything has been said already
tadzik Celelibi: the spirit :)
flussence "server-side AOL"
19:59 adu joined
moritz Celelibi: lots of things. Like 'my' and 'our' for lexical and package variables, sigils, strings with interpolation, m/../, s/// 20:00
Celelibi: subroutines with 'sub', arrays with [], hashes and blocks with {}
Celelibi there are... small syntatic details.
gfldex brrt: you don't need to say anything mean about PHP, the folk guilty of making PHP do all the work for you: www.kb.cert.org/vuls/id/520827 20:01
moritz Celelibi: curious. Most people make a big deal about disliking sigils :-)
Celelibi the more I read about perl6 the more I think I will have to learn it as a new language, and not as an improvment of perl5.
gfldex Celelibi: that's a pretty good idea
masak Celelibi: yes. 20:02
moritz Celelibi: anyway, there lots of perlish ideas (not necessary perl5ish ideas) that are also in p6. TIMTOWTDI, DWIM, safe defaults, the ability to write "baby perl"
Celelibi: and the fact that operators dictate the type of operations, not the type of the operands
masak Celelibi: but I still place Perl 5 and Perl 6 in the same mental compartment in some way, even though they are very much distinct languages with different semantics.
moritz (most dynamic languages do that differently)
brrt gfldex: o good lord
gfldex brrt: keep in mind that there are quite a few sites that have SQL login in the PHP source 20:03
Celelibi Wht about refs? In perl6 ? 20:04
brrt brrt: i know. most, in fact, in some way or another
Celelibi It looks like lists of lists now exists and are not flattened.
masak Celelibi: everything is refs, so you don't see them.
20:04 geekosaur left
masak Celelibi: lists of lists still flatten. 20:04
20:04 geekosaur joined
moritz Celelibi: perl6advent.wordpress.com/2011/12/1...nces-gone/ 20:04
masak r: my @a = 1, 2, (3, 4, 5); say +@a
p6eval rakudo cf40f2: OUTPUT«5␤»
20:05 yohan joined
Celelibi r: (1, 2, (3, 4), 5).perl.say 20:05
p6eval rakudo cf40f2: OUTPUT«(1, 2, (3, 4), 5)␤»
masak well, they didn't flatten yet there.
they will if you assign it to an array.
yohan i tried to build the book,, i got this : Pod::PseudoPod::LaTeX version 1.10165 required--this is only version 1.000 at lib/Perl6BookLatex.pm line 7 .. is there a ready pdf version with latest content?
masak or do some other stuff.
moritz yohan: the pdf in the download section on github is not too much out of date 20:06
yohan: ie there hasn't been much improvement since that last version
brrt is still working on a p6 epub book builder
yohan moritz: i see.. it is strange that my Pod::PseudoPod::LaTeX that comes with fedora 16 is not new enough to build the book 20:07
Celelibi hum... ok. It's not the construction of the data structure that flatten the list of lists, its the assignment to an array.
masak right.
Celelibi But the... Why?
masak :) 20:08
gfldex Celelibi: you may want to watch ontwik.com/perl/perl-6-update/
masak Celelibi: because sometimes we want the non-flattening behavior.
moritz yohan: well, I too find it strange that Fedora insists on shipping such an old version
gfldex Celelibi: or in general wiki.perl6.org/Categorized%20External%20Videos
yohan moritz: fedora 17 will have 1.000 as well
Celelibi why would it be different to use @a than the data itself? No one would expect that.
moritz Celelibi: note that nearly all uses will show the flattened behavior 20:09
Celelibi gfldex: your first video is 1 hour long. I won't watch it now.
moritz r: say (1, 2, (3, 4, 5))[2] 20:10
p6eval rakudo cf40f2: OUTPUT«3␤»
moritz see? it pretends it's falt
*flat
so unless you write .tree or so, you won't even notice the nesting
but it's there if you need it.
20:10 yohan left
flussence r: say $(1, 2, (3, 4, 5))[2] 20:10
p6eval rakudo cf40f2: OUTPUT«3␤»
flussence r: say (1, 2, $(3, 4, 5))[2]
p6eval rakudo cf40f2: OUTPUT«3 4 5␤» 20:11
flussence there
moritz one of the design principles of Perl 6 is not to forget stuff too quickly
in p5, if you do my $x = @a; then $x doesn't know anything about @a except the number of elements
20:11 geekosaur left, eviltwin_b joined
moritz in p6, you have full information of @a in $x, but if you use $x as a number, it works just like in p5 20:11
20:11 eviltwin_b is now known as geekosaur
moritz so, best of all worlds 20:12
Celelibi hum... ok
moritz thinks he has explained those things quite a few times by now :-)
20:14 havenn left 20:16 havenn joined
Celelibi And what's this $ in (1, 2, $(3, 4, 5)) ? 20:16
moritz r: multi trait_mod:<is>($m, :$a!) { multi y(|$) { say $m }; $m.wrap(&y) }; sub foo is a {}; foo
p6eval rakudo cf40f2: OUTPUT«sub+{wrapped} foo() { ... }␤»
flussence Celelibi: that's like saying [] in p5
moritz Celelibi: it's the "I'm a scalar, don't flatten me" marker
Celelibi ok 20:17
flussence r: say (1, 2, $(3, 4, 5)).flat.perl
p6eval rakudo cf40f2: OUTPUT«(1, 2, $(3, 4, 5)).list␤»
moritz star: multi trait_mod:<is>($m, :$a!) { multi y(|$) { say $m }; $m.wrap(&y) }; sub foo is a {}; foo
p6eval star 2012.04: OUTPUT«Null PMC access in find_method('gist')␤ in sub say at src/gen/CORE.setting:6555␤ in sub y at /tmp/zLdGhbLqGY:1␤ in any enter at src/gen/Metamodel.pm:2685␤ in method postcircumfix:<( )> at src/gen/CORE.setting:1601␤ in <anon> at src/gen/BOOTSTRAP.pm:815␤ in any…
moritz star: multi modify($m) { multi y(*@) { say $m }; $m.wrap(&y) }; modify sub foo { }; foo 20:18
p6eval star 2012.04: ( no output )
moritz hm, seems I really need a trait to reproduce that bug :/ 20:19
star: multi modify($m) { multi y(*@) { $m }; $m.wrap(&y)}; modify sub foo { }; say foo() ~~ Callable
p6eval star 2012.04: OUTPUT«False␤»
moritz star: multi trait_mod:<is>($m, :$a!) { multi y(|$) { $m }; $m.wrap(&y) }; sub foo is a {}; say foo() ~~ Callable 20:20
p6eval star 2012.04: OUTPUT«False␤»
moritz and I need say() too?
20:21 geekosaur left, eviltwin_b joined
moritz star: sub identity($x) { $x }; multi trait_mod:<is>($m, :$a!) { multi y(|$) { identity($m) }; $m.wrap(&y) }; sub foo is a {}; say foo() ~~ Callable 20:21
p6eval star 2012.04: OUTPUT«Nominal type check failed for parameter '$x'; expected Any but got Mu instead␤ in sub identity at /tmp/aOztWShfn1:1␤ in sub y at /tmp/aOztWShfn1:1␤ in any enter at src/gen/Metamodel.pm:2685␤ in method postcircumfix:<( )> at src/gen/CORE.setting:1601␤ in <anon> …
20:21 eviltwin_b is now known as geekosaur
moritz too tired to test and close rt.perl.org/rt3/Ticket/Display.html?id=112664 20:22
masak g'ah, Perl guy in video knocks Esperanto. :/ 20:23
implicitly ridiculing it as "purist". wtf... :) 20:24
20:25 raiph left, birdwindupbird left, geekosaur left, eviltwin_b joined, uvtc joined 20:26 eviltwin_b is now known as geekosaur 20:27 plobsing left
rurban FWIW Building perl6 with AddressSanitizer needs max 4.8G RAM (64-bit DEBUGGING src/gen/CORE.setting). 20:31
20:32 fglock_ left
masak Python guy's #1 reason for Python being better is "It's available". even people who know better use binary black/white rhethoric. :/ 20:33
20:34 eviltwin_b joined, geekosaur left
masak I'll also not that in general, the Perl guy focused on the strengths of Perl, whereas the Python guy kept coming back to how Python beats Perl in various ways. 20:34
of course, it's in line with Perl being Python's "nemesis language" -- but the reverse is not true. 20:35
20:35 eviltwin_b is now known as geekosaur
Celelibi Who cares about that actually? :) 20:35
I started to learn python last tuesday.
(But sorry, I won't drink pepsi, Coke FTW! :p) 20:36
dalek p/toqast: 6f91998 | jnthn++ | src/QAST/Block.nqp:
Generate compilation-unit unique IDs for QAST::Blocks (or allow explicit setting of one).
 20:37
p/toqast: feb325c | jnthn++ | src/QAST/Compiler.nqp:
Code-gen for outer chain.
p/toqast: 5f34bad | jnthn++ | src/QAST/Compiler.nqp:
Generate with :anon flag to avoid a Parrot namespace installation.
sjn ooh! *binary* black/white rethoric! Does that mean there's no middle-ground there too? How about grey areas? :D
THAT'S AWSOME/HORRIBLE! :)
20:37 MayDaniel left
Celelibi If you can actually merge pepsi and coke and make something edible, you can't really merge Perl and Python and make something runnable. :) 20:38
20:38 pmurias_ joined
sjn wonders what the "Pepsi" programming language would like 20:39
Celelibi Where is your grey now? :p
sjn would look* like
sjn is too late at $work, need to get home (brain apparently turned off) 20:40
masak sjn: I would be surprised if there weren't at least two languages out there called "Pepsi".
it's just too good a name to pass up.
sjn Pepsi: The next "P" language! 20:41
Celelibi this website doesn't know them : 99-bottles-of-beer.net/
uvtc masak, what's this in regards to: " I'll also not that in general, the Perl guy focused on the strengths of Perl, whereas the Python guy"? 20:42
masak er, note*
20:42 pmurias left
masak uvtc: a video that I just saw that was posted in the backlog. 20:42
uvtc: www.youtube.com/watch?v=oeghXnAGJV4 20:43
uvtc masak, ah. missed th...
well there it is.
oh yeah, I remember that.
masak, You answered _sri a couple of days ago: irclog.perlgeek.de/perl6/2012-05-02#i_5530090 . Then earlier today you wrote irclog.perlgeek.de/perl6/2012-05-04#i_5542058 . 20:49
dalek blets: dcf63ba | (Herbert Breunung)++ | docs/index.txt:
polish index
 20:50
uvtc If there's no "good example" of what a Perl 6 module should look like, then potential module authors are going to take that as a sign that Perl 6 is not yet ready for them to be writing modules for it. 20:51
jnthn lichtkind: Will there be a czech index too?
uvtc masak, and you're going to get those "not yet a reality" sentiments.
pmichaud surely modules.perl6.org has some "good examples" ?
lichtkind jnthn: to vis ze ne
pmichaud or, I suppose that we expect "best practices" to be established before there's been a lot of practice. :-/ 20:52
uvtc You've gotta start somewhere.
pmichaud I think my point is that the start has already occurred. modules.perl6.org 20:53
uvtc pmichaud, I wrote this a few days ago: wiki.perl6.org/Create%20and%20Distr...%20Modules
pmichaud uvtc++ # excellent!
uvtc pmichaud, it's actually already linked to from modules.perl6.org. 20:54
pmichaud I wonder if that page would be better maintained in the modules repo itself
instead of on the perl6 wiki
come to think of it... what's the origin of the Perl 6 wiki? I don't think I've ever really seen that before. 20:55
s/Perl 6 wiki/wiki.perl6.org/
masak uvtc: I was saying to _sri that we don't have distributions. 20:56
uvtc: we do have plenty of modules. but they're just that: modules.
we haven't really bootstrapped our way up to distributions yet.
uvtc pmichaud, A few days ago, I linked to a previous version of that article, and mentioned that if there was a wiki, I'd put it there, but I didn't want to use SocialText.
masak (well, Rakudo Star is a distribution, but that's a different kind)
uvtc A number of others chimed in about not really being interested in the previous wiki. Some (like me) liked the idea of writing in markdown. 20:57
pmichaud, anyway, long story short, the wiki software is very nice, easy to install (moritz set it up), and uses a particularly nice flavor of markdown (pandoc-markdown).
pmichaud uvtc: I agree that the previous wiki was very un-usable. 20:58
uvtc masak, when you write, "we haven't really bootstrapped our way up to distributions yet.", you mean module distributions?
20:58 noam left 20:59 noam joined
pmichaud uvtc: by "module distribution" do you perhaps mean "module releases"? 20:59
uvtc pmichaud, most everyone polled agreed. And markdown seemed popular.
pmichaud uvtc: my comment is more along the lines of "we tend to use github for all of our other shared projects.... why a wiki at all?"
although the ability to make links between pages is likely a big selling feature of wiki over github 21:00
anyway, the "create and distribute modules" page looks to me like something that is easier to maintain as a part of the perl6/modules repository than separately in the wiki. but that's just me.
dalek p/toqast: 7de618b | jnthn++ | src/QAST/ (2 files):
First crack at code-gen for :op('call').
uvtc pmichaud, the wiki that was set up uses gitit. gitit uses pandoc to process the markdown into html, and supports important enhancements making it suitable for writing larger docs (such as tables, definition lists, ...) 21:01
21:01 skids left
uvtc pmichaud, gitit and pandoc are a pretty amazing combination. I sold moritz on it. 21:01
21:02 geekosaur left, eviltwin_b joined, eviltwin_b is now known as geekosaur
uvtc pmichaud, I even started writing some documentation there: wiki.perl6.org/Mostly%20Harmless%20Docs/Main 21:02
pmichaud uvtc: okay, I'll look into it a bit more.
uvtc pmichaud, the way I see it now though, the Perl 6 effort is sending 2 pretty strong signals that it's "not yet ready" for everyone to start using. 21:04
21:04 geekosaur left, eviltwin_b joined 21:05 eviltwin_b is now known as geekosaur
uvtc it. Using it. 21:05
pmichaud waits for the description of the two strong signals
uvtc Oh, Ok. One is the issue I just remarked about above, re masak and modules. 21:06
pmichaud can you summarize that issue? I didn't get a clear understanding of it.
uvtc If potential module authors are told, "no real standard yet", it's sending the signal, "not yet done".
pmichaud standards should be drafted after experience is obtained, not before. 21:07
see: Internet standards
uvtc Unfortunately, if you don't give potential module authors *something*, it sends the signal that Perl 6 is not yet ready.
21:07 geekosaur left, eviltwin_b joined
pmichaud fair enough, what should that something look like? (more) 21:08
21:08 eviltwin_b is now known as geekosaur
pmichaud it seems to me that wiki.perl6.org/Create%20and%20Distr...%20Modules is pretty much that something, yes? 21:08
moritz I thought we've long had something
and it's been linked to from modules.perl6.org for quite some time too
what exactly is the problem now?
uvtc The problem is that someone came here asking "what's a good example module", and there was no answer. That sends the signal "not yet ready to contribute modules". 21:09
pmichaud did someone at least suggest looking at modules.perl6.org ?
masak yes. 21:10
moritz uvtc: example module != instructions
pmichaud what kind of answer would you like to have seen that would indicate "ready to contribut modules"?
*contribute
moritz in all honesty, when somebody asked me what a good example p5 module was, I wouldn't have any idea either
uvtc pmichaud, "Here's the document explaining how to do it, and here's an example. Get to it!". 21:11
Create a trivial example module if you have to.
pmichaud uvtc: so, you're saying it would have to be a specific module example, as opposed to pointing to the set of available modules?
uvtc: what answer would you give for Perl 5 for that question, ooc?
"What's a good example module for Perl 5?"
masak Acme::Bleach.
:P 21:12
moritz Acme::EyeDrops
pmichaud (I'm asking because the answer will guide us into what we ought to provide that would be at least familiar to P5 module authors.)
uvtc (sorry, just a sec, {family})
pmichaud I'm looking through learn.perl.org for "What's a good example module?" and not finding anything that really answers that. :-/ 21:14
I have to go for a while again... but a comment on 21:16
moritz maybe Date::Simple, if you want something with a bit of XS and a bit of perl too
pmichaud 21:11 <uvtc> pmichaud, "Here's the document explaining how to do it, and here's an example. Get to it!".
okay, we have a document explaining how to do it, at wiki.perl6.org/Create%20and%20Distr...20Modules.
for the other part, we need a line that says "A good example module is <modulename>" in the modules.perl6.org page. 21:17
or there can be a badge that indicates a module particularly well suited for understanding
uvtc pmichaud, what I was saying is that, if a user asks on IRC "hey, any good examples", there needs to be one to point them to. A link in that module creation doc would be fine.
pmichaud I think a link on the modules page itself would be good, too. 21:18
I also wonder if there are any good p6 examples in github.com/perl6/perl6-examples
s/p6/modules/
uvtc Otherwise they leave and think, "geez, it's been **10 years** and no one can even point me to a good example module?! Bah."
pmichaud uvtc: nobody's pointed me at a good p5 module yet.
"maybe Date::Simple" 21:19
is the best I have so far.
uvtc pmichaud, in P5 land, I think there's dist::zilla, no?
pmichaud you think dist::zilla is a good example module for people to learn how to write modules?
uvtc I mean this: Perl 6 needs module authors. They either have to create a module from scratch or use a tool to create one. Perl 5 has tools for this. Perl 6 doesn't yet. 21:20
rurban Dist::Zilla is cleartly the worst of all
pmichaud dist::zilla doesn't strike me as being a good "example module" for newbies... but perhaps I don't understand it
rurban Numbers to verify this at my latest blog post
uvtc I don't mean to use dist::zilla as an example. My understanding was that it was a good idea to use it, but I don't use a lot of Perl 5 these days. 21:21
rurban blogs.perl.org/users/rurban/2012/04...sizes.html
pmichaud uvtc: my point is that nobody here seems to be able to >quickly< answer the question of "What's a good example module for Perl 5?" either
lichtkind jnthn: but i already pondered to use more words like germing hungry..
uvtc pmichaud, right. And I think that's a problem, and it sends a signal to potential users (module authors). 21:22
pmichaud if p5 doesn't have a good answer for it, I'm not sure it's something that can be quickly answered for p6 21:23
rurban search.cpan.org/dist/Task-Kensho/ is a typical recommendation for modern perl
pmichaud not to say we can't work on it, but it's also a little disingenuous to say "p6 should already have these answers ready" when the p5 universe doesn't seem to have them.
uvtc pmichaud, ok. I think the documentation on it is much more important than the example anyway. I'm I'm glad it's there. I only hope it can be kept updated and that new users will be pointed to it. 21:24
rurban www.enlightenedperl.org/ 21:25
pmichaud thus back to my original response: "Did someone at least suggest looking at modules.perl6.org?"
If yes, then that page contains the link that explains how to create modules, as well as providing some modules to look at.
21:26 plobsing joined
uvtc pmichaud, no, I don't think so. 21:27
pmichaud you don't think someone answered modules.perl6.org, you mean? 21:28
uvtc pmichaud, it starts here at irclog.perlgeek.de/perl6/2012-05-02#i_5530090 . 21:29
21:30 havenn left
uvtc pmichaud, The other signal which Perl 6 sends (which suggests it's not ready/reality) is the lack of user docs. I realize this has come up many times before, but regardless, it's a signal to new users. 21:30
pmichaud uvtc: at irclog.perlgeek.de/perl6/2012-05-02#i_5529240, moritz++ had already suggested modules.perl6.org to _sri
perhaps it needed to be repeated at _5530090 21:31
the lack of user docs is due to a lack of user doc authors. I'm not sure how to resolve that.
althouch lichtkind++ and a few others are working on some pieces of that puzzle 21:32
I'm sure a userdoc champion for perl 6 exists somewhere; we just don't know who it is yet (and the champion likely doesn't know yet either) 21:34
21:34 geekosaur left, geekosaur joined
pmichaud still, I think that "lack of user documentation" is an entirely valid complaint. Our current team doesn't seem to have that slot filled I guess. 21:35
s/complaint/criticism/
uvtc pmichaud, regarding how to resolve the user docs issue, I don't know how to resolve it either. :( But here's been my experience: I started writing those "mostly harmless" docs I linked to earlier. My problem though was that the Synopses are very long, detailed, and dense. They are not tutorial in nature. So it was either toil away trying to understand them, or ask questions here...
21:36 brrt left
lichtkind pmichaud: im working on al of it :) just today added a FAQ section 21:37
pmichaud yes, I agree with your experience. many people have expressed similar tales.
uvtc pmichaud, and eventually it annoys people to be barraged with questions.
pmichaud I don't know that that's true on #perl6, though. We generally like questions here, even barrages.
still, it's annoying to the newcomer to have to ask lots of basic questions 21:38
uvtc pmichaud, yes. it makes them feel like they're being annoying.
pmichaud I find it odd (and a little frustrating) that it seems to be hard to come up with good written tutorial examples for Perl 6.
although teaching is/was my profession, written exercises/teaching was never my forte. I'm at my best in a lecture/presentation environment. 21:39
lichtkind uvtc: you can help me restore my perl 6 tutorial
its really easy entry
rurban my parrot github.com/rurban/parrot/commit/1e5f20eefce patch seems to have fixed overflow nopaste.snit.ch/137433
PerlJam pmichaud: I started on another book entitled "Pragmatic Perl 6" that was supposed to be a "get you started in Perl 6" kind of book. It was meant to show how to install the Perl 6 implementations and do useful, practical things (open files, parse data, etc.)
pmichaud: I never got very far with it though 21:40
pmichaud if I had some sort of forum where I could give lectures on Perl 6 and have them videotaped and then transcribed, that would end up making for a good userdoc basis, I suspect.
lichtkind PerlJam: yes its lot of work
pmichaud the problem is that to even get out of the basics takes hours of presentation 21:41
21:41 geekosaur left, eviltwin_b joined
PerlJam I wonder what it would take to get merlyn and bdf to actually finish "Learning Perl 6" 21:41
21:41 eviltwin_b is now known as geekosaur
pmichaud well, there's a bit of a disconnect here worth exploring (more) 21:41
a lot of people come to Perl 6 already knowing a language, so they want a tutorial that lets them get to the cool stuff quickly
21:42 pmurias_ left
pmichaud but something like "Learning Perl 6" tends to start a lot more basic than that, I suspect. 21:42
dalek blets: e5ec4dc | (Herbert Breunung)++ | docs/appendix-a-index.txt:
just enforce new formating rules
 21:43
pmichaud I should also review the tutorial materials that TheDamian++ did in his P6 tutorials at OSCON a couple of years ago
uvtc pmichaud, if you're not up for writing, then perhaps a series of screencasts.
pmichaud uvtc: I've not had much luck with screencasts. I somewhat need an audience to tell me where I'm overlooking stuff. 21:44
although maybe I'm selling myself a little short there.
uvtc pmichaud, I think you may be. :)
PerlJam szabgab did some screencasts ... maybe use them for inspiration?
pmichaud anyway, even as screencasts I suspect it's on the order of many weeks of work just to get out of the basics
PerlJam (or get him to do more)
uvtc pmichaud, perhaps add examples to wiki.perl6.org/Mostly%20Harmless%20Docs/Main ? 21:45
I'd love to see people add to that.
21:45 eviltwin_b joined, geekosaur left
pmichaud I'd like to see that page and perl6-examples unified better, maybe. 21:45
21:45 eviltwin_b is now known as geekosaur
pmichaud however, that page already has "are looking for lots of examples, see perl6-examples" 21:46
PerlJam pmichaud: why do you think it takes so long to get out of the basics?
pmichaud PerlJam: Nature of the beast. I suspect that's true for any language, not just p6.
PerlJam: think about how many college-level lectures it takes before students are able to write programs in a new (to them) language 21:47
PerlJam pmichaud: It seems to me that with some judicious omission you could wend a path out of the basics relatively quickly, then introduce the omitted concepts later when you're covering something that needs them.
uvtc pmichaud, *Right*. I was hoping for the mostly harmless docs to be a distillation of examples that someone could work through in a reasonable amount of time and end up with a grasp of the basics.
PerlJam,++ 21:48
pmichaud I think that perl6-examples (which was somewhat stale until the oslo hackathon) is aiming for the same sort of thing
uvtc PerlJam++
pmichaud PerlJam: "judicious omission" is the really hard part, though, if you don't have an audience handy to let you know when your "judicious omission" ends up being the hurdle
uvtc pmichaud, perl6-examples looks pretty huge to me. You want to be able to point new users to a concise set of examples that will guide them through learning Perl 6. Not just a giant collection of examples.
pmichaud uvtc: I'm saying that perl6-examples needs a reorganization, yes. 21:49
PerlJam pmichaud: what I just described is how I remeber every math lecture I was ever in: "Here's how you do X (but really I'm lying to you and we'll tell you to do X correctly in Calc. III)"
or some such
pmichaud PerlJam: sure, but each math lecture tends to cover one very small topic
(more)
Consider that to get through Calculus I takes on the order of 47 hours
of lecture time
21:50 bluescreen10 left
pmichaud if you just take the first 20% of the course, that's about 9 hours of lecture time 21:50
lichtkind uvtc: where you get the #=> syntax from?
uvtc pmichaud, perhaps what would help you is to create a sort of syllabus --- organize how you might group topics (one per screencast), include subtopics you plan to cover, and post a gist here of the overview. Then you could get feedback aot.
pmichaud uvtc: well, in some sense a syllabus just looks like the table of contents for Learning Perl or Programming Perl, but revised for P6 21:51
uvtc lichtkind, Not sure. I may have made it up? Seemed appropriate.
pmichaud, Right. I suggested it because you said writing wasn't your forte. So, write the ToC, see how everyone here feels about it, then start the screencasts. 21:52
PerlJam pmichaud: stealing ideas from others is always encouraged :) 21:53
pmichaud sure... my response is that it still requires hours of screencasts. for me, those are (precious) hours that I can either use screencasting or that I can use to fix bugs and continue some of the heavy programming lifting for rakudo
PerlJam pmichaud: how long would it take to do a single screencast do you think? (total time)
lichtkind uvtc: just wanted know if its a kind of standard 21:54
pmichaud also, I suspect that there's at least a 3x factor involved for good screencasting -- i.e., a 1-hour screencast requires about 2 addt'l hours of prep
well, prep and/or post-edit
PerlJam A good screencast won't be 1 hour long :)
but I agree with your estimation.
pmichaud see my point earlier about lecturing
21:54 alester left
PerlJam though the factor of 3 probably decreases with time. 21:55
pmichaud there's a reason that college courses use up 40+ hours of lecture/contact time
one can do smaller screencasts -- e.g., one screencast per subchapter of the syllabus... but that doesn't reduce the overall amount of time needed to cover the topics 21:56
PerlJam pmichaud: Sure, just amortize it over a year or two. 21:57
uvtc pmichaud, you may find that, once you've gotten yourself set up, they go pretty quick. A 5-minute one here, 10-minute one there. There could be a wiki page to categorize them, sorta like wiki.perl6.org/Categorized%20External%20Articles . 21:58
PerlJam It feels like you're looking at the mountain and saying "that's too big" rather than looking at the steps in front of you an saying "I can manage that".
sure it takes away from Perl 6 hacking (or other activities), but at (say) 30 minutes a week, it's probably worth it. 21:59
21:59 geekosaur left, eviltwin_b joined
pmichaud PerlJam: given that my p6 hacking over the last eight months has averaged about 30 minutes per week.... that's saying a lot. 21:59
21:59 eviltwin_b is now known as geekosaur
PerlJam pmichaud: okay, perhaps I'm being a bit optimistic :) 21:59
uvtc pmichaud, There's another hidden value in doing them: someone may come after you and transcribe them. :)
PerlJam uvtc: or someone else may take up the torch! 22:00
pmichaud uvtc: I'm not disputing they'd be valuable. I'm disputing my $tuit availability to make it happen.
PerlJam pmichaud: there's the rub. I'd encourage you to do the screencasts because I know you'd be excellent at it, but there's only so many hours in the day. 22:01
uvtc pmichaud, Well, I don't mean to try and convince you. All I can say is that not having user docs implicitly sends a message of "not ready yet". And you're good at exposition.
PerlJam though if someone could plan out a series of screencasts and get a little community feedback, I wonder if we could convince sazbgab to deliver them. 22:03
pmichaud I agree with you that not having user docs is a problem. I'm not sure I'm in a position to directly address that problem as things stand now. If there's an overwhelming sense from the rakudo team that my time would be best spent working on userdocs/screencasts, I'd likely be amenable to that.
PerlJam pmichaud: if that means less hacking from you, I'm totally against it ;)
pmichaud but I'm pretty sure that people would prefer me to do (1) critical regex engine fixes, (2) finish documenting lists / S07 first 22:04
22:04 NamelessTee left
PerlJam yep 22:04
pmichaud PerlJam: it absolutely would mean less core code generation and hacking from me, as things stand now in my personal life.
uvtc IMO, user docs > screencasts. 22:05
Screencasts take longer to get information from. I think they came up because Pm suggested having lectures taped.
(Erm. Recorded. Did I say taped? No, I'm not old enough to remember "tape".)
PerlJam wonders if we can get Salman Khan to do some Perl 6 screencasts ;) 22:06
pmichaud uvtc: anyway, your suggestions are duly noted and very much appreciated. This conversation has been helpful, to me at least. I don't know that we have any immediate answers for what we can do differently in our current situation, but at least there's some agreement on the issues. 22:07
masak PerlJam: that would be awesome.
pmichaud I have to head off now. if anyone reading logs/backscroll has comments, I'll try to remember to read the backscroll (or email/msg me them directly :) 22:08
afk #errands
PerlJam I would be *really* awesome if we had a Perl 6 learning site that was just like khanacademy.org
22:09 uvtc left, bluescreen10 joined
PerlJam oops, home time for me. 22:10
PerlJam *poof*
masak 'night, #perl6 22:11
22:12 wolfman2000 left 22:16 eviltwin_b joined, geekosaur left, eviltwin_b is now known as geekosaur 22:30 mspaulding left 22:31 thou left
dalek blets: f19d179 | (Herbert Breunung)++ | docs/appendix-f-faq.txt:
some questions for the faq
 22:34
blets: 3d5674f | (Herbert Breunung)++ | docs/appendix-f-faq.txt:
format faq
blets: 51625b1 | (Herbert Breunung)++ | docs/appendix- (2 files):
still more format fixes and more to come
22:39 mtk left 22:44 skids joined 22:46 eviltwin_b joined, geekosaur left
rurban There is regression in S05-modifier/counted.rakudo tests 71-80 matched values 1-7 22:46
22:46 poincare101 joined
dalek blets: 436fded | (Herbert Breunung)++ | docs/ (2 files):
all G links are now formatted
 22:51
22:57 adu left 22:59 eviltwin_b left 23:02 snearch left 23:03 geekosaur joined 23:05 whiteknight joined 23:15 geekosaur left 23:16 geekosaur joined 23:19 araujo left, eviltwin_b joined, geekosaur left, eviltwin_b is now known as geekosaur 23:30 eviltwin_b joined, geekosaur left 23:31 eviltwin_b is now known as geekosaur 23:46 geekosaur left, eviltwin_b joined, eviltwin_b is now known as geekosaur

Powered by Cro and the Raku® Programming Language.

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

Thank you!