pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
dduncan My impression is that, while the Haskell backend could be viable longer term, it may ultimately have just been a rapidly developed prototype for Perl 6, letting us work out some design kinks through practice, and that the Parrot one will shortly take over as the main implementation. 00:15
So optimized for fun vs optimized for longer term use. 00:16
At the moment I'm not putting much stock in the Haskell version progressing much longer, and that the Parrot one will be the most viable implementation within the next few weeks/months. 00:17
markstos dduncan: thanks for the viewpoint. 00:19
dduncan I think for a foundation Parrot is also a lot easier to install than Pugs is, the former having perhaps much simpler or common dependencies ... C vs Haskell.
markstos "apt-get install ghc" is rather easy.
I have poked through the parrot bytecode files supporting that implementation of Perl 6, and so far I have't find them accessible to hack on. So far, Haskell looks more interesting to me to learn. However, perhaps with NQP, more Perl 6 development will be happening in a language that looks Perl6ish 00:21
dduncan well sure, Haskell is a much more advanced foundation, and arguably a better language ... but still a larger dependency 00:22
as for *today*, though, I think Pugs is still further ahead if you simply want to write Perl 6 00:23
but I think that by the time I migrate from Tiger to Leopard, I may not bother installing ghc on the latter, we'll se
markstos Yes, although the slow performance of pugs was frustrating. The parrot backend seems to be faster now, but has less features than pugs does. 00:24
dduncan that's a big reason itself ... currently Perl 6 under Pugs performs about 200X slower than equivalent Perl 5 00:25
its really hard to do Perl 6 development with that
I'm currently assuming that Parrot will add enough features soon that I can develop under that instead ... if it doesn't already have enough 00:26
markstos And hard for me to conceive what it would take to optimize it to improve performance to be reasonable close.
dduncan I'm hoping that Parrot has a better chance of reaching native Perl 5 speeds ... or at least within 10X in the short term
Khisanth dduncan: it's kinda dead in the water if it doesn't ... 00:58
dduncan indeed 01:00
Khisanth on other hand it has been said perl5 has been getting slower with each release so if they wait long enough ... ;) 01:01
jjore-m But 5.10 is supposedly faster than 5.8. 01:02
jjore-m except for my ( $self ) = @_ which is much slower (but only due to a bug and not through anything intrinsic) 01:04
use.perl.org/~jk2addict/journal/35189 fwiw.
lambdabot Title: Journal of jk2addict (4946)
dduncan apparently it has a bug making list assignment slower though
jjore-m Jah, that's what I linked to. 01:05
I looked at the code for deciding the COMMON flag but looked away when it turned out to be ugly.
meppuru good night 03:00
Limbic_Region @tell TimToady Happy New Year - in S02 "with multiple identical bracket characters, which must by closed by the" s/by/be/ 05:43
lambdabot Consider it noted.
Limbic_Region @tell TimToady in S02 "If any native type is explicitly initialized to C<*> (the C<Whatever> type), it is left uninitialized." This is confusing since a native type can't be undefined. It might be better to say is initialized to the default values as shown above. That is of course if that's what happens - if not, I don't understand. 06:25
lambdabot Consider it noted.
duzy greeting, :) 08:26
Mips_ hi 09:35
pmurias what would be a good way to map perl6 calling convention to perl5? 13:29
pugs_svn r19298 | pmurias++ | [kp6-v6] 13:33
r19298 | pmurias++ | cleaned up KindaPerl6::Runtime::Perl5V6::Runtime a bit
r19298 | pmurias++ | a simple calling convention benchmark
clkao win 10
krinke #ruby-de 15:15
pmurias krinke: ? 15:24
krinke utlo?
Limbic_Region @messages 15:27
pmurias krinke: utlo means?
krinke yes
Limbic_Region @tell TimToady ref S02 I meant to ask you about KeySet as well but I fell asleep. It says increment = true and decrement = false. If you decrement an already false value does it stay false? If so, should the doc explicitly state this? I am thinking of the silly $|++ here 15:37
Limbic_Region TimToady ping 16:30
Juerd_ I just killed an elvis and a pugs on feather 16:33
The elvis was owned by root, the pugs was owned by andara
fyi
I'm incredibly tired so I'm not going to investigate further :)
Limbic_Region sleep well, and happy new year
Juerd_ Thanks and ditto 16:34
Juerd_ last woke up yesterday around noonish. It's now 17:32
afk # zleeb
pmurias if i pass a number to substr how does it know if i want end or length? 19:16
Gothmog_ could you be somewhat more specific? i.e., which substr, and which argument are you referring to? 19:19
pmurias Gothmog_: Functions.pod: 1355 19:20
docs/Perl6/Spec/Functions.pod 19:22
Gothmog_ by dispatching, you may either call it with a StrLen or StrPos type argument. 19:23
our Str multi substr (Str $s, StrPos $start, StrLen $length?) is rw
our Str multi substr (Str $s, StrPos $start, StrPos $end?) is rw
pmurias but a number is converted in either Len or Pos 19:26
1363 19:27
pmurias i'm implementing it as 01-sanity/07-substr.t 19:29
pmurias the first form is used 19:31
there
Gothmog_ hm, that seems ambiguous then. but shouldn't you do the cast in substr but before the call? 19:45
Gothmog_ either way, there will be default if you don't explicitly cast your number. 19:47
pmurias Gothmog_: it should be mentioned 19:51
Gothmog_ ack
pmurias TimToady: what does substr(Str,Int,Int) do? 19:52
pugs_svn r19299 | pmurias++ | [kp6-v6] subs withought arguments, a bit more runtime 20:13

Powered by Cro and the Raku® Programming Language.

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

Thank you!