svn switch --relocate svn.openfoundry.org/pugs svn.pugscode.org/pugs/ | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com | dev.pugscode.org/
Set by putter on 11 February 2007.
Limbic_Region salutations audreyt - still hospitalized? 00:00
audreyt yeah. 00:01
going to get today's test results in a couple hours
to see if I can be discharged. about 50% chance I think 00:02
Limbic_Region keeps you in his prayers 00:03
audreyt thank-you :)
Limbic_Region I also sent you a private /msg too 00:04
[M]erk TimToady, I think that S05 assumes that the reader is already familiar with perl5 regex. This sort of defeats the purpose of having a specification. 00:10
moritz one could argue that pcre are a quasi standard ;) 00:18
but I agree that it's not optimal
audreyt so. not quite ready to be discharged 00:59
probably going to be discharged in two days
I'm getting better but they're not confident enough to let me roam the streets...
audreyt happily goes back to sleep &
gaal p6 Range also has two kinds of endpoints, doesn't it. Inclusive or exclusive... 05:17
tene Yes. 05:18
?eval 1..6
?eval 1..^6
evalbot_r15306 (1, 2, 3, 4, 5, 6) 05:18
(1, 2, 3, 4, 5)
tene ?eval 5.5 ~~ 1..^6 05:19
evalbot_r15306 Bool::False
gaal so it's represented by more than "data Ord a => Range a = MkRange a a" 05:37
audreyt moose 05:38
we don't have Range yet
gaal more like MkRange { r_start :: a, r_end :: a, r_startInc :: Bool, r_endInc :: Bool }
I know :)
moose :)
audreyt also either end can be Inf 05:39
as in *
gaal no matter the type of the other end?
audreyt aye
when both end is *, it always matches
I guess that's a DeRange
tene ?eval 5.5 ~~ 1..*
audreyt that's a ulimit test... 05:40
evalbot_r15306 (no output)
audreyt good. ulimit works
tene Will it always be?
gaal hehe
audreyt nope.
tene Just checking.
audreyt I should blog, maybe
gaal how deranged.
audreyt or I should attack MO
gaal i should go to work :(
audreyt oi :/ 05:41
tene the ulimit settings in evalbot don't protect against "fork while fork", though.
audreyt I should have no work to do for a month :)))
gaal moose!
audreyt tene: fork happily is Unsafe
gaal tene: but fork's unsafe
hee, audreyt's still the fastest moose in the moose
tene ?eval fork
evalbot_r15306 0␤32358
gaal uh-oh
tene evalbot doesn't complain about forking.
audreyt uh oh. 05:42
gaal fixed.
should async also be unsafe?
audreyt no.
async is in-process microthread
gaal yy
committed. 05:43
audreyt gaal++
svnbot6 r15307 | gaal++ | * mark fork as unsafe. tene++ for noticing.
gaal tene++
work&
brad__ hi 06:05
audreyt brad__: greetings! 06:06
brad__ audrey, i have always wanted to meet you 06:07
i hope you are feeling better
not often i get to meet a luminary
i am a perl6 lurker and pugs player-wither
sometimes commenting on your blog as "grumpy" 06:08
audreyt well, this particular one is more flickry than luminary at the moment
oh hi!
brad__ i was wondering what happened to you
then again after your rash of insane output, i was assuming burnout
which is fair 06:09
audreyt I thought so too
but burnout usually recover within a few weeks
brad__ hopefully pugs will be fun again soon
no more politicking
audreyt and this time my nonmotivation and fatigue gradually worsened
so I'm glad it has a much simpler explanation 06:10
and I'm on my way to recovery
brad__ awesome, we need you if only for your kick-ass presentations
but your code is nice too!
audreyt well, it's a pity that the onset happened on the night before my google.tw talk :/
but we'll schedule another talk next month probably :)
brad__ cool 06:11
how far out of sync do you think pugs has become with the state-of-the synopses?
audreyt I wondered if grumpY! works at yahoo (hence the Y!) for quite a while
brad__ yes i am old-time yahoo.
i wrote a bunch of the early parts of the site 06:12
audreyt aha. still @yahoo?
brad__ 11 years+ now
yes still there
audreyt cool!
brad__ larry built my house
audreyt I heard that the early parts were all perl
brad__ yes!
and i still root for perl there
audreyt :D
so, re out of sync. ~3months
brad__ when we were trying to find a replacement for our internal server-side lang, i chaired to mod_perl side
i lost to php, its been downhill ever since 06:13
audreyt well I was told that parts of Yahoo is still in mod_perl...
but php is "canonical"
brad__ yes and yes, they still provide some support to mod_perl, but they'll likely kill that over time
too bad too
but its water under the bridge
they intentionally wanted an idiot language 06:14
and they got one in php
audreyt re politicking, yeah it took most of the fun away from me when I was in .br 06:14
renewed my resolve to ignore all politics this time around
brad__ and the reddit postings must not help
i am sure you have seen
tene explain "intentionally wanted an idiot language"?
audreyt no, I had not 06:15
brad__ programming.reddit.com/info/157m0/comments
lambdabot Title: Perl 6 and Parrot: Things I Probably Shouldn't Say But No One Else Seems To ...
audreyt nor do I actually want to
brad__ tene - a motivation to get a tool that had some power but not as complex as perl
which i will rewrite as wanting an idiot language
my own editorial there
at least with php4 i can say you can feel your iq drop as you use it 06:16
brad__ in any case audrey, i credit you with leading me to haskell 06:17
if perl6 never pans out i will get something out of this
i am in no danger of being in your realm of expertise of course
room has gone silent...was it my cologne? 06:18
tene The room is often silent.
audreyt well, hopefully not anymore :)
brad__: do you prefer the nick "brad" or "bjc"?
(when committing to pugs)
brad__ audrey - let me commit to this. i will read the pugs code and try to understand it, then i will humbly take a commit bit for something harmless 06:19
it is important for you to understand how far you are ahead of the rest of us 06:20
audreyt an invitation is on its way to your inbox regardless... :)
brad__ which one?! you can send to [email@hidden.address]
oh boy, i put an email address in a irc log
game over for me
audreyt done 06:21
brad__ i will promise you that i will crack open the pugs code again
audreyt woot!
brad__ my kids grant me three hours a night for me stuff
audreyt are you not also bjc @ yahoo?
brad__ i gave up that email because of too much spam, i also gave up [email@hidden.address] years agao too (i had it before yahoo mail existed, when yahoo used @yahoo.com) 06:22
audreyt (I sent an invitation there, too, with nick "bjc". the b7j0c one I sent an invitation too with nick "brad".)
gotcha.
brad__ thanks audrey! i will not waste the opportunity
audreyt wow, such politeness :) 06:23
brad__ well i legitimately think you are a genius. so i am giving deference
audreyt re the reddit post. after 3mo of rest I can honestly say I have no hard feeling wrt chromatic anymore... he sometimes unintentionally exaggerates things for effect (such as the parrot's release frequency), but overall he is cool. 06:24
brad__ i sensed that, i think some people just like a little drama 06:25
audreyt aye. so it's all fine now
please test that svn works by adding yourself to AUTHORS and commit -- our new invite.pugscode.org system havn't seen much use, unlike the old one 06:26
and welcome aboard :)
brad__ thanks!
hey, i thought you would be a shoe-in to use darcs
audreyt there is also darcs.pugscode.org. 06:27
brad__ i tried it out of haskell-snobbery and am sold on it
audreyt read-only at the moment, though
but people do use it all the time to track our commits
brad__ i am interested in what os you use, out of curiosity
audreyt (as seen in http logs)
freebsd for a long, long time.
finally decided I can't hack all layers of computing
so I'm on darwin/osx now
brad__ hmm, i am on freebsd6.2 now 06:28
audreyt which is like freebsd except it doesn't prompt me to learn obscure computing knowledge that often
offby1 yeah, just getting my wireless card working on FreeBSD 6.2 was a challenge
brad__ really? i have an ANCIENT laptop and wpa_supplicant did the trick!
with a netgear card 06:29
audreyt and I had to hack snd_ich to unbreak the sound card that's broken during a _point release_
etc, etc. :)
brad__ hmmm, yeah multimedia has never been a bsd strong suit
but then again my UI of choice is gnu screen
so my needs are modest
audreyt ditto.
offby1 brad__: I have no idea what wpa_supplicant is.
brad__ screen over dwm! 06:30
audreyt oh and darwin runs OmniGraffle :)
offby1 I naïvely expected it to just work when I plugged it in. Silly me
brad__ offby1 - in debian and freebsd you have to do some twiddling for wireless. ubuntu is what you want if you want wireless out of the box
there are good reasons for this
ubuntu will use binary blobs
debian won't 06:31
freebsd will
but they don't care that much about laptops (yet)
tene And yet they will all run Perl6!
brad__ that they will
audreyt wonders about writing device drivers with perl6
brad__ in any case, my UI is designed to be a console done with xft fonts. that is all i want 06:32
i REALLY use elinks as my browser
audreyt elinks > w3m because... ?
(w3m is my browser of choice.)
offby1 brad__: I had been using ubuntu for just that reason. Foolishly upgraded to feisty, and ... wireless stopped working :-(
brad__ elinks is just a weee bit easier to customize
and also, it had lua scripting
audreyt ah. Lua.
very important that. 06:33
brad__ but that was before i discovered PRIVOXY
now i do not need lua scripting
PRIVOXY FTW
pcre-enabled net proxy!
brad__ also elinks has numbered links 06:34
numbered links are important in a text browser
crucial imo
in any case i still beat the drum of having parrot be the vm of firefox 06:35
because i hate js
but it will never happen 06:36
tamarin FTW :(
brad__ okay audrey - created a passcode at pugscode. what now? 06:38
tene brad__: add yourself to AUTHORS and commit
audreyt re tamarin FTW. yes definitely.
also JS2/ECMA4 is really, really not bad. 06:39
I can't bring myself to hate it at all :)
brad__ i know the next js is not bad, but i hate being denied choice 06:40
why must there be only one client language?
audreyt well as long as we target tarmarin
it's just another vm
brad__ yes but is js a viable compile target?
audreyt oh definitely.
brad__ i wonder
we'll see
audreyt it may be that tamarin bytecode is easiest to target 06:41
but in any case we need to get 6.28.0 out of the door... an object model bridge can only exist when _both_ sides exist
brad__ hopefully, actually the world does not give a poop what i think, the web platform is settled
audreyt which reminds me. I should get hacking :)
(bbiab) 06:42
gaal audreyt: how much newVal changes should go into 6.28.0?
audreyt gaal: all of it
brad__ okay with that i will let you sign off, i got my fandom in for the night, i met audrey. audrey i will check out the pugs tree and get reading
gaal need to revisit much of Parser.*
audreyt gaal: though I'm still not sure about the VPure/VMut/VExt variance typing
I intend to get
gaal oh?
audreyt vv('some string').reverse
working using MO today
yeah. it's better to carry the P/STM/IO type annotation directly within the Val type 06:43
then we only have the Val type.
(for now.)
since we're optimizing for consistency not for speed
brad__ bye everyone, thanks for putting up with me
audreyt the intrinsics and undefs can just be normal Val for now
brad__: see you soon!
brad__ good night! 06:44
gaal bye
I don't see quite how it goes
audreyt I'll code instead :)
easier that way :)
gaal haskell+++
hehe, I didn't even mean to put an extra + there
"haskell" ++ repeat '+' or something 06:45
tene Or perhaps you were endorsing something called haskell+ 06:46
gaal +++. Is that preendorsing or postendorsing? 06:48
++++ 06:49
svnbot6 r15308 | gaal++ | * improve decode-utf-chunks.t to make the TAP telling. 08:33
gaal could anyone run a 15308 pugs against t/xx-uncategorized/decode-utf-chunks.t and see if they get those weird backslashes too? 08:35
"ok 2 - chunk \#0"
audreyt it's not weird 08:36
TAP protocol mandates it
otherwise it's comment like
gaal ahhhh
audreyt #SKIP
gaal right, right
audreyt so just drop the # from the tests
gaal haha, i was thinking it's a pugsbug
yeah
fixed.
audreyt k
svnbot6 r15309 | gaal++ | * declutter TAP output by dropping the "#" 08:37
gaal this runs very slow on a 'make fast' pugs 08:38
making a non-fast pugs...
gaal considers renaming 'make fast' to 'make soon' 08:39
audreyt gaal: do it :) 08:49
gaal done. we should also have "cheap" and "good" targets maybe. 08:53
svnbot6 r15310 | gaal++ | * Makefile.PL ~~ s/fast/soon/, as this target builds a slow pugs. 08:54
Juerd make tuits
gaal make love 08:55
audreyt not war? 08:56
"make test, not war"
so vv('foo') is close to working
jisom don't you want a battle hardened pugs?
audreyt but then I discovered my attention span, though much improved, is still not normal 08:57
audreyt faints into unconsciousness for a bit &
gaal make sweet-dreams
actually "make fast" builds a shipworthy pugs 08:58
that doesn't shift as the ship heels etc.
I don't have any good comparison points, but isn't >20sec slow for t/xx-uncategorized/decode-utf 09:14
-chunks.t ?
wolverian Juerd, I see feather has bitlbee installed. how does one use it? :) 10:56
Juerd, it's not in inetd.conf afaics..
Juerd Preferrably not at all, if I were you. It's a huge privacy hazard. 10:57
But see man bitlbee
wolverian oh. it is?
Juerd Run it with -D
wolverian is it because of bitlbee itself, or the method it is used on feather? I was just setting up it on my own server, too.
Juerd Sure - people on instant messaging will assume you're in for intimate personal talk, while feather is hardly secure.
wolverian ah. true enough. 10:58
hrm, well. irssi does something it doesn't like and the connection fails, anyway. 11:02
and on my server, from inetd, it just refuses the connection. sigh :) 11:03
I love daemons without logging..
svnbot6 r15311 | fglock++ | kp6 - added Code.pm 12:00
avar What are some of the posix regexp implementations? glibc, ..? 12:52
avar ot:)
-/w 5 12:54
eek
moritz POSIX defines regexes?
avar yep 12:55
moritz in this channel I learn more than in 13 years school ;)
moritz TimToady: will there be a "Programming Perl 6" book? 13:20
pdcawley would be surprised if there weren't. 13:23
xdg moritz: I think TimToady needs to finish the language spec first. :-) 14:22
Though, really, I assume a compilation of all the synopses, etc. will be the base of any such book.
Khisanth pokes Someone at 85.197.228.236 17:28
svnbot6 r15312 | gaal++ | * very broken UTF-8 caused a crash. added a test and fixed the crash. 17:44
r15312 | gaal++ | (Note: please do not add the usual svn props to pathological-8bit.dat)
cj moritz: the prototype for the Programming Perl 6 book is the Perl6 Apocalypses: dev.perl.org/perl6/doc/apocalypse.html 18:35
lambdabot Title: Perl6 Apocalypses - perl6
cj right, that.
ingy seen audreyt 18:43
jabbot ingy: audreyt was seen 9 hours 46 minutes 41 seconds ago
qmole 3 20:00
ingy qmole: I think you want #python ;) 20:55
(if you insist on discussing the Holy Hand Grenade) 20:56
ingy hi buetow 20:58
buetow hi
putter TimToady: is there a spec way to select p6 dialect within a p5 re? perhaps (?p6) (?p6:...) ? Don't ever need (?-p6) ;) 23:29
(?5) (?5:...) 23:30
err, no. (?6) (?6:...)
moritz putter: uhm, from p5 I don't know such a construct, so why should it exist in p5-regexes in p6? 23:31
putter: I thought they were mainly for backwards compatibility and convenience...
putter: and if you need more advanced features you can fall back to p6 rules 23:32
putter symmetry? it's a backward compatible mod of p5 re syntax.
if you have flavor x and y, and you can embed x in y, it's just asking for headaches to be unable to embed y in x. 23:33
re "mod of p5", no pun intended. really. 23:37
moritz ;) 23:37
moritz well, I'd say that both points are valid and that it's just a matter of taste/convenience which one to choose 23:38
(of course this beeing perl I know what the answer will be ;)) 23:39
putter lol
moritz putter: and beeing (becoming...) a physicist I appreciate symmetries as well ;) 23:43
putter Basically why I'm here. Perl's eschewing of the otherwise pervasive "no, you may NOT have that power, because it {can't be compiled efficiently, can't be type checked, is ugly, is likely to be abused, confuses newbies, etc, etc, etc}". 23:45
Hmm. General eschewing. I still think not being able to define methods on ARRAY,HASH,etc, was... well, closed off one rich direction for language evolution. 23:48
tene putter: explain more? 23:48
moritz putter: can you at least inherit from {ARRAY,HASH,etc}? 23:49
putter there were p5 patches to permit {package HASH;sub foo{} } {}->foo; it was (perhaps correctly) thought this would cost more in bugs and confusion than it was worth. 23:50
wolverian I think it's a pretty universal sentiment that tie() was not the right solution either :) 23:52
moritz but still a clean interface allowing just that would be cool ;)
putter basically it might have permitted rubyesque "everything is an object" style.
putter moritz: re "inherit from ARRAY, HASH", I don't believe so. I believe you have to use tiearray,etc, and annoying tie games. 23:54
putter :) 23:55
moritz well, time for bed ;) 23:59
TimToady_ putter: I certainly wouldn't stop anyone from adding a (?p6:...) to p5... :)

Powered by Cro and the Raku® Programming Language.

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

Thank you!