Run Perl6 Now! | pugscode.org ('Overview', 'Journal') | pugs.kwiki.org | channel log: xrl.us/e98m
Set by autrijus on 6 March 2005.
metaperl_ I went there when I lived in Taiwan 00:00
mugwump purl, #taipei.pm? 00:02
doh :)
obra #elixus on irc.tw.freebsd.org 00:03
mugwump how long did you live in Taiwan for, metaperl_ ? Are you a good person to ask for I18N tips from? :-) 00:40
metaperl_ oh no, I know nothing about that? 00:42
Limbic_Region I lived in S. Korea for a year and a half and the only other country I visited was the Philippines - shame
metaperl_ s/\?/\./;
what did you do in S. Korea? 00:43
It costs money to travel and nothing is close by land to Korea if I'm not mistaken
arcady russia is close enough to korea to have ferry service 00:49
metaperl_ are you serious? 00:53
Limbic_Region sorry - was paying attention to #perl
metaperl_ you arent the arcady from LA.pm are you?
Limbic_Region I was a contractor to the US Army
my gf at the time got stationed there
so I told my boss to get me a job there or I would quit 00:54
metaperl_ oh I see. I enjoyed Taiwan and miss it. My plane stopped in Korea on the way back to the US but that's it
Limbic_Region 1 month before we were supposed to come back to the states - she sent me a dear john email
Limbic_Region was well compensated
besides - a large portion of foreigned earned income is tax free (under the right conditions)
oh - if we are talking layovers or 1 day trips, HongKong and Tapei should be added to the list 00:55
3 countries in Europe, 3 in North American, and 3 in Asia 00:56
metaperl_ I knew an American who teaches in Japan. He simply quit responding to his creditors, including student loans. He said they assume you are dead after 3 years 01:00
and just nullify the loans
metaperl_ quit too --- time to leave this office and head home!
adios fellow hackers 01:01
ingy hi metaperl_ 01:12
arcady wow, I didn't even know there's an LA.pm... or another arcady for that matter 01:14
stevan added a basic early support for isa_ok() in Test.pm if anyone is interested. r542 01:19
cls_bsd ōæ½xB5Lōæ½xAAkōæ½xB3sōæ½xB1ōæ½xB5ōæ½xA8ōæ½xECōæ½xA6ōæ½xF8ōæ½xAAAōæ½xBEōæ½xB9 (svn.openfoundry.org) # cannot connect..... 02:04
rooneg i was getting that before, seems to work now though 02:05
seano :t ws 02:40
by which I mean "hi"... 02:41
hcchien 95/1360 subtests failed 02:54
arcady and the test results have some weird numbers like 178% of subtests failed 03:17
obra yeah. there's been lots of test un_todoing 03:19
04:01 metaperl__ is now known as metaperl_
autrijus greetings on another frabjous day! 04:06
obra hi! 04:07
seano calloo callay, homes. 04:08
autrijus :)
seano Has either of you been using GHC 6.4? I just realized that both FiniteMap and Data.Map look buggy... 04:09
autrijus I've been testing against it
but for daily work I'm still using 6.2
mostly because 6.4 is still in early rc
seano hm... I guess everyone has his own definition of "rc". 04:10
Parsec is sure a fun tool, but it seems to take a different kind of discipline to make grammars legible. 04:20
autrijus explain? :)
seano It almost makes me wish for a BNF -> Parsec tool, like yayacc or something. 04:21
autrijus ahh. I see.
seano With BNF, your description and your language are very similar.
autrijus seano: write a Rule->Parsec compiler? :)
seano With Parsec, it's very tempting to add helper functions, folds, maps, and tables to generate the thing, which makes it hard to read. 04:22
Yeah, something like that...
For a real mind-bender, write a BNF parser in Parsec that generates a Parsec parser.
I dare you ;) 04:23
autrijus that is how Rule->Parsec compiler must be written :)
well, there is already Happy
flw autrijus: ōæ½xCEōæ½xD2ōæ½xB7ōæ½xA2ōæ½xC1ōæ½xCBŅ»ōæ½xB7ōæ½xE2ōæ½xD3Ź¼ōæ½xFEōæ½xB8ōæ½xF8ōæ½xC4ōæ½xE3ōæ½xA3ōæ½xACōæ½xD4ōæ½xDA [email@hidden.address] 04:24
seano True, but I think the staticness makes it unusable for "is parsed foo".
flw: wtf with the row of boxes. Is that perl6?
autrijus and there is already www.cs.chalmers.se/~markus/BNFC/
flw: please try to keep #perl6 as utf8... if you'd like to write gbk we can talk at #perlchina :) 04:25
flw: and yes, I did get your mail.
flw: can you provide a telnet server or ssh? 04:26
so that I can get on your machine and debug?
seano autrijus: cool, didn't know about that one.
autrijus otherwise, I'm trying to lure autarch into implementing Build.PL for us
in which case your problem is more easily solved.
flw seano: ??? 04:28
seano non-ASCII in Emacs looks like boxes.
mugwump Heh, confusing Perl 6 with Chinese. "It's all to me!" 04:36
į¼™@ā@ā@į¼”@vį¼”@KĪŗ 04:42
lol
mugwump tries iso-8859-3 (B&%B(B&KB(B&KB(B&GB(B&MB(B&IB(B&JB 04:43
mugwump gives up
autrijus heh :) 04:50
mugwump did you get my modified sendmoremoney.p6 script, autrijus ? 04:51
autrijus mugwump: I did. I also wonder why you did not commit it yourself :)
mugwump My message was asking if I could have access, in the form of a token change. I grokked what you were trying to do with the sendmoremoney.p6, and as I'd recently explored the problem space, I'd show the script 04:52
s/show the script/add my findings/, sorry 04:53
autrijus oh. sendmoremoney.p6 is not my baby :)
mugwump oh, duh.
autrijus so I was briefly confused :)
viirya autrijus: why re-tranalste? You made changes in PA01?
autrijus I think it's nnunley's
viirya: yes, there is one single paragraph rewritten by Stevan
viirya: see yesterday's journal
mugwump: an invitatil should be @ your mailbox now. 04:54
viirya autrijus: ok. I see it. 04:55
autrijus thanks!
mugwump waits 300 seconds for the greylisting timeout ...
autrijus mugwump: welcome aboard. sorry, I thought I've invited you earlier; mea culpa 04:56
viirya autrijus: it's done, how do I give you it? 05:18
Darren_Duncan greetings humans 05:24
I've just checked in an updated Locale::KeyedText ... 05:27
autrijus Darren_Duncan++
Darren_Duncan some bugs are fixed, partly thanks to some questions I got answered on p6l
autrijus yeah, larry seems excited
it's good to excite larry :)
<- but I gotta run.
Darren_Duncan but one of the more interesting things is that the new version uses new Perl 6 features 05:28
subtypes I think they're called
lets me cut out a lot of input checking code
autrijus also "where" clause I think is useful. 05:29
Darren_Duncan assuming I used them correctly
my subtypes use the "where" clause
autrijus ah. cool
Darren_Duncan I just didn't want to repeat it for every variable
autrijus nod nod.
Darren_Duncan I consider subtypes like SQL domains
you can pre-declare a set of constraints and use it everywhere 05:30
autrijus yup. logic programming, here we come
I just submitted YAPC::NA proposal :)
obra !
oh. canada.
Darren_Duncan of course, that's one more thing Pugs'll have to support before my mod runs
autrijus obra: yes, the very reason :)
Darren_Duncan other end of the country from where I am
I'm going to try to be at OSCON this year 05:31
my first ever big Perl conference
it's location is relatively close to home
and some of the biggest stuff happens there
autrijus have fun. :) ingy will speak there for me on pugs
Darren_Duncan but I suppose its more expensive
what are you talking about at YAPC? 05:32
same thing?
autrijus yeah. "Apocalypse Now"
but more of a how-to-hack-pugs tutorial
Darren_Duncan eventually I'll try to visit all the english-language Perl conferences, as I have the means to do so
autrijus :) 05:33
Darren_Duncan it seems like some people go to all of them
obra is trying to cut down on conferences
autrijus obra: I figure 3 days can't hurt :)
Darren_Duncan I could do that later, but gotta go up before I can come down
autrijus runs &
obra + how may days of travel ;)
Darren_Duncan I figure its a great opportunity to meet all these Perl folks I only see on the internet before that 05:34
obra Indeed.
Darren_Duncan that said, I have met a couple names in person
obra has spent a bit too much time at the worthless conferences.
Once I hire minions, things will get better, though
At least in theory 05:35
Darren_Duncan If any of you have heard of Peter Scott, who wrote Perl Debugged (name?) and spoke at OSCON in 2004 about taming legacy perl, he lives in my home town and we meet at the monthly Victoria.pm gatherings
obra Ah
Darren_Duncan Also, about 1.5 years ago, a Linux cruise stopped by in Victoria and I briefly met ... Mike Schwern (name?) ... and I met Linux Torvalds 05:36
obra But yes, come to conferences. Meet more people. Share more ideas
Darren_Duncan not mike schwern ...
Randal Schwartz I met 05:37
that's about it
wolverian well, Larry Wall has replied to my mailing list post! so nyah! 05:38
wolverian cries softly
Darren_Duncan he replied to mine too
and he said thanks for my efforts
so nyah! 05:39
anyway, Larry is one of those I'm most looking forward to meeting when I get the chance, for several reasons 05:40
mugwump #1. to bow down and chant "we're not worthy"
Darren_Duncan won't list em
mugwump #2. Hit him up about why there can be no $%@#$% operator in Perl 6
Darren_Duncan who says there can't
I seem to recall Perl 6 will let us redefine its own grammar 05:41
I think that was Larry's idea
obra Oh wow. The ruby track at OSCon is going to be great.
seano because that's a magic variable that stores the last failed grammar rule.
mugwump Is that a flying saucer? No, wait, I think it's a pie
Darren_Duncan is there another bet?
statico mmm... unidentified pie...
Darren_Duncan like pie-thon the return? 05:42
or the son of pie-thon
or the bride of pie-thon
mugwump pie-in-the-sky-thon
nothingmuch good morning 05:52
hoowa morning 06:14
thesaboteur morning
hoowa afternoon in china. 06:15
thesaboteur 1am here
hoowa where u local?
thesaboteur east coast usa 06:16
hoowa woo :0
morning morning!!
thesaboteur yea haha
hoowa alexe?? 06:17
06:33 Paler2872 is now known as cnhackTNT
Darren_Duncan about ... stuff? 06:34
thesaboteur mmmm...... invisible stuff. 06:39
hoowa stuff? 06:44
Darren_Duncan that was a joke ... 06:50
it is what someone can say to try to start a conversation while being fully non-descriptive
thesaboteur hmm is there a way to get more verbose error output from perl? im running it with -w but is there anything else I can do to get more info from the errors? 06:53
#!/usr/bin/perl -w
Use of uninitialized value in split at /Library/Perl/5.8.1/Net/AOLIM.pm line 2406. 06:54
that's the error I get
&{$imsg->{'callback'}}($tp_type, split(/:/,$tp_tmp,$SERVER_MSG_ARGS{$tp_type})); < line 2406
Darren_Duncan the only way I've found to resolve that is to split up the offending line into several shorter lines 06:55
not with whitespace, but with the use of a few temp variables 06:56
thesaboteur yeah
obra it might make sense to print out the variables just before that line
or use the debugger, if you dare. but really, #perl is a better place for perl5 questions
Darren_Duncan sometimes printing them helps, sometimes not
thesaboteur ah ok that I can do
Darren_Duncan good idea 06:57
thesaboteur kool thanks
Darren_Duncan I've just committed some more LKT upgrades
that's it for tonight ... tomorrow I'll upload a test suite for it
it looks a lot more Perl-6ish now
obra Night, Darren.
Darren_Duncan same to you
obra Oh. it might be worth taking it and annotating it and posting it on your use.perl journal or somesuch 06:58
are you blogging about perl6 hacking? if so, I'll add you to planetsix.perlfoundationlorg
Darren_Duncan I should post what there?
no, I don't have a blog for any reason
usually, stuff like this that I want to share, I email a related mailing list 06:59
obra so. you're one of the first folks actually _coding_ in perl6.
Darren_Duncan probably
obra It might be interesting for folks to read about the stuff you're doing.
Darren_Duncan maybe, but blogging never really interested me, and it takes time
obra use.perl.org provides free blogging for perl-related people. and we've got public blog-aggretators so the general public can see what's up
ingy hola
obra hey ingy
ingy hi jesse
obra Understood. 07:00
obra is getting very excited about the hualien hackathon
Darren_Duncan still, I've got a little problem where it takes me a lot more time to do things than the average person, so doing extra things like a regular blog can be significant 07:01
on the other hand, I may be pleasantly surprised
ingy Darren_Duncan: maybe you irc too much :P
Darren_Duncan part of it is just about figuring out what to say without sounding boring
obra *nod* you could also just write to mailing lists and copy and paste your messages into the blog ;)
so. I'm a module author. I'd love to hear about what makes perl6 so much better for coding in. 07:02
since I've been paying only a bit of attention and it sounds like you've found it nicer than perl5 in some ways
Darren_Duncan it takes getting used to, but I certainly appreciate a lot of Perl6 features
the subtypes can save a lot of input checking code for example 07:03
assuming they do what I think they do
I mean, subtypes in combination with a 'where' clause
sort of like a SQL domain
and I'm quite into databases, if you hadn't noticed from my modules 07:04
anyway, while I *could* start a blog, I can't promise that I'll write in it very often
ingy obra: hualien looks awesome 07:05
obra nod 07:06
to both of you
ingy it will be a fun drive down the coast 07:10
Darren_Duncan as for public-visible stuff, fortunately all the Perl mls and this IRC channel are all archived on the web, so everything I say is sort of published
ingy I can't wait
obra driving? last time, we just took the train 07:12
Darren_Duncan question: yesterday there was a p6l post by Shu-chun that said the ternary operator wasn't working in pugs, ?? :: ... is this known to be an outstanding issue or known fixed 07:13
actually nevermind, I'm leaving 07:14
nothingmuch morning 07:37
08:14 sorje_ is now known as sorje
nothingmuch hola larsen 08:18
Khisanth ack! 08:19
Khisanth gives up on haskell and tries perl6 instead :)
larsen nothingmuch: Hello. 08:21
ingy Khisanth: what haskell stuff have you read? 08:25
Khisanth ingy: just following stuff on haskell.org 08:27
ingy Khisanth: I like www.haskell.org/onlinelibrary/ 08:28
just the facts
Khisanth "A Gentle Introduction to Haskell" seems nicer than YAHT
ingy agreed
but both take to long to get to the nuts and bolts
nothingmuch what is long? 08:29
the first 40 or so pages of YAHT, which is 200 pages
were not that dense
Khisanth ingy: maybe we're just impatient :)
nothingmuch so they went quickly
i spent about 90 minutes learning haskell so far
and excercises i felt i should skipped i did 08:30
but then again, i sort of skimmed
nothingmuch feels cunning 08:55
make test now works on OSX 09:02
whomever++
(ingy?) 09:03
ingy did it not work?
nothingmuch it didn't work yesterday 09:05
PERL6LIB problems
make purge didn't help
but now it does, maybe because i upgrade ExtUtils::MakeMaker
maybe not 09:06
eitherway, karma is up
dada hola 09:19
nothingmuch hola dada 09:25
has anybody but autrijus taken a look at the fatal (even in eval) casting of hash to string?
i'd like to comit the second iteration of t/op/smartmatch.t
which covers everything but rules & objects
and some stuff i didn'
t understand
theorbtwo Morning. 09:26
nothingmuch hola theorbtwo
autrijus hi. 09:40
nothingmuch morning autrijus
autrijus where is the stringification to hash specced again?
I have fixed not()
and various related parsing bugs
and hopefully made flw's nmake problem go away.
nothingmuch i would like it to just not die horribly in an eval, 09:41
appearantly in S02 at the top
as arcady noted
it says $key\t$value\n
autrijus "There is no such thing as a hash list context." 09:42
oops. so pugs is wrong.
autrijus makes a note
nothingmuch a hash in list context == a list of paris, no?
don't make a note, i'll write a test 09:43
oh
not hash in list context 09:44
list context into a hash
autrijus hash stringification implemented as r554. 09:45
autrijus goes out to find food 09:46
food &
nothingmuch *hug*
thank you!
autrijus :)
it's AST.hs line 158 if people are interested. 09:47
it's one line:
vCast (VHash (MkHash h)) = unlines $ map (\(k, v) -> (vCast k ++ "\t" ++ vCast v)) $ fmToList h
nothingmuch \n?
autrijus unlines = \n
nothingmuch ah
autrijus gone & :)
theorbtwo Anybody SubEthaEditites around here? 09:59
nothingmuch is here 10:00
but i can't punch work nat
you'll have to arrange for that
theorbtwo I actually wanted an ethereal or tcpdump trace of an anonymous subethaedit session. 10:02
nothingmuch anonymous? 10:03
i.e. rendezvouz?
theorbtwo Anon IE without encryption.
nothingmuch i doubt there is encryption either way 10:04
subetha uses zeroconf to find other subethas on the LAN
subnet
and aside from that it's just a propietry protocol
theorbtwo I know, but it uses BEEP for the actual communications, and BEEP will do SASL.
nothingmuch tcp/ip
ah
theorbtwo SEE 2.0 or 1.0?
nothingmuch that i didn't know
nothingmuch is outdated, appearantly 10:05
theorbtwo <g>
nothingmuch well, not really
software wise i'm up to date
i just was not aware of any changes
you want to glue see operability into something else?
theorbtwo xemacs.
Not that I've got the elsip chops, but perhaps I can figure out the protocol, and get somebody else to do it. 10:06
nothingmuch search.cpan.org/~autrijus/Kwiki-Edi...ubEtha.pm?
theorbtwo Looks like that just scripts a subethaedit instance. 10:07
nothingmuch oh, right 10:08
hmm
funny way of doing it
if i connect to you, you can just tcpdump and see if it does sasl
theorbtwo Um, I could, if I had a subethaedit, but the point is that I don't. 10:09
Connect to see://www.codingmonkeys.de/ ?
Juerd_ hates his wrists and hands 10:16
theorbtwo Type with your feet?
Juerd_ I wish I could
theorbtwo Should learn. 10:17
I've never tried to type with my feet, but typing with my nose works OK. 10:18
Of course, it makes it hard to watch the screen while typing.
nothingmuch theorbtwo: oh, i thought you were running pearpc
so you want me to connect there, and give you tcpdump | grep codingmonkeys.de? 10:19
theorbtwo No, have to set it up, I suppose.
That won't give me payloads. Hold on a sec, let me figure out how to convince tcpdump to give me what I want.
nothingmuch tcpdump -vv? 10:20
theorbtwo tcpdump -w host codingmonkeys.de 10:21
luqui just had an insanity attack
theorbtwo Er, tcpdump -w host codingmonkeys.de > see.tcpdump
Hmm?
luqui hmming at me? 10:22
theorbtwo Aye.
nothingmuch tcpdump won't accept 'host codingmonkeys.de'
luqui my most recent message on p6l is purely insane
nothingmuch one minute
luqui but perl6 has been open to insane ideas before :-)
theorbtwo try "host 212.204.60.78" 10:23
Oh, that should probably be quoted.
nothingmuch ah, yes
i thought maybe -w wants a file arg
theorbtwo D'oh, yes, it does. 10:24
I haven't use tcpdump directly for quite a while -- I just use ethereal, and let it do the hard bits.
nothingmuch theorbtwo: i have dumps of simple actions 10:28
connect, disconnect, open file
theorbtwo Great! 10:32
nothingmuch hold on, got to do it all again
theorbtwo Put them someplace http-accessable, or mail to james-at-mastros-dot-biz?
nothingmuch it was the argument to -w, and now you have all my other traffic aswell =)
theorbtwo Oh. Um, might want to do something about that. 10:34
nothingmuch uhuh, redumping is easiest 10:36
ok 10:37
FYI, i am syeeda.int.hyperroll.com
which is my laptop, internal net, 10:38
ip 10.1.1.xx
i forget
here is the progress, || denotes tcpdump was restarted
connect || close || connect || open file || move around file || add line || <--- overwritten some data, state ruined --> || remove line, by repeatedly deleting chars, and by selecting and deleteing || quit app 10:39
theorbtwo So you only have data from the <--- overwritten ---> bit forward? 10:41
nothingmuch no
i accidentally overwrote the dump where the line was deleted 10:42
so i redid that
but the state was not exactly the same
nothingmuch.woobling.org/subethadumps/
between add_line.dump and delete_line.dump the line was deleted and added
theorbtwo Ah, much much thanks.
nothingmuch good luck with it 10:45
and feel free to ask for more specific interaction as necessary
i'm going to lunch
theorbtwo Bon appitit. 10:46
DOH! 10:50
theorbtwo stops messing with nothingmuch, installs pearpc.
lumi Mornin' 10:51
Umm, I could send you dumps in ~ 10 hours
theorbtwo Mornin, lumi. 10:53
I can have a pearpc install up and running quicker then that, I hope.
nothingmuch FYI - subetha may be free for noncommercial use, but i'm not sure whether it's EULA allows this kind of stuff 10:57
you should check
nothingmuch would donate to theorbtwo if the said monk would produce a 'libnetedit' which every editor would bind to
theorbtwo The EULA is irrelevant; the wire is mine. 10:58
nothingmuch isn't this reverse engineering their protocol, which is sort of like data files?
buu Yes! 10:59
theorbtwo Yes; just like data files it's legal. If it weren't, the samba guys would have been sued into the stone age by MS.
nothingmuch hmm 11:00
so why are DVDs illegal?
program binaries?
if i disassemble the windows kernel on a sun machine, the machine code is no longer executable
i could argue it's not a program as far as i'm concerned
forget it
i don't really want to know this stuff 11:01
i wasn't asking
just be careful =)
theorbtwo The reason is that the copyright on the program files belongs to them. If I write something in Word, and save it, does MS have a copyright interest in my words? 11:03
Nope.
Juerd_ I'm sure that a careful reading of the EULA brings up a possible interpretation that says they own everything, including the user's thoughts for the time they use the program ;) 11:06
nothingmuch autrijus: ping 11:25
is(~%hash9, ("1\t2\n2\t3\n"), "hash can stringify"); 11:35
this passes
but it's not correct
is(~%hash9, ("1\t2\n2\t3\n" | "2\t3\n1\t2\n"), "hash can stringify");
this does not
not ok 53 - hash can stringify
# Failed test (t/op/hash.t at line 133, column 1)
# got: '1 2 11:36
2 3
'
# expected: '2 3
1 2
'
oh shit
i see
ok 52 - hash can stringify
not ok 53 - hash can stringify
ok(hash eq junction)
sorry, everyone!
12:08 Aankh|Clone is now known as Aankhen``
wolveria1 I wonder what the 'shall' auxiliary trait keyword in A12 does. 12:22
nothingmuch hmm 12:24
no clue
it's not in s12
or s13, which forked off 12:25
p6l?
wolveria1 Thomas Sandlaōæ½xDF's message in the thread Adding linear interpolation to an array wonders about it too, but I don't see any answers. hmm. 12:27
nothingmuch =P
nothingmuch just realized he's at work, and can run pugs tests on a 2x3ghz machine, instead of his 1x400mhz laptop 12:28
Aankhen`` Dual 3 GHz.? What CPU? 12:29
nothingmuch model name : Intel(R) Xeon(TM) CPU 3.06GHz
hyperthreading
twice
Aankhen`` Kewl.
nothingmuch uhuh
Aankhen`` You lucky dawg.
nothingmuch and that machine's swap disk just got zapped
so no one can use it for anything else either =) 12:30
Aankhen`` Awesome. :-P
nothingmuch is there an svn rpm for only the client somewhere?
theorbtwo Nice.
nothingmuch we also have some tougher boxes
at the US site there's some quad and 8 cpu aix and sun boxes
and we have an old alpha server 12:31
which in theory was much more powerful
it looks like a minibar
Aankhen`` I want an 8 CPU Opteron server.
And I want a life too, while you're at it.
nothingmuch ... make -j5 ... 12:35
=D
Aankhen`` ?
nothingmuch oh nothing
Aankhen`` Awww, c'mon, tell me... 12:36
nothingmuch compiles svn
wolveria1 svk? :)
nothingmuch svk - i've been having interface incompatibilities with it
i don't know how to use it yet
svn is like cvs which i already know
wolveria1 it is a bit odd.
nothingmuch and i have no time to learn something new
so it's on hold 12:37
even though i would like to use it eventually
time make optimized ... 12:46
116.590u 2.020s 2:02.58 96.7% 0+0k 0+0io 52996pf+0w 12:48
not bad
csh makes it undreadable
time make test 12:49
16.990u 1.330s 0:18.23 100.4% 0+0k 0+0io 104230pf+0w
woot
does the default var ($_) need any declaration in p6? 13:12
rgs I don't think so 13:13
fayland hcchien: if i find some mistakes at the zh_tw/S13.pod, what can i do? 13:19
hcchien_: maybe my mistake? should i contact the author directly? 13:21
nothingmuch fayland: change and commit, IMHO 13:25
if it turns out to be a wrong correction, it can always be rolled back
fayland nothingmuch: but my encoding is diffrent, is that ok?
nothingmuch ah
utf-8 is the norm
pugs won't eat anything else 13:26
perl6 says it's a good thing
and i think hcchien_ converted them yesterday
fayland but the words appears different
zh_tw is big5, i use gb2312 13:27
nothingmuch where is S13 in zh_tw?
both should be utf-8
if i understood correctly
fayland the words appears big5.the encoding is utf-8
nothingmuch ah
you mean it was reinterpreted as utf-8 13:28
but it really is big-5?
fayland appearance is different
nothingmuch and meaning?
also, duh
nothingmuch is a bit tired, please forgive
hmm, i think you'd better ignore me for now 13:29
fayland :) soory for my poor english.
nothingmuch i think i should apologize for my poor mental skills 13:33
scw fayland: Hello, what's the problem you discovered? I'm from Taiwan so may be able to help :)
fayland scw: such as "Likewise operators on your new type can appear in the class:" 13:40
scw: i upload my translation to zh_ch, may this helps. 13:41
scw: i think we need discuss that.
scw fayland: Are you talking about some file already in the svn tree? 13:42
fayland scw: yes. 13:43
scw: the done translation work.
scw fayland: So you are stanslating t/Synopsis/*.pod in to Chinese?
fayland scw: that's right.
scw fayland: I think the file should better use UTF-8 encoding 13:44
fayland yes. i use utf-8 encoding.
scw fayland: but with different codes for those words which are different between simplified Chinese or Traditional one 13:45
fayland: Just as what t/docs/zh-tw and t/docs/zh-cn do :)
fayland scw: so i upload my document to zh_cn.
scw s!t/!!g
fayland svn.openfoundry.org/p6l10n/zh_tw/S13.pod 13:46
i just think there may has some mistakes.
and ask what can i do
contact the author directly or something else?
scw fayland: if you can commit, commit directly should be ok :p 13:47
fayland scw: but i use diffrent glossary and the appearance of simp.chinese and trad.chinese is different 13:48
lumi Why are there Chinese translations? 13:50
I'm not complaining, I'm enquiring
scw fayland: different glossary? Do you means those word translated in different term? (ex: computer)
lumi Would other translations be helpful?
fayland maybe for the man who is not good at english
of course.
theorbtwo The zh_tw translation should be traditional chinese; the zh_cn translation should be simplified chinese, but both should be in utf8 encoding.
fayland scw: yes.i mean that 13:51
scw fayland: That should be ok, l10n including culture :)
theorbtwo As far as I know, there's Chinese translations and not others simply because there were volunteers for the Chinese translations first.
fayland theorbtwo: we are use utf8 encoding.but the appearance is different.such as ōæ½x98ōæ½xCBōæ½xEE} as trad. and ōæ½xB1ōæ½xEAōæ½xCCōæ½xE2 as simp. 13:52
scw: that's fine.:)
scw fayland: in fact, if there are some term different in simp. in pugs/docs/zh-cn, update directly, too. 13:53
nothingmuch lumi: thinking of hebrew? latin? finnish?
scw fayland: it is translate by trad. author and convert to simp. by program :)
nothingmuch thinks STM haskell should be the de-facto academic computation language on linux clusters 13:54
it can't be slower than the java overhead
which is a pill most are willing to swallow
but it looks so much more useful
lumi The only language I could translate well into is Hebrew, but I don't know of any Perl hackers who don't have decent English skills
fayland scw: autrijus ask me to change the terms. :) i'll do it later.
scw fayland++ # I am, in fact, afraid of translating :) 13:55
fayland scw: ask u a question. does pugs support the %= op? 13:56
yet
nothingmuch will write a test 13:57
scw Err, seams not.
fayland and what about multi in deep?
scw fayland: What do you mean? multi-what? 13:58
nothingmuch fayland: multi in deep?
fayland such as "multi sub infix:<%> (Us $us, Them $them) is deep { mymod($us,$them) }" 13:59
is deep
theorbtwo Objects aren't yet supported, so you can't do Us $us, Them $them.
fayland maybe i should write test myself
how about Int
multi sub infix:<%> (Int $us, Int $them) is deep { mymod($us,$them) } 14:00
ya, i write test now. see u all later.
scw I wonder if `is' is implemented or not :p 14:01
nothingmuch i think not
i submitted t/var/is_constant.t as a test for is applying 14:02
but t/op/is.t is not really there
ugg 14:10
t/op/arith.t is not nice
someone please sanity check r560 14:23
autrijus greetings folks 14:39
autrijus squashed more test failures
still testing, commit in a bit
theorbtwo Looks sane to me, nm.
lumi The failure is kinda useless, it always gives the testing sub's line number 14:40
autrijus hi stevan.
stevan hello autrijus 14:41
Coke-afk ~~ 14:42
hcchien_ autrijus: is length a perl6 build-in function?
14:42 Coke-afk is now known as Coke-DarnitThisI, Coke-DarnitThisI is now known as Coke_
autrijus hcchien_: string length? 14:44
hcchien_ autrijus: yes
dada hcchien_: no length anymore :-)
malaire wasn't length banned from perl6
dada AFAIK, yes
is spelled chars(string) or string.chars now 14:45
autrijus yeah, and underlying it's
theorbtwo Or bytes(string), depending which it is that you care about.
autrijus .bytes, .codes, .graphs, .langs
the "chars" one is aliased to one of the four
depending on the unicode supporting level.
dada wasn't that .graphemes?
autrijus no it's graphs. 14:46
dada oh, ok
autrijus: one question
autrijus oh btw, my ($a, $b) = (1,2) now works...
I'll commit in a bit
that is my first use of parallel comprehension
and it rocks :)
dada how can I distinguish an arrayref from an array?
stevan nice 14:47
scw autrijus: It works!!
autrijus dada: in general you cannot.
stevan actually array-refs usually show up as List types
autrijus I guess .ref maybe. but that's blurry.
dada mmm
I mean in Haskell 14:48
err, in the Haskell part of pugs, that is
the fact is 14:49
that to implement the splat operator
I need to distinguish an array from an arrayref from a list 14:50
autrijus oh. arrays are VArray
refs are VRef
you can use case analysis.
dada cool, thanks
and lists are VList?
autrijus right
t/01basic........................ok 14:51
2/17 unexpectedly succeeded
cool. seems to work.
my (@a, @b) = (1,2,3); # what happens?
does @a get 1,2,3? 14:52
dada I think so
autrijus my (@a, @b) := (1,2,3); # @a get 1, then?
dada no, wait 14:54
mmm, yes, it should
(@a, @b) := (@b, @a); # "this should swap two arrays" 14:55
autrijus okay. that's what I thought too.
I wonder why it's not an infinite loop ;)
dada autrijus: another question
larsen autrijus: me too 14:56
isn't := the binding operator ?
dada larsen: sure it is
larsen then shouldn't I use = to swap arrays ?
dada no
because = flattens
(@a, @b) = (@b, @a); # this would concatenate @b and @a and put them in @a 14:57
lumi And truncate @b? 14:58
dada probably, yes
autrijus: consider this: my *$a = @a; 14:59
autrijus: this should be "translated" to: my ($a) = @a;
autrijus: this means that the * should change the context in which the assignment is evaluated
autrijus yeah.
dada any clue about how to do that? :-) 15:00
can I assign to envContext from inside the reduce function?
autrijus sure 15:01
grep for enterContext
dada ahhh, yes
enterContext does that?
autrijus yup. 15:02
it's not assign, mind you
it's a pad
dada wait a moment
autrijus the difference being it reverts back after the scope exits.
dada is enterContext different from enterEvalContext?
autrijus see "All About Monads", the reader monad
enterEvalContext enters a context
then eval something.
see its type
enterContext enters a context
dada ok 15:03
autrijus it takes an action
eEC takes Exp
enterEvalContext :: Cxt -> Exp -> Eval Val
enterEvalContext cxt = enterContext cxt . evalExp
dada autrijus: "see its type" sounds like "see how many legs martians have" to me :-)
autrijus I hope that is clear :)
enterEvalContext is just, first enter the context specified as first arg, then eval the exp in the 2nd arg. 15:04
dada sure, I get it
enterContext :: Cxt -> Eval a -> Eval a
autrijus it takes a context, and another eval action, then perform that action under the specified context. 15:05
clear as ink!
dada mmm
but I have no action to perform
autrijus you have something to eval, don't you?
the @a
dada yeah
no, rather the $a
autrijus aye 15:06
dada my *$a = @a;
ok, I will try to code it now :-)
autrijus my ($a, $b) := ();
what happens?
runtime error? 15:07
dada mmm
maybe it's just like ($a, $b) := (undef, undef);
so it's like $a := undef 15:08
autrijus so $a and $b are constant undefs?
dada yes
autrijus i.e. you can't assign to them?
ok, implemented as such
it's actually
$a := not()
theorbtwo That seems like very useless behavior...
autrijus so we get
@b := not()
as well.
dada I'm not sure about it 15:10
binding should be (more or less) the same binding used for subs 15:11
autrijus I mean,
@b := undef
will result in +@b being 1
which is certainly what we want.
so not() is right.
PerlJam autrijus: I think that my ($a, $b) := (); should be an error. 15:13
(not that what I think counts for much :-) 15:14
autrijus I was trying to find it in specs.
Apocs or Syns.
maybe warrants a p6l post again :-/
PerlJam I think you won't find it. IIRC, that behavior is up for grabs.
actually, maybe it shouldn't be an error, but also $a and $b shouldn't be bound to anything 15:16
theorbtwo I think my ($a, $b) := (); should be error, or nop... but I don't know what makes the most sense in.
@a=(); my ($a, $b) := @a;
PerlJam theorbtwo: I think nop makes the most sense now.
autrijus I'll use const undef for now
until a ruling is made on either p6c or p6l.
PerlJam theorbtwo: that would make $a an alias for @a and $b is bound to nothing 15:17
autrijus (read: easiest to implement)
PerlJam (ie, it's free)
autrijus s/const undef/const not()/
theorbtwo Do we have a mechinisim in place for generating warnings?
autrijus no we do not. do you want to design one? 15:18
ok, I think multiarg LHS for both := and = are done. 15:27
hash_ref fails, though. weird
my $a = (1,2); 15:29
is $a 1?
or [1,2]?
theorbtwo Are we checking ($a, $b)=($b, $a) ? 15:30
autrijus prolly not. 15:31
my () = ();
syntax error?
I'm making it a syntax error. 15:32
rgs [rafael@grubert ~]$ perl -e 'my()'
Can't declare stub in "my" at -e line 1, at EOF
nice error message :)
autrijus C:\work\pugs>pugs -e "my () = ()" 15:33
unexpected ")"
expecting variable name
pugs is superior! :)
wolveria1 autrijus: re: my $a = (1,2); I think apos say now that a list in scalar context doesn't flatten 15:36
15:36 wolveria1 is now known as wolverian
autrijus ok. 15:36
wolverian so $a would be a List presumably.
autrijus let me get this clear 15:38
my $a = (1); # List
my $a = 1; # Int
correct?
wolverian I have no idea. I hope it's not (1,) ;) 15:39
let me check the apos.
autrijus (1,) is sad ;) 15:40
(and I know where it comes from.)
wolverian (I share your dirty secret.) 15:41
autrijus cackles 15:42
this is too complex :( 15:43
and not well specced. we need tests!
(and post those tests to larry)
starting from
my () = ();
my $a = ();
my ($a) = ();
my ($a) = (1);
my ($a) = (1, 2);
my ($a, $b) = (1);
etc
and also s/=/:=/g 15:44
wolverian hrm. you need to ask this on p6-l
or I can do that, if you want
autrijus please do that.
autrijus is still catching failing tests left and right
wolverian as far as I can understand, the apos say that , in scalar context generates a list
autrijus oh, and also use @ somewhere too. 15:45
riight...
(1,) then?
wolverian or list(1)
autrijus shudders
or [1]
wolverian yes.
autrijus just to be safe
wolverian oh, why doesn't mutt have a history of addresses I've sent to :) 15:46
autrijus my ($a) is different from my $a it seems... 15:47
or I'm totally on crack.
my ($a) = (1,2);
my $a = (1,2);
my ($a,) = (1,2);
;)
ingy hola
autrijus hola ingy.
I'm on crack it seems :-/
the blue sugar cubes things 15:48
ingy haha
autrijus hi Shillo. 15:49
ingy going to breakfast&
Shillo Hullo!
Was reading the IRC logs and then asked myself why wouldn't I just connect. :)
autrijus =) 15:50
so, first time here?
theorbtwo First time caller, long-time listener?
Shillo Miroslav Silovic from the mailing list... I followed up some of your posts. Not very creatively. ;)
autrijus =)
cool!
Shillo First time here... ummm.. don't remember. I think no, but the last time wasn't much happening.
autrijus now here is a very happening place.
Shillo Yeah, saw in the logs. 15:51
autrijus decides to split ruleVarDeclaration into ruleVarDeclarationSingle and ruleVarDeclarationList.
theorbtwo It tends to be a lot more interesting around here when autrijus is up.
Shillo Hehe. :)
theorbtwo ...and about.
autrijus nah. I also have my fair share of backlogging :) 15:52
Shillo just wants Perl6 to have CLOS-style method combinators.
Easy to make me happy. ;)
autrijus write tests! write dialects! :)
or, just try to explain them here :)
Shillo He! use Lang::CL :)
autrijus there is no Lang::CL on cpan 15:53
6pan, maybe? :)
Shillo CL -> Common LISP
What impressed me about Perl6 is that it's the first mainstream language that has most of it.
(while staying practical) 15:54
pjcj Perl6 is mainstream?
Shillo pjcj: :p More than CL.
autrijus perl6 is practical?
hcchien_ t/op/sub_return_values...........ok # cool 15:55
Shillo Oh, method combinators...
method bla(Dog $x) is before {...}
method bla(Dog $x) is after {...} 15:56
method bla(Dog $x) is around {...}
method bla(Dog $x) { ... }
Order of execution of these would be 4, 1, 2, 5, 3
wolverian autrijus: sent.
autrijus cool.
nothingmuch 16:40] lumi: The failure is kinda useless, it always gives the testing sub's line number
autrijus oh. I thought it already has PRE and POST blocks
that you can hook on 15:57
Shillo If around calls next_method();, that is. Otherwise, just the 4th would run.
wolverian oh, wow, there is already a thread about those.
nothingmuch this is a bug tested for in t/magical/callerposition.t
wolverian dammit. :)
Shillo No, note that all the methods have the same declarations. They're all supposed to run.
autrijus dammit. sorry.
ah. multiple dispatch that all happens 15:58
I like it.
wolverian well, I'm a bit more specific than Aldo Calpini's post
so I don't think it's dangerous.
Shillo This becomes important if signatures aren't the same. For example, around method runs first even if it's not the closest match.
(and you want to use these to catch exceptions from all the other methods, or do BEFORE/KEEP/UNDO blocks for the whole thing) 15:59
15:59 rafl_ is now known as rafl
Shillo This kinda ties into Luke's post from today, about making MMDs into real objects. 15:59
autrijus r561 landed!
* lone blocks without trailing semicolon now runs; multi ple variable assignment now (sort of) works
we're down to 3 failed test scripts. 16:00
Shillo autrijus: Woot! :)
nothingmuch w00t
autrijus: i started some rules tests 16:01
starting with the really really basic crap
then converting t/op/re_tests from perl5 (doing that now)
autrijus nothingmuch: try reuse ambahal (?)'s tests
he already converted all for us
cls_bsd great!!
nothingmuch ah
i didn't know that
where are they?
autrijus see past p6c topic
a sec
nothingmuch ah
ok, i'll just play integrator then =) 16:02
in that case, i can get on to real grammars =)
autrijus nothingmuch: www.nntp.perl.org/group/perl.perl6.compiler/131
ah, so it's our very own malaire 16:03
I hear he've generated them by a script
so please coordinate. I'd very much like them to all land to pugs. 16:04
nothingmuch will do
autrijus procs the remaining three failures 16:06
nothingmuch ponders how to attack OOP tests 16:07
autrijus no p5 for those?
nothingmuch there are 16:08
t/op/method.t
but this is not very, err, perl6ish
autrijus time to play porters? :) 16:09
well, another vector is exegeses
nothingmuch i toyed with unit testing some stuff 16:10
but i think it's too early to start
the exegesis are purposely broad
and cover a whole lot of everything
autrijus it's not too early... OO is the next thing in sight
nothingmuch not for OO tests
for exegesis examples
autrijus ahh.
A12 also contains many test stuff. 16:11
nothingmuch for example, e02 has autochomping, is constant, my Int, sub sigs, traditional subs
i'll look at a12 more closely
Shillo Going through tests... #LEAVE { push @leve, [ $i, +@enter ] } <-- should be @leave. Trivial but anyway. :) 16:12
autrijus Shillo: if I give you committer bit, will you fix it? :) 16:13
Shillo No. I'm behind a firewall, miles high. :/
This is IRC over http. :(
autrijus eh? HTTP POST should be fine, no?
pugs uses a http as repository.
Shillo Hmm... Need to learn svn obviously. 16:14
Need to /install/ svn, too. ;)
autrijus subversion.tigris.org
or, tortoisesvn if you're on win32.
Shillo Just did urpmi subversion. ;)
Mandrake's currently sucking. 16:15
autrijus nod.
Shillo Sucking as is *sllllluuuuuurrrrppp*, not at in *sux* ;)
*grumble* Need to do urpmi.update 16:16
nothingmuch Shillo: - sorry, i couldn't really syntax check that since the line is commented out =)
i won't fix it though, so that you're forced to install svn/svk ;-) 16:17
Shillo Yeah. I noticed, 'swhy I said it was trivial.
Lemme actually take a good look at the whole thing before I accept the haching rights.
autrijus Shillo: also give me your email when you're ready :)
Shillo [email@hidden.address]
(that's for later ;) ) 16:18
autrijus har.
Shillo Hmm, need ghc, too.
autrijus ghc should also be part of urpmi'able things. 16:19
Shillo MDK repo has it.
Yeah.
autrijus actually, I think pugs is also urpmiable.
thanks to a certain rgs here.
Shillo Heh.
Although... I plan to switch to Ubuntu at work soon.
As in, the next 2-3 months.
rgs mhh
Shillo MDK 10.1... well... how to say it... Windows ME was positively stable and bug free in comparison. 16:20
rgs ubuntu is such a small distro. I'm sure they don't even have ghc.
autrijus looks at rgs :)
rgs but you have a point.
autrijus yeah, but it can have debian universe.
Shillo rgs: I have a 25 gigs of archive.ubuntu.com mirrored on my local drive. :)
arcady you can have the union of ubuntu and debian
it actually sort of works
rgs right.
arcady for such useful things as X.org 16:21
Shillo rgs: Although, to be honest, I'm really not sure why 10.1 sucked so badly for me. It just /can't/ be bad QA, some things were way too obvious.
Like, kdm hanging.
Or, zsh dumps core.
rgs ooh, weird.
Shillo NForce 2 chipset could be the reason, but that would just introduce flakiness, not insta-deaths. 16:22
rgs several in-house developers here use zsh. core dumps would have been noticed.
Shillo And then there were just plain weird problems.
-- I used prompt module.
Worked without it, but with a strange 1 sec startups.
As for the just plain weird... bluetooth dialup. 16:23
Spend 3 days trying to debug that. Traffic dumps were crazy.
It'd connect, exchange network setup, run the ipup script (I could be very wrong about the name of this one - the one that gets run after the machine IP address and DNS come down the pipe) and then it'd promptly disconnect. 16:24
(not that Bluetooth dialup is common, but still)
Ick! Need to go, and urpmi ghc has half an hour ETA. Talk to you all in a day or two! :) 16:25
nothingmuch ciao Shillo
those tests won't pass anytime soon anyway!
autrijus grins. 16:26
Shillo nothingmuch: :)
nothingmuch autrijus: different plans?
autrijus nothingmuch: no, same w/ you
nothingmuch ah
autrijus my plan now is to squash test failures
nothingmuch was getting optimistic 16:27
autrijus and try/catch/given
and reapply newest p6l clarifications.
small things.
but important to get the overall quality up.
nothingmuch (smartmatch!)
try can be just try, btw
autrijus I think it's easy enough to be done as an exercise for lambdafolk
nothingmuch without catch, for the meanwhile 16:28
PerlJam nothingmuch: doesn't it almost have to be because catch needs smartmatch semantics to really work?
I guess we could fake it with strings though 16:29
autrijus it's already stubbed w/ strings now
nothingmuch if catch is not a given block now, but just silently takes anything 16:30
i.e. try {}
which is eval {} for perl5 heads
then try works
without smartmatch
autrijus try{} always need to be followed by catch? 16:31
nothingmuch no, try without catch is like try with catch anything
autrijus ahh.
nothingmuch however, true catch requires given, which in turn requires smartmatch
which probably likes to have isa
and rules
autrijus nothingmuch: you want to implement try{}? 16:32
I can walk you thru :)
will take ~2min
nothingmuch i think my mind is too cooked to understand what i'm doing
so don't waste it on me =)
wolverian there is no catch
there's CATCH { }
autrijus there is no spoon
theorbtwo There is only fork().
autrijus ^W^W/me sobs.^W^W^W^W^W^W^W^W^W^Wnice :) 16:33
gah. control characters. bad ssh.
(or bad screen, or bad irssi.)
theorbtwo Ah, was just about to ask.
wolverian try { ...; CATCH { when StupidError { puts "U STOOPID!" } } }
er. s/puts/say/ ... 16:34
autrijus parses "ST", "OO", "PID"
wolverian (too much ruby for me.)
nothingmuch wolverian: see t/base/try.t
i think it was the first i wrote
PerlJam autrijus: ^U is your friend though ;-)
nothingmuch or at least the first i started 16:35
autrijus ;)
try{} parsed.
doing eval.
wolverian nothingmuch: nice :)
gotta run. cheers. 16:37
nothingmuch ciao wolverian!
autrijus see ya 16:38
implemented. 16:40
testing
nothingmuch thanks! 16:41
theorbtwo Ha! My pearpc box finally has working networking! 16:43
nothingmuch theorbtwo++; # subetha <-> xemacs in 8 hours! 16:44
autrijus gawd I love ContT.
theorbtwo 8 hours?
autrijus "try" -> do
val <- resetT $ evalExp (head exps)
case val of
VError _ _ -> retEmpty
_ -> retVal val
that's all it takes.
theorbtwo I think that's a little bit optimistic. I'm not autrijus. 16:45
nothingmuch ;-)
autrijus ok($lived, "try weakened death"); 16:46
nothingmuch dances
autrijus r562 16:48
nothingmuch yay yay! 16:49
hcchien_ should we put t/0*.t to t/base or others? 16:50
nothingmuch now you you don't owe me anymore
i'll have to write new tests
16:50 _metaperl is now known as metaperl
autrijus hcchien_: I think so now 16:58
16:58 hcchien_ is now known as hcchien
nothingmuch base seems to be basic stuff 17:10
perhaps not base, but something else?
i think it should be put in the blue sugar cube dir
hcchien nothingmuch: I think so. :) 17:11
metaperl_ can I create a hash whose values are arrays and derefence them in pugs? 17:18
and can I use each or keys on a hash in pugs?
autrijus 1. multidim is not there.
2. you can use
for %hash.kv -> $k, $v { ... } 17:19
malaire btw, I put my perl5-script which converts (actually: tries to convert...) re_tests into PGE-format to laire.info/markus/perl/re_tests.html , just in case anyone is interested in hacking with them. (I'm currently trying to learn Haskell, and I'm not sure whether I'm going to fix/change these scripts anytime soon...) 17:20
autrijus looks at nothingmuch 17:21
malaire: thanks!
nothingmuch sorry 17:24
nothingmuch will ttyl
dada how do I "cast" an Array to a List? 17:32
or, how do I get all the elements of an Array?
is enterEvalContext "List" vals enough? 17:33
or do I need doSlice? 17:34
autrijus you need a fromValue.
val <- enterEvalContext "List" exp
now val is of "Val" type
dada mmm ok
autrijus to get a [Val] from Val:
list <- fromValue val
in fact, you can use fromValue to obtain almost any time.
haskell well figure it out based on how you use it later. 17:35
if you want annotation:
(list :: [Val]) <- fromValue val
but you won't need them in most of the casts.
metaperl_ has the fat comma lost its stringification semantics? 17:37
my %mates = (
'jane' => 'joe',
sally => 'bob',
janice => 'bill'
);
autrijus I think so. 17:38
metaperl_ sally => 'bob' # yields parse error
autrijus not very sure.
metaperl_ oh
dada no no
fat comma should still do stringification
autrijus But C<< => >> still autoquotes any bare identifier to its immediate
left (horizontal whitespace allowed but not comments).
mea culpa.
so, a bug.
gah, even more lookaheads.
autrijus hugs parsec
metaperl_ autrijus, please look at this: hcoop.net/~terry/perl/talks/p6-junc...pleiter.p6 17:39
the for is throwing a long error message
my pugs is for 562 release
theorbtwo Put parens around the listy bit of the for. 17:40
Er... Hm.
metaperl_ bingo
refresh the URL to see a working program
thanks theorbtwo 17:41
autrijus oh btw. feel free to replace eval''
to try{}
dada I didn't got it right
autrijus in tests that has runtime error.
instead of parse error.
dada vals <- if isaType "Array" exps
then fromValue (enterEvalContext "List" exps)
autrijus I think the then needs to be indented after if 17:42
I think.
also, it's monadic
so you need
vals <- if isaType "Array" exps
then do
dada actually, it says
autrijus val <- enterEvalContext "List" exps
dada Probable cause: `isaType' is applied to too few arguments in the call
autrijus fromValue val 17:43
i.e. you can't just nest fromValue with enterEvalContext.
need a "do" expression.
(or a =<<)
dada yes 17:44
but I still miss a parameter to isaType...
autrijus you need cxt
I think.
you can obtain it from
cxt <- asks envContent
err wrong
you need cls
cls <- asks envClasses
dada I see, I see 17:45
autrijus ctx <- asks envContext
if isaType cls "Array" cxt
etc.
stevan so try {} is working, but not CATCH right?
autrijus please abstract them away.
I want a
bool <- currentContextIs "Array"
or something like that.
but you can roll by hand too. 17:46
stevan: yup.
stevan autrijus: so where is the error? $! $@
dada autrijus: no, I don't need the currentContext. I need to know if exps contains an array (or an arrayref) 17:47
autrijus stevan: should be in $!. a sec. 17:48
dada: oh oh.
that won't do then.
you need something more clever...
stevan autrijus: also this will die pugs -e 'my $val = try { "foo" }; say $val;' 17:49
autrijus valType val
this should work
i.e. it's a function that returns a stringified representation of val's type
metaperl_ can a hash have an array reference as the value of the key? This isn't working as I expect: hcoop.net/~terry/perl/talks/p6-junc...all-all.p6
stevan I am not sure if try{} will return anything, but I think it should not die
autrijus metaperl_: is shape() is not yet implemented. 17:50
tests welcome!
metaperl_ so I cannot use array references as hash values?
stevan metaperl_: t/op/multidimensional_array.t has some tests for this stuff in it, if you want to add to them
autrijus stevan: try{} is only a construct now
metaperl_ also, maybe I will create a test case for a hash having a trailing comma 17:51
autrijus it's not yet a proper Syn.
stevan: i.e. can't be Expr.
metaperl_ checks
autrijus stevan: just toplevel Stmt.
stevan autrijus: ok
metaperl_ %hash = (a => b,) ; # fails
stevan waits patiently :)
metaperl_ perl5 allowed that dangling comma
PerlJam metaperl_: perl6 will too at some point. 17:52
autrijus metaperl_: right. it's pugs' bug. just adda test :) 17:53
or grep for one.
stevan: r563. $! is now set.
autrijus starts to write today's journal
hcchien there is no $/? 17:54
PerlJam um ... perl6's $/ means something different than perl5's $/ which is now a method/attribute on the filehandle (I think) 17:56
crysflame journal? public?
PerlJam crysflame: use.perl.org/~autrijus/journal
crysflame neet 17:57
autrijus larry wants my ().
we'll do my ().
PerlJam What's my() mean?
dada sgrunt 17:58
Couldn't match `Val' against `[Exp]'
autrijus my () = (foo())
it could be reduced to
list(foo())
dada: mm? 17:59
dada vals <- case valType exps of 18:00
"Array" -> do
I can't do valType exps, because exps is [Exp]
autrijus yeah.
dada and valType wants Val
autrijus that is correct.
so you may wish to persumably evaluate the exps 18:01
into val
or something.
you can't tell, from looking at an exp
PerlJam dada: looks like you need to iterate ... er recurse ;)
dada in fact, I need to check before that exps contains only 1 thing, and that thing is an array
autrijus whether it will return a list or not.
obra 'morning
autrijus dada: ok. you can do that as:
case exps of
[Val v] | valType v == "Array" -> do 18:02
_ -> do
dada mmm cool
autrijus yo obra.
dada now that 18:04
what is an "Eval a" type? 18:05
Expected type: Eval a
Inferred type: [Exp]
autrijus you can cast a to Eval a 18:06
via "return"
PerlJam dada: what are you hacking into pugs?
autrijus you can cast Eval a into a
via "<-"
dada PerlJam: unary *
PerlJam oh! dada++ keep ahcking
er, hakcing
autrijus "retVal" is also commonly used if you are returning a value.
PerlJam er, hacking dammit!
Limbic_Region notes autrijus's sleep cycle slowly shifting in phase with the US and notes pretty soon he will be sleeping when everyone else is awake
PerlJam Limbic_Region: or he just won't sleep.
autrijus that too.
Limbic_Region s/everyone else/everyone else in Taiwan/
obra that better not happen. since folks from .us and .eu will be in .tw soon 18:07
luqui are we going to get slice assignment working soon?
I've been trying to figure out how to do it... unsucessfully 18:08
s/c/cc/
autrijus luqui: 18:09
my ($a, $b) = (1,2);
this works now.
($a, $b) = (1, 2); # not.
need to unity both. 18:10
now we have a ruling from larry, that is
luqui 'my' should be an inline operator 18:11
my ($a, $b) is like do { my $a; my $b; ($a, $b) } everywhere, barring scoping issues
rgs my($a, $b) would be (my $a, my $b) 18:12
luqui right
which is illegal 18:13
since you can't use my in an expression
(but it shouldn't be)
Something of a problem for the parser is the ability to do foo(my $x, $x)
you have to be smart about how the parse tree communicates scoping stuff 18:14
cabal design call today; is there anything I should mention? 18:15
PerlJam That's an interesting way to say "undef"
luqui if foo accepts references, it might not be
autrijus luqui: parrot AST? :)
luqui er, rw params
autrijus never mind
luqui: MMD is still blurry. I think it's the most blurry part. 18:16
luqui as you can see, I'm working on fixing that
We'll definitely be talking about my recent thread
autrijus yes you are. I'm eager to see new specs.
on and get unary = committed in Syns
if you had not yet done that.
s/on/oh/
luqui I'm on it 18:17
autrijus 1395 tests now. fear
hcchien tests++
luqui wow
how many does perl have?
dada countless
autrijus much more.
PerlJam luqui: is the meaning of ($a,$b) := () well-defined? 18:18
rgs the magnitude order is 100 * more
Juerd_ say "no_plan"; say "$_ ok" for 1...;
Testing infinity ;)
rgs but it depends on platform, compilation options etc.
autrijus Juerd_: lol.
hcchien I need a faster laptop to do so many tests, since I have to make three times a day. :p
luqui PerlJam: yeah, error
Juerd_ How else would you test that it really loops from 1 to inf? :)
luqui it's like sub foo ($a, $b) {...} foo()
autrijus hcchien: "make clean; perl Makefile.PL; make optimized"
PerlJam oh good. 18:19
autrijus luqui: try write that down!
luqui: I mean, the := and = semantics
the "is like" thing is of crucial importance, yet glossed over in Syns
Juerd_ autrijus: := uses the same semantics as sub signatures
autrijus Juerd_: I know.
Juerd_ ok
luqui fixing unary = first
but ok
PerlJam Juerd_: that fact is just sub-documented.
autrijus ($a, $b) := ($b, $a);
vs 18:20
@fib := (1, 1, (@fib Y @fib[1...])).map:{ $^a + $^b })
(eager vs lazy evaluation)
one is wrong.
not sure which one is.
luqui isn't either 18:21
autrijus @fib := (1, 1, (@fib Y @fib[1...]).map:{ $^a + $^b })
luqui: so bring it up to cabal :)
that is high level enough I think :)
"abstract policy"-wise, that is
PerlJam I'd think that second one needs a little more syntactic help to manufacture a generator (but that's just MHO)
luqui starts a list of things to talk about today 18:22
autrijus PerlJam: it Just Works in haskell :) 18:23
PerlJam: but yeah, if syn help is needed, I'd like to know how it looks like.
luqui PerlJam: map is a generator 18:24
another thing is whether it's COW. Not an issue in haskell
dada autrijus: I don't understand how doSlice works
doSlice :: [Exp] -> [Val] -> [VInt] -> Maybe (Val, [VInt])
luqui @fib = (1, 1, (@fib Y ...))
@fib[3] = 42
dada autrijus: why 3 arguments?
luqui what's @fib[4] 18:25
Juerd_ Let's just hard code @fib
The numbers won't change any time soon ;)
PerlJam heh
autrijus heh
that's an legacy
just put [] in the first slot.
metaperl_ so how do I iterate through an array, assigning each element of the array to a variable i and execute a block using i
autrijus sorry for that, btw.
dada autrijus: ok
autrijus I'll refactor it.
autrijus needs to sleep :) 18:26
metaperl_: foreach surely?
obra sleep is good for you, autrijus
PerlJam and when you say @fib[50] = 42; does it have to generate all of @fib[1..49] first?
obra you'll be a 10x programmer when you wake up.
Juerd_ metaperl_: for @foo { my $foo = $_; { ... }($foo) }
metaperl_ perl6 has no foreach
autrijus for, then. same difference
Juerd_ I don't see the point in not using aliasing
luqui he's only 8x now
metaperl_ checks t/for.t 18:27
luqui Juerd_: what do you mean?
hcchien 20x, but make some mistakes? :p
autrijus okay, foreach is now illegal.
committing
luqui is 0.5x with ECC
Juerd_ luqui: "assigning ecah element ..." 18:28
Why assign.
autrijus: &foreach ::= &CORE::for; # now it has ;)
autrijus except for is a Syn
not a App
so doesn't work.
you need macros for that.
now, if you can formulate for into a App... I'm all ears :) 18:29
luqui autrijus: I don't think for is a syn in real perl
PerlJam autrijus: template haskell :)
autrijus luqui: oh?
luqui: I'd like to know which is Syn and which is not.
luqui it's a statement:for, and it has standard statement parsing
wolverian yay, lwall replied.
autrijus wait, wait, statement: ?
luqui or something like that
autrijus never mentioned anywhere.
certainly not in synopses.
I can only implement specced things :) 18:30
luqui hmm so it's not
luqui adds that to list
metaperl_ I cant get this program to run... I think the error message is about the wrong lines: hcoop.net/~terry/perl/talks/p6-junc...all-any.p6
autrijus I'll really sleep now. have fun, and may your fruitful discussions bore crystal clear semantics!
autrijus waves &
obra night, autrijus 18:31
luqui night
PerlJam luqui: Are you working with pmichaud on pge? Or just the grammar and specifications?
luqui try $epsiolon > all(...)
PerlJam: I sortof am, but not actually 18:32
I was going to, but pm likes to commit in big chunks
so it's hard to make incremental progress
dada anybody knows how to coerce a `Maybe (Val, [VInt])' into a `Val'? :-)
luqui I'm sitting around until we have the machinery to write perl 6 in perl 6
then I'll get to do some serious work
(I'm already doing a little bit of that) 18:33
PerlJam luqui: That's strange. His other project (pmwiki) gets released early and often (so much so that there will sometimes be multiple releases in a single day)
luqui hmm, yeah, that is strange
PerlJam luqui: what pugs work are you doing? 18:34
PerlJam is interested in "merging" the development effort and thinks that pugs + p6rules is the quickest way to get us bootstrapping 18:35
luqui nothing in particular on the queue right now. I'm gradually doing builtins.kwid
PerlJam: no kidding
dada luqui: AAARGH
luqui that's excellent, it's exactly what we need
dada: hmm?
obra PerlJam: I think that everybody's already drawing from everybody else. patrick's certainly finding pugs useful
dada luqui: your reply on p6l
luqui: about *$a on LHS being illegal without a bind 18:36
PerlJam obra: sure, but if the people that know p6rules inside and out can get to implementing them in haskell, that would be best IMHO :)
luqui with my behind it, yes.
dada: why aaargh?
dada luqui: because I'm trying to make it work :-) 18:37
luqui PerlJam: the problem is that those people don't know haskell very work
s/work/well/ :-)
(reading and typing at the same time ;-)
dada luqui: and you're adding a whole level of complexity I didn't thought of
luqui the fact that that damn binding needs infinite lookahead kind of complexity? 18:38
dada no, the fact that I'm implementing unary * without worrying about where it is used
PerlJam luqui: indeed. Which is why I'm asking questions. Are you learning haskell? I'm wondering how much prodding I should apply towards pmichaud regarding learning haskell :-)
luqui PerlJam: I'm doing my best to learn it 18:39
I actually took a stab at starting the p6rule implementation the other day
and failed
PerlJam luqui: heh ... me too :)
PerlJam doesn't grok enough haskell yet either 18:40
luqui dada: I'm thinking, since binding creates variables, that we should have a bind operator
that takes special syntax, argument list syntax
Don't worry about the binding unary * yet
it's an ugly part of the language, and it ought to change 18:41
luqui adds that to the list
PerlJam but I was reading about template haskell last night and I'm completely convinced that haskell has all the machinery to not only make it happen, but for someone with the right haskell-fu, to make it happen relatively easily
luqui template haskell?
PerlJam www.haskell.org/th/ 18:42
luqui We might not even need that
we parse with monads, our return values are other monads
crysflame PerlJam: you've found autrijus's supply of crack!
crysflame phears
also i agree 18:43
luqui the problem I ran into was embedded codeblocks
because I don't know how Eval works at all
metaperl_ here is the talk on Perl 6 Junctions I am giving at our Perl Mongers meeting tonight. Thanks guys for all your help: www.hcoop.net/~terry/perl/talks/p6-...index.html 18:44
luqui metaperl_: nice 18:45
metaperl_ thanks
:)
PerlJam slide7 seems to have some spurious code at the top
(duped @pacific)
metaperl_ checks
oh yeah, I'm going to comment out the second @pacific 18:46
and show how it works differently
I couldnt use a hash with array refs as values or I would've done it that way
luqui & class 18:47
PerlJam seeing "$pass = =$IN" just makes my eyes hurt
(and shouldn't it be $*IN anyway?)
metaperl_ how about $pass = $IN.next
the * is optional
luqui in the absence of another declaration of $IN, $IN and $*IN are the same
metaperl_ but you are right it should be there
PerlJam is a fan of explicitness sometimes
luqui PerlJam: revel in the fact that $pass = =$IN isn't officially documented yet 18:48
PerlJam It would be cool in a sick sort of way if "$pass =$*IN" just worked :-)
luqui & class # really
metaperl_ wont $IN.next work?
many people are not gong to adjust to all the changes in Perl 6. they are going to be upset 18:49
but they will have nowhere to run because Perl will still be the most convenient power tool out there
theorbtwo Nowhere to run? They won't need to, just stay at perl5. 18:50
PerlJam or start using ruby
dada or (*cough* *cough*) piethon
theorbtwo But so long as TIMTOWTDI, they will hopefully find one that they like. 18:51
metaperl_ all it takes is looking at the usermanual of any other LAMP language for about 5 minutes for me to be grateful for Perl and comeback
PerlJam the more I look at "$x = =$*IN", the less convienent that shorthand seems.
I'm glad it's not documented.
metaperl_ compare with $x = <STDIN>
ihb PerlJam: what does it do?
metaperl_ ihb, read a line from stdin 18:52
dada ihb: what metaperl_ said
PerlJam ihb: unary = is the "read a record" operator
ihb wow
metaperl_ no, it's the iteration thingy
Schwern I've been confused with Randal.
metaperl_ oh I guess that's the same thing
Schwern I don't know if I should be insulted or flattered.
theorbtwo I'd go with insulted, m'self.
metaperl_ Randal is a great person.
unless they think you were the one hacking into Intel 18:53
Schwern <Darren_Duncan>Also, about 1.5 years ago, a Linux cruise stopped by in Victoria and I briefly met ... Mike Schwern (name?) ... and I met Linux Torvalds
<Darren_Duncan>not mike schwern ...
<Darren_Duncan>Randal Schwartz I met
PerlJam heh
Schwern metaperl: Or if you're within 50 meters of a Karoke club
metaperl_ i dont get that about karaoke.
PerlJam Schwern: flattered wouldn't be the right thing. 18:54
metaperl_ CGI::Prototype is awesome. Have you tried it?
Schwern Flattened?
metaperl_ I think it is what we needed for Perl/CGI long ago
Schwern metaperl: Randal likes to sing Karaoke.
PerlJam Schwern: neither would insulted. I'd just chalk it up to Darren_Duncan being really confused and/or drunk ;)
metaperl_ oh, I didnt know that... I can't imagine Randal singing.. that is terrifying and horrifying
does he have a good voice? 18:55
PerlJam metaperl_: go on a geek cruise and find out! 18:57
metaperl_ Schwern and Randal to look very similar though if you ignore the difference in height and hair length 18:58
and body fat
PerlJam metaperl_: what?!?
metaperl_ lol
jusssssst kidding
PerlJam metaperl_: merlyn and Schwern look the same ... on IRC. If they're using similar nicks. And speak in short sentences.
ingy Schwern: you should change your nick to schwerlyn 18:59
PerlJam ingy: heh!
Though that sounds like something that jock do to geeks in high school.
"Hey man, let's go give that geek a schwerlyn!"
ingy haha
metaperl_ schwerlyn.... lol 19:00
and if every other sentence isn't "see my article on that subject"
Juerd_ Haha
metaperl_: Have you seen that PM writeup that is a merlyn-bot? 19:01
(Old)
metaperl_ no, I havent
PerlJam Juerd_: me either, where is it?
PerlJam always downvotes merlyn's nodes if all they contain is "see my article on ..." 19:02
Juerd_ I'll try to find it
metaperl_ that's not cool: referring to those articles is a good thing
Juerd_ PerlJam: I don't. I read the article and vote as if the article was the post
metaperl_ they are well-written and cover the topic
PerlJam I can see tchrist's point that merlyn is all about self-aggrandizement
Juerd_ PerlJam: Usually down, because I don't like his programming style :)
PerlJam: I refer people to my own documents too 19:03
ihb PerlJam: "aggrandizement"?
Juerd_ It's to avoid having to write the same thing over and over
While still helping people
PerlJam Juerd_: sure. Perhaps I've absorbed tchrist's (and other's) bias against merlyn always referring to his own work rather than the work of others 19:06
malaire data: do you still want to know how to coerce a `Maybe (Val, [VInt])' into a `Val'? 19:13
err,, dada: do you still want to know how to coerce a `Maybe (Val, [VInt])' into a `Val'?
dada yes! yes!
malaire myCoerce (Just (val, _)) = val -- seems to be working 19:14
dada mmm
malaire perlbot: nopaste
perlbot Paste your code here and #perl will be able to view it: sial.org/pbot/perl
pasteling "malaire" at 217.119.39.217 pasted "coercing a value..." (3 lines, 133B) at sial.org/pbot/8048 19:15
"dada" at 193.203.230.22 pasted "Implementation of unary * from Eval.hs" (11 lines, 567B) at sial.org/pbot/8049
Juerd_ perlmonks.org/?node_id=437958 19:16
Sorry, got distracted
Posts like that one so terribly get on my nerves
dada malaire: how does that fit in my snippet?
Juerd_ continues the search for the writeup he was looking for
malaire I'm checking that... (I'm still quite new on Haskell)
ihb the formatting on that paste page doesn't go well with haskell :-)
malaire where in that snippet do you need it? 19:17
dada malaire: at the end of line 374
doSlice [] vals 0 is returning that Maybe stuff
doSlice :: [Exp] -> [Val] -> [VInt] -> Maybe (Val, [VInt]) 19:18
pasteling "malaire" at 217.119.39.217 pasted "coercing a value..." (3 lines, 100B) at sial.org/pbot/8050 19:19
dada looks fine
malaire does that case-code help? You need two cases - whether is has real value or 'Nothing'
Juerd_ metaperl, perljam: perlmonks.org/?node_id=61933
metaperl_ LOL 19:20
I was going to do something similar with Template Toolkit syntax
[% INCLUDE std_disclaimer.h %] 19:21
dada ouch
src/Eval.hs:375: parse error on input `='
metaperl_ [% IF stonehenge.article_on(subject) %]
see my article
Juerd_ metaperl_: We know the syntax, really
metaperl_ [% ELSE %] [% INCLUDE snide_insult %]
malaire dada: sorry, that should be -> instead of = 19:22
dada malaire: right 19:24
I'm having a damn hard time making GHC happy about my code 19:40
malaire can you paste it to sial.org/pbot/perl ? 19:42
dada sure
theorbtwo Stick a 6 on the end of that. 19:43
pasteling "dada" at 193.203.230.22 pasted "Eval.hs snippet" (12 lines, 604B) at sial.org/pbot/8051
dada theorbtwo: a 6?
theorbtwo sial.org/pbot/perl6 19:44
dada theorbtwo: ah, ok
now I get: Couldn't match `ContT Val (ReaderT Env IO) t' against `Val' 19:45
malaire I need to quit now... Can't really help there.. 19:49
dada oh
I have no hope :-)
Juerd_ Sex is fun, 19:50
but it probably doesn't solve all your problems.
-- lwall
Latest quote :)
Odin- ... uh.
Okay. That doesn't even need context. :)
Juerd_ Do you mean that sex needs no context? 19:51
'cause I disagree :)
Odin- Juerd_: The quote. I wouldn't have a clue on the other matter. :p 20:02
Juerd_ Hello Limbic_Region :) 20:38
LTjake pugs ascii logo: www.inkdroid.org:8888/391 -- warning, i've already been told it sucks =) 20:39
hooray for "art". 20:40
Limbic_Region suspects he has been caught reading slides 20:42
salutations Juerd_
Juerd_ Limbic_Region: slides? 20:44
Limbic_Region: I said hi because of your remark in the PM CB
LTjake: It does 20:45
jdv79 Juerd_, where's the next version of PLP
Juerd_ jdv79: I won't be working on PLP until my RSI is mostly gone, which can take years. 20:47
I have better ways to spend the little time I have 20:48
The next version of PLP will either have user contributed changes or a complete rewrite.
Patches are welcome, especially a (backwards compatible!!) patch to make it work with apache 2
theorbtwo Use one of your daughter's odd input devices?
ihb Juerd_: i don't remember, what was PLP about? Or: why did you create it? 20:49
theorbtwo ...or would that just cause you to get RSIs in other parts of your body?
jdv79 RSI?
Limbic_Region oh - which comment - I am kind of all over the place ATM Juerd 20:50
theorbtwo Repetitive Stress Injury.
jdv79 PLP is like PHP
Juerd_ Limbic_Region: About you going to catch up on #perl6 :) 20:51
Limbic_Region oh - well, I found lots of links here that I was following
Juerd_ ihb: I created it because I needed something template-language-like (PHP-like, if you want) that was both very fast and very easy to use (no weird compilation with subs causing all kinds of problems with lexicals and other subs) 20:52
s/to use/to start using/
ihb Juerd_: which other templates-language-like things existed at invention-time? 20:53
Juerd_ ihb: embperl, mason, eperl, iperl, s///e
ihb Juerd_: in which way/under which circumstances is PLP preferable? 20:54
Juerd_ The first incarnation of PLP just used s/<:(.*?):>/'; $1; print '/g; eval "print '$_'", only escaping some more than this :)
ihb: I have no idea.
ihb: I use it for sites I build
ihb Juerd_: :-)
Juerd_ ihb: It's in use by the dutch ISP "Internet Online" for their administration of rackspace, switches, routers, servers and bandwidth 20:55
But that's just because a good friend works there and wrote this system
I will maintain PLP, because it's used in many important production environments
But don't expect too much 20:56
If you're writing something big, use Template or Mason or Embperl
If you need the speed, PLP's great, though.
theorbtwo Juerd, consider finding somebody else to pawn it off on^W^W^W^Wmaintain it.
Juerd_ Cheap computers have become much faster since PLP was made though
theorbtwo: I don't like that.
ihb Juerd: thing is i don't write big things, which is why i'm interested in other solutions. 20:57
Juerd_ I want to be in control, because everyone who has some opinion about PLP wants to introduce backwards incompatible changes, and someone without an opinion is by definition a bad maintainer.
theorbtwo Mm, point. 20:58
If they can find a way to make backwards-compatable changes, then that'd be good. 20:59
But loosing back-incompat is a big -- unless you cleanly fork.
Juerd_ They can't because PLP needs big changes
ihb Juerd_: would you disrecommend someone starting fresh from using PLP? 21:00
Juerd_ ihb: Only if you want mod_perl 2 support or PLP to do something it currently does not.
Feature requests are left unanswered for the time being.
I use PLP all the time
So how could I recommend against using it?
It's PHP-ish in many ways though 21:01
And that requires some tolerance for ugliness
And impurity
OTOH, it does use perl and it is fast.
Which are important for me.
dada I go, goodnight 21:48
mugwump hmm, can't commit to svn.openfoundry.org even after completing registration ... will try again later 21:52
obra :/ 22:03
Steve_p autrijus, are you using the ErrorMonad to implement try...catch?
22:35 buu is now known as perldoc, perldoc is now known as buu
obra schwern? 22:44
Coke_ what abuot schwern? 23:11
obra was looking for him 23:13
Coke_ I miss purl. 23:15
theorbtwo perlbot, tell Coke_ about the meaning of life. 23:16
perlbot the meaning of life isn't something I know about, theorbtwo
theorbtwo goes to bed. 23:17
metaperl_ is Sam Vilain here? 23:57
obra 18:57 -!- idle : 0 days 2 hours 5 mins 48 secs [signon: Mon Mar 7 23:58

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!