Run Perl6 Now! | pugscode.org ('Overview', 'Journal') | pugs.kwiki.org | channel log: xrl.us/e98m
Set by autrijus on 6 March 2005.
crysflame hi ingy 00:09
ingy hi crysflame 00:55
mugwump can any functions in pugs return lvalues? 01:41
autrijus technically yes, practically "is rw" trait is not yet parsed. 01:55
mugwump I was just adding a test for the junction example in S09 01:56
autrijus it just need to set the env isLValue bit on.
mugwump when I realised that substr hasn't been implemented yet :)
autrijus sure, feel free to test it
it's a todo_ test thought :)
it's not a misimplemented feature, it's just unimplemented
autrijus reading (google cache of): kw.pm.org/talks/2004-10-24-Perl6talk.ppt 01:58
(Perl 6 is) Ready to ship in { (localtime)[5] + 1902; }.
that slide is quite helpful. 01:59
cls_bsd autrijus++ 02:03
autrijus :))
cls_bsd: if you're looking for src/ job 02:05
I just committed a XXX in Eval.hs line 325
it deals with the fact that
cls_bsd okk
autrijus @a[1] = func(); # scalar context
@a[1,2] = func(); # list context 02:06
currently both passes list context to func()
what needs to happen is to write a "cxtFromExp"
of type:
Exp -> String
and replace ths "List" in line 327 with "cxtFromExp exp"
the way to write the cxtFromExp function is to pattern match 02:07
cxtFromExp (Syn "," _) = "List"
cls_bsd hmm
autrijus cxtFromExp (Var ('$':_)) = "Scalar"
cxtFromExp (Var _) = "List"
etc. feel free to finish the rest :)
cls_bsd go to read haskell doc 02:08
autrijus YAHT may help. some people prefer www.haskell.org/tutorial/ 02:10
cls_bsd great, this tutorial 02:11
autrijus bbiab & 02:30
ingy hola 02:49
mugwump hi ingy 02:51
hcchien ingy: would you mind to send me your flight schedule again. :p 02:53
mugwump ./pugs -e 'say ["c","a"].join("")' 02:54
ac
:-/
ah, it's todo, nevermind 02:55
pjcj svk ci 02:56
oops
stevan_ howdy ingy 02:58
I had a kwid question earlier, but I figured it out :)
mugwump just how *do* you test that two junctions are the same, anyway? :) 03:03
if ($j == $k) # doh! :)
I just did a `svk push', but got ` t/03operator.t - skipped' 03:10
hcchien do you svk pull before? 03:11
t/03*.t doesn't exist anymore. :)
mugwump oh, that'd do it :-}
I was naively using `svk update'
and `svk sync -a' 03:12
hcchien pull may be your friend. :)
up; sy; and you have to do the 'sm'
autrijus just "svk pull" ;) 03:14
mugwump where do you think I should put the junction tests... t/op? 03:17
t/voodoo perhaps 03:18
autrijus heh heh
t/junction/
hcchien if the value is assigned as undef. what should we expect when try to print it? 03:29
autrijus empty string
crysflame is there a function junction 03:34
HA HA
autrijus :D 03:35
hcchien "for my $foo (@bar)" doesn't work? 03:41
autrijus for @bar -> $foo { ... } 03:42
that way reads better
for (@bar) -> $foo { ... } # this works too
for @bar { $^foo ... } # too
hcchien or for ($foo | @bar) ? :p 03:43
autrijus bwahaha
crysflame heh
autrijus [ say $foo | $foo <- @bar, $foo > 3]
crysflame hcchien++
autrijus: is that print grep { $foo > 3 } @bar? 03:44
autrijus @bar.grep:{$_ > 3}.map:{say $_}
crysflame let me rephrase 03:45
what does what you said do? [ say ..
say every $foo that is > 3, using @bar as a source?
autrijus yup. 03:46
except say is invoked many times
crysflame right
autrijus instead of being passed a list
crysflame so map { print } grep { $_ > 3 } @bar in perl5
neatr
hcchien is pugs always 'use strict' now? :) 03:47
autrijus perl6 is always use strict by default.
but "perl6 -e" may relax it.
that is not currently implemented in pugs. tests welcome
reboot, bbiab & 04:02
04:14 root_____ is now known as nnunley
metaperl nnunley are you in England working for fotango or in America? 04:21
mugwump if he's in England he won't be up now ;) 04:22
nnunley metaperl: I'll be working for fotango in the UK come monday. But currently in America.
metaperl why not? they have acid house parties all night :)
nnunley grins.
crysflame fotango :)
autrijus really. :) 04:23
clkao: hope you come back alive :D
nnunley nods.
clkao and I will be in the flesh coworkers in about a month.
autrijus mmm acid-induced software 04:24
nnunley Could be interesting. Perhaps a little hard to debug.
autrijus yeah, you'll keep seeing grid bugs. 04:25
nnunley Heh. Speaking of that... anyone play darwinia, yet?
clkao yawns
nnunley will be heading to bed shortly. "Held a going away salon/scotch tasting..." 04:27
Just checking in to see if anything interesting is happening.
autrijus ah. lots actually. 04:28
#perlchina is pushing pugs to parse utf8 literals.
I'm preparing a "It's a wonderful hacking day" slide, due ~3 hours 04:30
which means I should probably deIRC and Work
metaperl nnunley, you have played darwinia? the graphics look blocky 04:34
autrijus bbiab & 04:35
metaperl i'm d/l'ing the demo now
scw autrijus: in fact, I noticed the problem on VInt on the way home that day, but totally forgot to fix it, sorry 05:07
obra heya david 05:48
mugwump puts some finishing touches on a *working* junction-based sendmoremoney.p6 05:50
06:04 khisanth_ is now known as Khisanth
mugwump sighs at just how long finishing touches can take sometimes 06:40
send+more=money is now junctionally functional! 07:06
crysflame heh 07:08
ingy perlbot: seen Schwern? 07:31
hola amigos 07:32
Khisanth perlbot does not have that capability :/ 07:33
ingy I forgot 07:35
crysflame i do 07:45
ingy ?
crysflame not anywhere in my history have i seen schwern mentioned here
ingy schwern has definitely been on the channel 07:46
crysflame weak history it is then 07:47
Khisanth well there is always seenserv 07:51
autrijus mugwump++ 08:02
going to the talk &
I still had not done my slides! 08:03
clkao have fun
autrijus needs to pull an ingy
ingy hahaha
pull my finger
nothingmuch good morning 10:16
dada hola 10:31
autrijus the talk is about to begin 10:34
wish me luck!
autrijus hopes to lure more pugs developers
rgs sends good waves to autrijus 10:35
larsen in *%$ of the whale! 10:36
malaire Hello, is here anyone who could check my patch about hyperops? Whether it's ok or not? 10:57
(www.nntp.perl.org/group/perl.perl6.compiler/292)
nnunley mugwump: Sweet, re the working send+more=money 14:14
malaire GHC 6.4 has been released today... 15:00
DrHyde maybe it will compile on my box then, cos 6.2 sure as hell doesn't 15:14
rgs bootstrapping is not easy 15:15
can you bootstrap 6.4 with 6.2 ?
ingy hola 15:47
autrijus: how was your talk?
dada is the talk finished yet? :-) 15:50
ingy dunno, I hope so 15:51
I mean a 5 hour talk with no preparation would be... interesting
larsen " 11:39 < autrijus> the talk is about to begin "
and now it's 16.56 here
ah. 5 hours
I didn't know that
ingy ? 15:52
dada oh, maybe he's racing for the World's Longest Talk
ingy I seriously doubt the talk was more than an hour or two
larsen I thought you were talking about some form of extended tutorial 15:54
malaire pugs gets quite many warnings with GHC 6.4, but seems to work OK. 16:00
(same 2 test scripts failing)
nothingmuch could someone please clarify some haskell stuff? 16:08
err, maybe i should go to #haskell instead ;-) 16:09
ingy nothingmuch: yeah, don't interupt us here! We're all trying to sleep! 16:10
nothingmuch beh
ingy ;)
hcchien ingy: did you see my message here? 16:11
ingy hcchien: you want my itinerary?
hcchien ingy: yes. 16:12
ingy hcchien: so you can pick me up from the airport?
nothingmuch wants to be around airports 16:13
hcchien ingy: I don't have car. :/
ingy hcchien: what do you need my info for? 16:14
hcchien ingy: when will arrive and leave.
ingy hcchien: I know, but why do you want to know? :) 16:16
hcchien ingy: good question, I may have to make sure you don't lose. :p 16:17
ingy ? 16:18
anyway I arrive March 22nd at 07:50
and I leave April 5th 16:20
hcchien :) 16:19
ingy hcchien: where do I go, when I arrive?
hcchien ingy: we will pick you up when you arrive Taipei. 16:20
ingy wow! thanks!
obra hi 16:21
Perla somebody's here?
hcchien it's what we should do. :)
ingy hcchien: the evaair info is very unclear, but I assume if I am leaving Seattle at 02:55 on March 22 and landing at 07:50, then it is still March 22nd, right? 16:22
obra: confirm
and if I leave Taipei on April 5 at 18:20 and arrive in Seattle at 14:00 it is then April 6 16:23
hcchien: note I leave Taipei on April 6 at 18:20 (not 16:20) 16:24
hcchien I think it is 3/23 07:50 in Taipei
obra TPE->SEA, you get in before you leave 16:25
ingy hcchien: but I am flying west... hmmm, I cross international date line?
obra SEA->TPE, you lose a day
yes
ingy I see
so hcchien, it looks like I arrive on 07:50 March 23rd :( 16:26
hcchien ingy: I think so. :)
ingy less days to write talk and party 16:27
hcchien: at least you do not bug me for my materials this year :P
obra arrives TPE at 0600 on 3/20 16:28
ingy at YAPC::Taipei all talks are written JIT
obra gets *3* days on me, no fair
hcchien ingy: yes. so I won't have processings this year. :p
ingy this is a conspiracy!
obra kept asking for my flight dates 16:29
and then moved his earlier
Taipei will be all used up by the time I get there
obra we'll take pictures
ingy :P 16:30
maybe I'll change my flight to the 15th
muhahaha
hcchien ccCc
ingy *_*
obra do it
get taipei ready for us 16:31
ingy gives obra her ICY STARE OF DEATH
obra now you sound like boojum, ingy
ingy I learned a thing or two from booj
er, b**jum 16:32
whiteg ingy: are you sure? 16:33
obra ow. my travel to TPE is going to be 24 hours, starting at 6pm 16:34
crysflame .. 16:59
aha
ingy hi whiteg 17:39
hcchien: ping 17:40
hcchien ingy: yes?
ingy do we have an official wiki for yapc yet? 17:41
hcchien ingy: not yet.
ingy I can set up yapc.taipei.kwiki.org 17:42
hcchien oh, cool.
ingy++
ingy hcchien: do you have some grapics or a static site that I can theme the kwiki with? 17:43
I can just use the stuff autrijus sent me
unless you have something nicer 17:44
obra bummer
hi luqui 17:47
autrijus hi :) 17:49
hcchien ingy: how is the official site of yapc 17:52
ingy hcchien: url? 17:53
malaire autrijus: Could you check my hyper-ops patch, whether it's ok or not? 17:54
luqui hi ingy
malaire (www.nntp.perl.org/group/perl.perl6.compiler/292)
autrijus !!
lots of things happened
# autrijus.org/hackday/
slides for today's talk
just had a brief electricity outage here...
finally got some stable network now
interesting approach. 17:55
I think it works good as first-try 17:56
eventually we'd like to parse it as part of op2 parser
^Hbut I think it can go in.
please commit
wolverian autrijus: what is the encoding on hackday/? I can't get it to show properly.
autrijus utf-8.
it's also all in chinese.
s/all/almost/ 17:57
wolverian ah. still shows as ?. I guess my font doesn't have chinese characters.
autrijus makes sense.
malaire I have no commit-rights yet
autrijus really.
autrijus checks
gugod: you want to commit your test too? I'll make you a committer
greetings Ovid-san
Ovid Good morning, autrijus :)
ingy hi gugod 17:58
Ovid Hi ingy.
ingy hi Ovid
justatheory autrijus: No distracting ovid, he needs to be working right now!
autrijus what work needs him now
?
malaire: invitation sent. please reply "thanks applied" to yourself :)
justatheory The training I'm about to give him.
autrijus err
s/invitation/committer bit added/
Ovid I pretend to work, Theory pretends to pay me. We're all quite left-wing over here.
justatheory autrijus: Oh, have you implemented that op yet? 17:59
autrijus what op?
ingy oh no, Ovid is getting sucked into bricolage
justatheory err
blown, more like
Ovid What an apt description.
ingy grabs Ovid's leg
Ovid That's not my leg.
justatheory grabs ovid's hair
ingy you're right!
justatheory All right, all right! That's enough!
ingy but it's so big 18:00
autrijus gugod: welcome aboard, too.
justatheory gugod++ # My Unicode tester!
justatheory & # Work
ingy gugod: I am coming up with ideas to use all of your time during YAPC
gugod hello theory
hello all
obra hi 18:01
gugod ingy: .....
autrijus ingy: surprisingly I am coming up with ideas to use all of your time too
malaire ok, I did my first commit, v588 ...
autrijus malaire++
I'll base generic hyperop off that.
luqui ack
ingy autrijus: I'm creating the yapc kwiki to start writing down all these plans
autrijus sure. 18:02
luqui needs to propose a talk for yapc
rgs which yapc ?
luqui na
luqui can't afford international flights 18:03
ingy luqui: think of it as an investment
luqui no, as in, I actually don't have that much moeny
not, I don't have money to spend on that. I just don't have it.
obra luqui: some international conferences fly interesting people in
rgs apparently my employer agreed to pay for me to go to oscon.
luqui rgs: that's always nice 18:04
rgs luqui: can you afford oscon ?
luqui not if I don't get my talk accepted
rgs right
luqui I'm counting on Damian to vouch for me (he's on the selection committee)
hcchien ingy: yapc.elixus.org/
ingy hcchien: yapc.kwiki.org/taipei/ 18:05
rgs 'pparently my talk is mostly accepted
obra luqui: is it the conference fee for OSCon or the other expenses?
luqui obra: hmm?
obra that makes it too expensive
luqui mostly the conference fee
ingy hcchien: is Dan Kogai comming? 18:08
hcchien ingy: sadly, no. :/
obra yapc.tw people, join #yapc.tw 18:10
pjcj d 18:18
luqui just submitted his talks 18:24
tiw hello, i am using pugs. I have a question, whether pair is implemented in pugs. 18:25
theorbtwo pair?
tiw $a = :key('value')
appear err
theorbtwo Ah.
crysflame it might be in a test somewhere 18:26
theorbtwo Well, from a quick check, no, at least not in that format.
tiw and in which format?
crysflame if not, if you write up a test, autrijus will implement it for you (usually within a few minutes of seeing the test)
dada $a = (key => 'value');
tiw $a = 'key'=>'value' is not ok too 18:27
theorbtwo Hmm? my $a = 'key' => 'value'; say $a.
Works for me.
dada for me too 18:28
autrijus that is the only form of pair syntax that works now.
tiw Undefined variable $pair
Val VUndef
autrijus I think we've got tests for the rest of the forms
tiw: oh. strict is on by default.
so, "my $pair"
tiw also 18:29
i got it
thanks lot
theorbtwo Hm, looks like the time prim is defined wrong -- it should be 0 at Jan 1 2000 (and floating point)... unless Larry changed his mind.
autrijus it is indeed wrong. 18:30
fix it? :)
tiw yes, it work now.
and how can turn off strict
autrijus write a test
so I can write ext/strict/ 18:31
theorbtwo Hm, I'm somewhat worried that I can't think of a way to test it. 18:32
autrijus "no strict;"
luqui time 0 ?
oh that test
eval 'no strict; $x = 10;' 18:33
in a list of patterns with |, how do you specify a default?
theorbtwo I meant testing time().
luqui in haskell?
theorbtwo | otherwise = whatever
luqui theorbtwo: yeah, time(0) should do it
(if I remember how time works properly...) 18:34
(which I'm sure I don't)
nevermind
autrijus luqui: if you are fixing hyperop base case, I just committed
luqui oh, okay, I'll see how you did it 18:35
I kept getting type errors
I don't think that's the proper behavior
tiw I can not use no strict, with following :No compatible subroutine found: &no
autrijus tiw: right, so that doubles as a test for 'no'. 18:36
luqui I think hyper on two scalars just does the operation once
that makes writing things recursively really nice
autrijus luqui: maybe you can help haxx0ring Lexer line 383ish 18:37
and make >> and << work on all binary OPs
so it can work without us munging the Prim decls
if you don't feel like doing it I can do that tomorrow
luqui I can focus on that when I get home 18:38
that sounds like a fun problem
autrijus cool. it's mildly complex, yeah
dada autrijus: do you want me to commit the unary * as I've coded it? 18:39
autrijus dada: as you nopasted the other day?
dada yes
autrijus along with Syn * etc?
and tests?
if so, sure!
dada yes
ok, I do
luqui how do you run a single test anymore? 18:40
no module 'Test.pm'
malaire ./pugs t/op/hyper.t works for me
dada set $PERL6LIB
autrijus luqui: make install
luqui ohh... have to install it
autrijus or, set PERL6LIB to ext/Test/lib
and HARNESS_PERL to ./pugs
then "prove".
or just set PERL6LIB and ./pugs
take yer pick :) 18:41
ok. I think 6.0.11 on Sunday or Monday.
that gives me a weekend to tie loose ends up.
nothingmuch morning
autrijus greeintgs nothingmuch.
nothingmuch hulloh
nothingmuch is sorry for not living up to promises 18:42
crysflame autrijus: pugs isn't mentioned at en.wikipedia.org/wiki/Perl
nothingmuch been an annoying half-week so far
crysflame and it mentions Perl6 as a parrot project
autrijus crysflame: you can fix that
crysflame but i won't be. just figured i'd mention it.
crysflame is researching re contextedness for work
autrijus nothingmuch: aw :/ 18:43
theorbtwo Woo, my first commit: r592, fixes time(). 18:44
nothingmuch theorbtwo++
luqui & class
nothingmuch wonders how you test that
autrijus theorbtwo++ 18:45
nothingmuch: time(0)
theorbtwo I just asked pugs for time, asked perl for time, subtracted, asked units to convert to years, saw 30, and liked it.
autrijus pugs> time
crysflame heh
autrijus 163881920
perl -e 'time' # 1110566723
crysflame that doesn't look right
autrijus something is wrong.
(win32)
theorbtwo That's right.
Number of seconds since jan 1, 2000. 18:46
autrijus 163881920 is right!?
theorbtwo Not 1970.
Wait, what?
nothingmuch why did that have to change?
theorbtwo No, that's not right.
nothingmuch oh god
i want Time::Piece
autrijus jan 1 2000!?
nothingmuch 6.2! get me 6.2!
autrijus where is that mentioned?
where is it said that p6 counts with y2k as epoch? 18:47
ingy oh my god! there is another ingy, and she is also from Seattle!
(google hit #2)
autrijus ingy: yeah, I noticed the other day too
guess you're really her
and I've been meeting someone else
who knew you well to fake an identity as ingy 18:48
ingy oHH NO!
Coke huh. for some reason I though "ingy" was a guy's name.
theorbtwo searches.
ingy Coke: there are *2* girls named ingy in Seattle that disprove your assumption 18:49
obra ingy == twiggy?
Coke It just sounded masculine. =-) My bad. =-)
ingy oh crap... another ingy www.geekhive.com/ 18:50
autrijus oh wow. y2k is for real.
Coke vaguely wishes english had better gender discrimination built in, though the PC folks would hate that, he's sure. =-)
malaire where is y2k mentioned?
ingy I need a new name!
autrijus theorbtwo: is it mentioned in apoc/syno?
nothingmuch was really offended when nothingmuch was already taken on livejournal 18:51
a friend has been pestering me for about 6 months to sign up
theorbtwo I'm searching.
nothingmuch so finally i was like 'fine fine fine'
turns out a month before that she signed up
autrijus malaire: 0rz.net/5c0dN
we live in interesting time()s. 18:52
nothingmuch must remember to try and ruin her google ranking some more
theorbtwo Can't find it. 18:53
autrijus not in apocs it seems.
more importantly, not in Syns.
theorbtwo Not in any of the three, that I found. 18:54
autrijus neither did I.
theorbtwo: can you post to p6c?
announce that the time() has been changed
theorbtwo c?
autrijus citing the url above as reference
and ask for codification
yeah, it's c I think.
I'm still living by the "always start with p6c" rule of thumb 18:55
nothingmuch ooh, googling for myself: forums.photographyreview.com/member.php?u=239051
autrijus and larry seems quite serious on that 2004-03-03 post...
nothingmuch damn, i love it when i'm artsy in the bio stuff 18:56
autrijus theorbtwo++ # actually remembered and implemented the y2k ruling. wonderful. 18:57
crysflame the y2k.. . ruling? 18:58
autrijus www.nntp.perl.org/group/perl.perl6....nals/21717 # better reference url than 0rz above
crysflame: see that url
crysflame click
nothingmuch wonders why not 0
autrijus C:\work\pugs>pugs -e "time().say"
crysflame oh, nice
autrijus 163882705
nothingmuch but 2000
obra cries
nothingmuch if at all
obra Y2K IS NOT THE EPOCH
Coke obra; prove it. =-) 18:59
crysflame it is now
autrijus it is the pugs epoch now...
...unless you can convince larry the other way.
nothingmuch and why not 1970 to retain compat to data stored in many places
Coke is tempted to implement dan's epoch for parrot from that post.
obra What was nov 7 1858?
Coke that was for /parrot/ 19:00
obra er 17
sure
Coke and dan was being DanCranky.
obra :)
autrijus obra: that's a VAXism.
Coke ah,I misread your send. Yes, it was 11/17/1858. =-) 19:01
autrijus "This base time of Nov. 17, 1858 has since been used by TOPS-10, TOPS-20, and VAX ... "
crysflame you can always just set the epoch by wrapping time..
time.epoch(VAX)
autrijus and we all know where Dan's root was.
crysflame time.epoch(MAC)
Coke ... "ew" =-)
crysflame time.epoch(UNIX)
dada use Epoch 'Neanderthal';
autrijus time.epoch('Discordian')
crysflame right! 19:02
malaire time.epoch(UNIVERSE)
nothingmuch discordian i can handle
obra wants to know what caused that choice originally
autrijus obra: ask!
nothingmuch maybe we should find out if larry is not a cabbage or something
autrijus I think larry is at least 15,352,192 distinct different cabbages.
nothingmuch arbitrary values setting 0 are silly. they are --><-- with logic
obra autrijus: I'm asking the intarweb
nothingmuch Data::Dimensions is something i like 19:03
obra nothingmuch: so you want the discordian epoch?
nothingmuch because it is very carefully designed to not do that stuff
obra --><-- is the five fingered hand of eris.
theorbtwo Once we can, time(:epoc=>UNIX) should work. But we can't yet.
nothingmuch orbra: out of all the possible solutions it is the one i'll have the lease problems remembering =)
theorbratwo. beh
tiw i example of pugs, there have #!perl6, is perl6 a link of pugs?
nothingmuch should use tabcompletion and believe it when it doesn't
autrijus theorbtwo: the first goal is to get it codified :-/ 19:04
tiw: no it is not. not until pugs is rewritten in perl6.
and even after that, not until larry blesses it as the real perl6.
crysflame that's the first time i've heard that come up
Khisanth wonders if anyone is working on a vim/emacs "mode" for perl6 19:05
tiw then why write it in example?
autrijus tiw: because that's the suggested style in the Apocalypses and Synopses.
obra perlbot: november 17, 1858 is www.inwap.com/pdp10/usenet/julian-day
perlbot added november 17, 1858 to the database
theorbtwo OK, sent.
autrijus tiw: it also means that the examples can be shared between different perl6 implementations.
theorbtwo++
obra that makes some sense now
tiw i see
autrijus I think we took that tradition from seano's original P6C.pm. 19:06
theorbtwo Oh: Does openfoundry use your i10n framework, autrijus?
autrijus l10n? yes. 19:07
which is just rt's.
theorbtwo was wondering about a fallback that used babelfish.
rgs lsc
obra theorbtwo: we've talked about it
theorbtwo: it's against their aup
theorbtwo Pitty. 19:08
obra autrijus: what's the state of foundry-3.4?
autrijus obra: the UI part is pending more meetings
obra once that happens, i think that'll be the fast way to get cpan-3.4
autrijus but I think pretty strongly that tabbedui and classicui and foundryui should merge.
obra yay 19:09
it's time
autrijus it is. and cpan-3.4 can then base on that.
obra hell, it can just be foundry
autrijus we should code for real toward that when you're here
sure thing.
ingy has good ideas about per-user svn space and kwiki and blog etc space
obra yay
autrijus which I am very interested in
so we should also do that. 19:10
obra -> yapc.tw
ingy yes!
autrijus :) 19:12
tiw is type system in pugs not complete, or? 19:15
theorbtwo Do we not do prefix ++ and --?
I don't see them in op1...
autrijus it's just ++ and -- 19:16
tiw: the type system in pugs is very confusing.
however, I don't think perl6's type system is less confusing. :)
theorbtwo Ah.
ingy heh
I spent hours reading about Haskell's type system last night
tiw confusing? what is mean 19:17
autrijus tiw: i.e. we have a type system, but it is only used as context hints
i.e. sub foo (Str $x) { ... }
you can pass a Int into foo()
it will just stringify it.
dada ingy: I gave up at page 39 of Gentle
ingy dada: I'm half thru chapter 7 19:18
tiw i can not explicit define a Int variable now?
autrijus tiw: no, there is a test for that.
that is todo.
ingy ie I survived reading about types, without committing suicide
Haskell has an entire OO system just for types! 19:19
a fullfeatured OO system
tiw oh, i should see the test.
ingy: you don't like types 19:21
ingy tiw: sure I like types, I just wasn't prepared for Haskell's types! 19:22
theorbtwo Is there a purticular reason that readline is defined to be =, rather then vice-versa? 19:23
dada uhm 19:29
ghc complains about not finding 'System.Posix.Types'
autrijus version and platforms?
dada 6.2.2 win32 19:30
fresh installed
autrijus ingy: there is also fundeps, existentials, rank-n, and GADTs
.msi?
dada yes
autrijus C:\ghc\ghc-6.2.2\imports\System\Posix
Types.hi
is it not there? it is there for me.
ingy autrijus: great :| 19:31
autrijus ingy: these are not haskell 98 and pugs makes no use of them (yet)
so you don't need to do seppuku just yet
dada fixed
autrijus "fixed"?
theorbtwo Please, don't get guts all over the chan.
dada just needed to close the console and open a new one :-)
autrijus ah. 19:32
dada it messes with %ENV apparently
autrijus weird.
oh. the .msi
that's too fresh :)
dada yes, the console was opened *before* installing GHC :-)
autrijus figured :) 19:33
tiw ~= is also a todo feature?
autrijus yeah, the >><< modifier and the post = modifiers are todo.
~= etc are trivial to implement 19:34
just look at --
not sure if they are tested
dada why doesn't make test pick up my testfile? 19:35
autrijus perl Makefile.PL 19:36
will refresh the list of test.s
theorbtwo: you posted to p6c?
theorbtwo Yeah. Not sure why it hasn't shown up; I should be whitelisted.
autrijus hmm, maybe needs some time to propagate
theorbtwo I posted via the NNTP.
autrijus not sure nntp works.
but ok :)
dada autrijus: yes, sure 19:37
ingy gugod: ping
autrijus 3:37am here. guess gugod's asleep.
I need to sleep soonish too.
ingy gugod never sleeps 19:38
gugod: join #yapc.tw when you wake up
nothingmuch anyone care to help me with YAHT excercise 4.7?
i've been delaying it all day
autrijus malaire: mm, the >> form is not parsed, only french form? 19:39
malaire yes, I only added french form
theorbtwo Which one is 4.7?
nothingmuch #haskell people are not listening to me
Tuples + Either
what does Left x construct?
a value that can be used in place of Either? 19:40
which means...?
dada I'm about to commit
nothingmuch fromTuple (One a ) = Left (Left a )
malaire Left x constructs a value of type Either, as does Right x
autrijus Left :: a -> Either a b 19:41
nothingmuch what i'm looking for is a practical example of where Either makes sense
autrijus i.e. for whatever b, it can make a "Either a b" from a
nothingmuch: oh. sure. Bind.hs :)
nothingmuch in the hopes that i can generalize form that
autrijus Either lets you returns a string for error
nothingmuch autrijus: be reasonable please
autrijus and a whatever type you want for normal "correct" answer 19:42
"Either String Val" means
theorbtwo wonders how many people have filled in affiliation~~m:i/pugs/ in the YAHT download form... and how many of them are me.
autrijus "something that can fail with a String, or succeed with a Val"
nothingmuch so the return of function bzzt, being either means 'either bzzt was happy, or it was sad'
malaire Example 4.7 is *not* practical example... Once you learn about Monads, there are many examples where Either is used.
dada how do I set the username with svn?
autrijus svn ci --username maybe? 19:43
nothingmuch ok, what does fromTouple mean, really?
dada autrijus: exactly that
nothingmuch either it's a value, or it's a pair, or it's a triple, or a quadruple?
given any one of the Touple variations? 19:44
ingy dada: svn help checkin
er
nothingmuch how does that become, err, logical?
ingy dada: svn help commit
rather
dada it worked! committed! r593! hooray!
ingy dada++
dada ingy: thanks :-)
ingy np 19:45
Limbic_Region perlbot karma for autrijus
perlbot Karma for autrijus: 6
theorbtwo You can write multiple versions of the same function that take different arguments.
Limbic_Region seems low
autrijus
nothingmuch rephrases
malaire notrhingmuch: It's jsut a example how a single function has return 4 different types of data (i.e. you can't do that, unless different types are put within some other type, e.g. Either)
theorbtwo So tupple1 Tupple a can do something different then tupple1 Tupple a b.
Limbic_Region arggh - but this client eats plus symbols so I can't karma anybody
nothingmuch how do i use the return value? 19:46
ok
dada actually, t/op/splat.t contains 1 test and 2 todo's (and 4 more commented out because they don't even parse)
nothingmuch i think i get it
theorbtwo autrijus++ # proxied
nothingmuch then i pass it on to other polymorphic functions
and it just works?
theorbtwo Then again, I'm not real sure my solution was the best one.
nothingmuch data if you can get something that doesn't parse to todo_ok in an eval '', then please do 19:47
theorbtwo BTW, appendix d has solutions.
nothingmuch i looked at the solution, and didn't understand it
autrijus is this the "tup1" exercise?
nothingmuch uhuh
i got passed Maybe
that makes sense, sort of
autrijus oh. it's not the tup1 yet. I have a favourite solution for tup1 :) 19:48
probably won't help you, though :) 19:49
# wagner.elixus.org/~autrijus/tup1.hs
theorbtwo Oh, my solution does match the sample one pretty well. That's pretty sad, as I thought my solution was an ugly hack.
nothingmuch grudgingly continues hoping to feel he got it later on 19:50
tiw is match a regex not implemented? and is there anywhere a list, in which i can find what is implemented and what not, i do not want to waste your time on such trivial question
nothingmuch tiw: this list is a work in progress
tiw: to be quick, grep -r todo_ t/ 19:51
autrijus tiw: you can help filling in annotations for t/Synopsis/*.pod
nothingmuch tiw: and check if there's a corresponding file for the feature you want
autrijus it's a task that can be done piecemeal
dada nothingmuch: I have another exercise for you: learn to spell my nick :-)
autrijus ask nothingmuch and stevan for the appropriate S<> syntax.
I'd be happy to walk you through, but I need to sleep now :) 19:52
nothingmuch da<tab>, i think
autrijus waves &
dada nothingmuch: no, "data" is not here :-)
nothingmuch that's unambiguous at the moment
dada bye autrijus
nothingmuch ciao, autark-jp
damnit
autrijus:
dada [20:47] <nothingmuch> data if you can get something that doesn't parse to todo_ok in an eval '', then please do
nothingmuch is having a very very bad day 19:53
all(users).any(forgive, ignore)
dada pats on nothingmuch's shoulder
malaire nothingmuch: Once you get to YAHT chapter 9, you might want to read "All About Monads" (www.nomaware.com/monads/html/) at that point. -- At least I found that a lot easier introduction to Monads. But it requires you to understand basic stuff first. 19:54
nothingmuch will try to remember
malaire Goes to sleep... 19:56
nothingmuch ciao malaire
rgs Ryusei-Kacho ?
dada speaking of Haskell, I didn't understand what a "constructor" is 19:57
nothingmuch as I see it a constructor is a way to take some things
dada but
nothingmuch and combine them in an undefined way (does the compiler control this?) 19:58
into a tyupe
dada data Point a = Pt a a
nothingmuch which other functions have a standard way of taking apart
dada "Point" is the type here
and what is "Pt"?
nothingmuch Pt is the name of the constructor
dada and is the body of the constructor defined elsewhere? 19:59
nothingmuch a a is the body
it takes a a
and converts it into Point
dada mmm
nothingmuch somehow
dada but aren't names beginning with a capital reserved for types?
nothingmuch well, it's not really a function 20:00
since it doesn't do things you control
or rather, it's not really user defined
i don't know if it isn't a function
dada the Gentle tutorial isn't very clear about this point
rgs autrijus: nice slides. 20:01
nnunley data: It's a signature, not a function. 20:03
So you know the shape, but not the construction.
dada argh! 20:04
nnunley You would need to write a function that in turn builds the data for you.
dada why is everybody calling me "data" today? :-)
nnunley dada: Good question. Muscle memory.
Or you could see it as a dadaist experiment. ;)
dada even :-) 20:05
so, "data Point a = Pt a a" is not enough to build a point?
nnunley dada: You'd have a simple function that did someting like: makePoint x y = Pt x y 20:07
dada: Note, the type isn't defined yet for a -- it could be just about anything, as long it as it was the same for both slots. 20:08
dada yes, this is clear 20:09
but!
Prelude> data Point a = Pt a a
<interactive>:1: parse error on input `data'
should I use "dada" instead? ;-)
nnunley Put it in a file.
dada oh, ok
nnunley The interpreters are a little dull when it comes to defining new functions and types.
tiw nothingmuch: i have checked matching regexp in t/, but find nothing about it 20:10
nothingmuch tiw: regexp are not implemented
tiw and also in todo 20:11
not in todo?
dada well, I still don't understand the need for a "Pt" there
but maybe I will...
nothingmuch i was supposed to write that, tiw, but i feel like crap and can't concentrate 20:12
nnunley dada: I think it's just to give you a place to hang things. If you had other options for what a Point could be, say... data Point a = Pt a a | Pt3d a a a | Pt1d a 20:14
dada mmm
this makes sense 20:15
nnunley data: I suppose it makes more sense once you start playing with the codebase.
That is, once you start playing with a real code base. 20:16
Blah. Did it again. 20:18
dada I'm getting used to it :-) 20:23
I go, goodnight 20:34
theorbtwo Resending time() post via email. 20:37
Is it just me, or is the hyperoperator bit in Prim.hs completely wrong? 20:41
The first codepoint given is capital a circumflex, not <<. 20:42
Morever, it gives <<op>> instead of >>op<< unless I'm quite mistaken.
Which is quite possible, I should note.
tiw bye, goodnight 20:43
theorbtwo G'night, tiw.
t/op/hyper.t would be the right place, right? 20:49
Oh, hmm... 20:50
nothingmuch needs alcohol and vast quantities of it, at that 20:51
sorry for not getting to rules yet, everyone
nothingmuch--
adios
theorbtwo Allo, Schwern. 21:42
Schwern Hi 21:46
theorbtwo Hm, do all binary operators have an assignment form?
Er, binary infix.
Oh, wait, they're already there, why aren't they parsing... 21:48
Hm, this is sticker then I thought. 21:49
The problem with $a ||= 3; is that it thinks it's $a || =3; 21:50
That is, that the = readline operator is getting applied to the constant 3, and that's getting logically ored with $a. 21:51
(And all of this in void context.)
nnunley Schwern: Hey. 22:32
Schwern Hey 22:49
Limbic_Region 17:35, nnunley serves to Schwern 22:50
17:51, Schwern returns serve 22:51
this is the longest game of pong I have ever seen
Schwern Sorry, Pugs is slow 22:58
Juerd_ hehe
I've played pong via e-mail once
Once, because a 15 point game takes way too long
theorbtwo And playing it by email speeds it up? 22:59
Juerd_ No
Not quite
Real time games are hard to play turn based
theorbtwo Of course. 23:00
Juerd_ It worked though
Schwern Let's play tetris on paper.
theorbtwo (This being why D&D gets a big -- in my book vs LARP.)
Juerd_ Schwern: I've done that!
In school
During maths
Schwern I was going to say, you must have been reaaaaaally bored.
Juerd_ Of course 23:01
theorbtwo Trying to create an optimal algo, or what?
Juerd_ No
Trying to out-score my opponent
We used all blocks except the square one
A simple D6 determined the block 23:02
If I had to choose between paper tetris and e-mail pong, it'd be paper tetris 23:03
theorbtwo /metoo. 23:04
Juerd_ we mostly played hangman though
but not with easy words
theorbtwo antidisestablishmentarisim? 23:05
I know, in Dutch.
Juerd_ Like that, yes
theorbtwo I wonder -- does Dutch tend to form reallylongcompoundwords like German? 23:06
Juerd_ Our chem teacher once tried to play hangman with formulas
lumi It does
Like telefoonbeantwoordaparaatvrees
Juerd_ Like H2O -> 2H2 + O2
lumi (Fear of ansaphones)
Juerd_ lumi: That's voicemailfobie in modern Dutch.
:)
theorbtwo: Yes. 23:07
lumi Really? My favourite word, demolished!
Juerd_ theorbtwo: rioolwaterzuiveringsinstallatie
theorbtwo Something to do with plumbing? 23:08
Juerd_ Somewhat
I don't even know the English word for riool
theorbtwo I know no Dutch. Hell, I barely know any German, and I've lived in Munich for over a year and a half.
Babelfish suggests sewer. 23:09
Juerd_ You could never learn Dutch either - you'll just notice everyone starts speaking English to you...
sewer water cleaning installation, then
You misspelled renaissance in your yahoo profile, I think 23:11
theorbtwo Me?
I don't really care.
Juerd_ Or someone invading your name space 23:13
(It's hard to type that without [])
webmind print $_
Juerd_ webmind: needlessly verbose
webmind: Just print
webmind uhm
yeah
sorry
Juerd_ :)
theorbtwo Fixed anyway.
lumi btw theorbtwo, how's the subetha doing?
theorbtwo Not too badly, but will be a while before it sees light at this pace. 23:14
lumi Need help? I've my mac now
Juerd_ lumi: with french fries?
Juerd_ is still waiting for the mac menu mini 23:15
lumi Um yeh, one of those
theorbtwo Not really. Ended up putting pearpc on another box as well.
Juerd_ I've been considering wanting a mac mini.
(Literally)
I still don't know if I want one
theorbtwo Strangely, networking worked much more easily with that one.
Juerd_ I want to want one though
theorbtwo I still don't know if I want an OSX box at all.
webmind Juerd_, why would you want one |? 23:16
Juerd_ theorbtwo: That's my problem too
webmind ?
Juerd_ webmind: Because it's a very neat and silent tiny box
theorbtwo There still seems to be a significant divorce between the unixy bits and the macy bits.
webmind well loads of silent boxes out there
theorbtwo The filesystems have different roots, though.
Juerd_ Not at this price, webmind
webmind www.myelectronics.nl/
yes
at that price
but no macs 23:17
Juerd_ Then not elegantly small
Or?
lumi Filesystems, different roots?
webmind Juerd_, check that url.. and then compleyr systems
the site is crap btw
theorbtwo They seem to...
Juerd_ I'll have a look 23:18
lumi I don't think so
webmind btw did the '...' operator still made it into perl6 ?
lumi You have / and it has your booted partition mounted, at least in my setup, and /Volumes where itautomounts stuff
But ask nothingmuchm he's the Mac guru 23:19
theorbtwo Oh... no, there's just a lot of hidden directories.
lumi There is that
Juerd_ webmind: Ehm. I can't find any system that's comparable to the mac mini in price, size and performance 23:20
webmind: Which one do you suggest I look at?
webmind ah
performance
Juerd_, well mini-itx case + pentium M board ?
not sure about the pricing then
webmind actually isn't sure about the mac-mini's performance 23:21
Juerd_ webmind: Pentium M is expensive.
webmind would you want to run linux or MacOSX on the mini btw ?
Juerd_, but fast 23:22
Juerd_ I'd like to try OSX
But I'll probably end up running debian on it
theorbtwo There's little point in buying a mac just to put debian on it.
Juerd_ theorbtwo: There is with this little machine
theorbtwo ...though I suppose if the ...
Juerd_ It's small and silent.
theorbtwo What I was about to say.
webmind doubt if debian runs as fast on a mac as macOS does.. but I honestly don't know
theorbtwo If you actually have a wish for a small silent box. 23:23
Juerd_ And not ridiculously expensive, as mini-itx systems are
webmind Juerd_, I sugest trying it out somewhere first then
theorbtwo I have only a minor wish for such a thing.
(Could do movies on it, probably better then the laptop I'm using now.)
Juerd_ Compare cheap laptops to mini-itx systems
webmind Juerd_, macmini is 500 right ?
Juerd_ And then realise the cheap laptops come with a big TFT screen built in!
That makes the mini-itx systems very expensive. 23:24
A mini itx board easily costs EUR 100
While a cheap micro atx board costs around EUR 40
Or even less
webmind: Yes
webmind: WITH optical drive!
theorbtwo There's the matter of upgradability, but a mac mini isn't any more upgradable then a laptop.
Possibly less. 23:25
webmind Juerd_, dvd drive right ?
which is slot load afaik :(
I guess it's nice
just not sure about the g4+slowbus performance 23:26
Juerd_ theorbtwo: I don't really care much about that. After two years, I just buy a new one if I think I need one
webmind especially with non-macos
Juerd_ webmind: You know what I use computers for, right?
webmind coding ?
Juerd_ I use a browser and terminals with ssh
theorbtwo I don't get new computers all that often.
Juerd_ That's about it.
webmind yes
Juerd_ And some music playing
theorbtwo And I very rarely do a complete upgrade at once.
webmind so why need the performance of a macmini ? 23:27
Juerd_ webmind: Because it's a hell of a lot faster than the annoyingly slow 533 MHz vias
I have two 800 MHz mini itx epias just laying around because I find them too slow to enjoy 23:28
webmind hm k
theorbtwo For me, I have two rooms. The noise budget in the living room is already blown. Putting a perm machine under the bed in the bedroom might be nice, but it'd have to be quiet, and damn cheap.
webmind sounds usefull
Juerd_ webmind: One will be our print server
webmind Juerd_, seen the dual mini-itx btw ?
Juerd_, hmk
nice
Juerd_ Although that'll be a step backwards from the P3-700 we have now
The good thing though is that it'll be mounted to the wal
wall 23:29
webmind and silent ?
Juerd_ And thus look good in the office :)
A little more silent than we have now. It's not fanless and we'll still be using a normal psu
webmind main advantage of mini-itx to me is size+noise+power usage
hmk
Juerd_ Which it all doesn't have. :) 23:30
webmind well size + power usage advantage? 23:31
theorbtwo Hm. 23:35
The 802.11 will not work under Linux. 23:36
I'd call that a good reason to not run it.
Juerd_ So I'll use an USB dongle 23:37
It's not a big deal. It doesn't come with 802.11 standard anyway
webmind does it have any interfacing / driver issues under linux for any other hardware ? 23:40
theorbtwo www.sowerbutts.com/linux-mac-mini/#support -- audio doesn't work, probably will be fixed soon, this guy thinks. 23:43
bluetooth and modem untested.
webmind audio 23:52
that sounds like crap

Powered by Cro and the Raku® Programming Language.

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

Thank you!