Run Perl6 Now! | pugscode.org ('Overview', 'Journal') | pugs.kwiki.org | chan now logged: xrl.us/e98m | openfoundry down. svn co svn.perl.org/perl6/pugs/trunk
Set by metaperl_ on 4 March 2005.
jdv79 Larry is a little confusing on this Config vars rant... 00:50
his rather
luqui hello? 01:39
hi nothingmuch 01:40
nothingmuch hola
nothingmuch complains about a horrible evening
luqui complains about a horrible yesterday
nothingmuch and with an agenda to actually have fun tonight, looks for something to do
luqui hack pugs
of course
:-)
nothingmuch well, duh, otherwise why am i here?
;-)
luqui Well I'm just standing as the commit proxy while I do other work 01:45
nothingmuch btw, what happenned to elixus.org? 01:47
. o O ( mmm... brahms )
luqui connection refused... what is elixus.org? 01:48
nothingmuch autrijus et al's webspace 01:49
where darcs repo is
svk wiki
luqui hmm...
well openfoundry is down
nothingmuch nevermind, i can wait till that side of the world wakes up
luqui so other sites that autrijus hosts might be down too
nothingmuch my agenda is to learn a bit of haskell tonight
luqui haskell is awesome
luqui learned it last week 01:50
er, most of it
er, some of it
nothingmuch it was up after openfoundry was down for a while, so that's why i assumed it was unrelated
nothingmuch is painfully reminded of having claimed to have grokked C two years ago
after having been leraning it for 24 hours
i still believe i was right
now that I work in it
i new enough other languages to cover all it's concepts 01:51
except that what as i was going along i kept saying to myself 'ah, this is cleaner in C'
luqui haha 01:52
nothingmuch well, cleaner syntactically ;-)
luqui than what?
nothingmuch it's an amazing language in that it defines very little, is rather strict, but allows a lot
inherently you come to accept odd constructs 01:53
luqui perl is an amazing language in that it defines everything, so you never need to define anything
nothingmuch (like duff's device, funny manipulations of things being treated as something they're not via type casting)
luqui well, that's the theory at least
like *(foo*)&mem?
nothingmuch for example
one of the first things i liked was when i learned of fscanf 01:54
the person i helped out (that's how i started learning)
needed to read some user data into an array
and some vars
and one var was a string that needed to contain two values
or something like that
luqui oh no, I might see where this is going
nothingmuch anyway, i started by fscanfing every group of fields that could be read easily together in one call 01:55
several lines into it i remembered that pointers are numbers
in the end one call was enough
with simple arithmetic for telling it where to put what 01:56
rooneg sends in another fix for given.t
nothingmuch ooh fixes
what have i done bad/
?
luqui looking at it
rooneg there are two unexpected successes that are succeeding for the wrong reason
nothingmuch inside the loop? 01:57
that was fixed
and a new test was made to test that bug
rooneg i'm still seeing the problem now...
nothingmuch i think it wasn't applied
rooneg ahh
luqui I wonder why ?? doesn't work...
or maybe it's the wrong precedence 01:58
nothingmuch i concluded it was a parse error
the problem is that 'my $t = (1 == 1) ?? "foo" :: "bar"' sets $t to "foo"
but if you drop the my, like in that example when it was declared in another scope
then it gets set to the result of (1 == 1), and the ?? :: isn't even evaluated 01:59
t/op/assign.t tests for that
luqui weird
nothingmuch Message-Id: [email@hidden.address]
luqui and it's a parse error if you enclose the ??:: expr in parens
nothingmuch that's the submission of that test
in p6v
that's another bug
tested in t/03operator.t at the top 02:00
luqui p6v
nothingmuch first todo
luqui ?
nothingmuch p6c, sorry
ok, i forgot to submit the test to given.t
sending now
rooneg - did you commit? if you did i will not send 02:02
rooneg doesn't have commit access
nothingmuch no reason to create a conflict
ok
luqui so should I go with nothingmuch's test?
nothingmuch well, there's two things
there's that test
which is not related
and a fix to given.t which i'm submitting right now 02:03
luqui okay, so both
nothingmuch (darcs is hanging trying to fetch _darcs/prefs/email from autrijus's repo, which is down)
rooneg has someone already looked at the unpexpected successes in shift.t yet? figured i should check before i start looking into it this time ;-) 02:04
nothingmuch i think stevan did earlier today
luqui rooneg: thanks, applied 02:05
rooneg bah, not interesting anyway, it's just removing some todos, nothing fun in this one ;-)
nothingmuch ok, after some fudging with darcs a bit, the patch was sent to p6c 02:06
(for given.t)
rooneg luqui: thanks
luqui nothingmuch: how the heck do I apply that? 02:08
nothingmuch uh
i can make a unified diff
=/
rooneg if anyone feels like fixing it, svn:ignore could stand to be updated in src (add pugs_version.h, pugs_config.h, and Config.hs) and ext/Kwid (add pm_to_blib), i'd send a patch, but it's not like patch'll understand it anyway
luqui nothingmuch: please do :-) 02:09
nothingmuch *disgruntle*
luqui er wait
I have darcs
I just don't know how to use it
nothingmuch i thought autrijus set stuff up for this
ah
luqui maybe he did, I know nothing about it
nothingmuch pipe the whole email or just the attachment to darcs apply
it will just work with either
02:10 edgewalk1r is now known as edgewalker
luqui ohhh... but my repo is svn 02:10
it needs to be a darcs repo?... hmm
nothingmuch uhuh 02:11
i'll make a unidiff
luqui thanks
nothingmuch should i mail to p6c? make available by http? 02:13
mail to you
?
luqui to me is fine
nothingmuch oh, crap 02:14
rooneg: your fix is correct
luqui hmm? 02:15
nothingmuch #3 should also be done
# TODO confusement - i didn't realize that
(i didn't see it not working on my Test::Harness, that is)
luqui I applied s/#/g/ for that region, against the patches instructions 02:16
shouldn't I have?
rooneg that seems fine, i should have noticed the other #, but i wasn't paying attention
luqui good 02:17
rooneg if i had been paying attention, i would have noticed that it misspells "interleaved" three times ;-) 02:18
nothingmuch you should have
that part i would have noticed
nothingmuch has the documents to prove he's dyslexic
will fax on demand
rooneg can't give anyone any crap about being able to spell/type correctly 02:19
nothingmuch for me the test name's only importance is to be distinct enough to grep for quickly
not that i'm well behaved in that respect at all times 02:20
i tend to forget and make them to englishy
luqui nothingmuch: applied 02:21
nothingmuch wait, if you applied rooneg's second given.t fix which is the same as mine you should have a conflict
luqui i did
i kept his 02:22
nothingmuch are you sure you're not lying?
HUH?!
ok =)
luqui okay, back to hard work 02:23
...on my game
:-p
nothingmuch =) 02:24
rooneg promises he'll get bored and let luqui get back to his game soon enough 02:41
nothingmuch just had an epipheny
foldr is sheer genius
if i got it right
luqui hehe 02:42
no, it's okay
I'm not really working on it... no inspiration
I think I'm going to try to implement your .kv method now
rooneg yay!
i must admit, this pugs thing is addictive. enough of it works that you can have fun, and when you find something that doesn't at least you can write a test and feel like you've accomplished something ;-) 02:43
nothingmuch functional programming is so pretty
luqui nothingmuch: until you try to do something hard 02:44
er, s/you/I/
nothingmuch well, my problem is that i've never actually programmed in FP yet
i love closures in perl 02:45
luqui as Larry says, many paradigms have very applicable situations, but procedural maps the closest to how our brains usually work
nothingmuch but they never add up
they always account for small individual parts of my systems
luqui me too
nothingmuch i've gone through some SICP lectures (video, from MIT)
luqui I write my grammars functionally
nothingmuch and it all seems to fragile
luqui you can design well with it, it just takes a different kind of brane 02:46
nothingmuch but i do keep thinking "oh this is so nice"
luqui like autrijus's
rooneg the trick is to harness people like him and point them towards interesting problems you want solved
nothingmuch i got flamed for plagiarising MAPL
crap
why didn't the other line get in
i said before '/me's first post to perlmonks was when he got closures' 02:47
i never seen anything like that before, and said 'woow, i've got to try to make this useful'
so i took the approx matching algorithm in MAPL and made it faster by constructing matchers
luqui mapl? 02:48
rooneg should i be able to declare a variable inside a loop? i.e. loop (my $i = 0; $i < 10; ++$i) { } 02:52
nothingmuch i was amazed by how hard it was to make people realize that it wasn't the point of the post
mastering algorithms with perl
i think that node still has a rep of -2 or so... what a way to start, eh?
crysflame & 02:56
is MAP out yet?
luqui rooneg: you can declare a variable anywhere
my is just a prefix operator (with some compile-time magic)
nothingmuch in YAHT
exercise 3.5
rooneg ok, in that case here comes a test ;-)
nothingmuch is foldr max 1 [ ] the way the author means to extrac the max out of a list? 02:57
luqui or 1 otherwise, I think 02:58
nothingmuch luqui: who were you talking to? 03:00
luqui you 03:01
nothingmuch in that case, explain what you meant... =)
luqui assuming [ ] means [...]
nothingmuch yes
(you can assume values of list are >= than 1)
luqui then it's just the max 03:02
but if the whole list is < 1, then 1 is returned
nothingmuch right
ingy hola 03:04
nothingmuch morning
buu Yay soon it's going to be illegal to link to a candidates website! 04:40
luqui hmm? 04:53
buu Hrm. Is `*bold*` supposed to be bold and code? Or just code? 05:00
luqui rooneg: you there? 05:27
oh, no, of course not, you left 05:28
Many new features in new r571 05:29
there was a wealth of unimplemented functionality hidden in that little .kv test
it's a friggin mortuary in here 07:39
hi chansen 07:41
chansen hi luqui
ingy hi luqui 07:46
buu: just code :)
luqui hi ingy 07:48
ingy buu: `*just code*` *`bold code`*
crysflame .
luqui confusing 07:49
ingy not really
luqui oh, just code with the stars included
no, not confusing
ingy luqui: correct
`a*b*c=d`
but that is contrived, because a*b*c has no bolding 07:50
'a *b* c' has bolding
luqui how might you get a<b>b</b>c ? 07:51
ingy 'a *b *c' does not
luqui: a{*b*}c
luqui nice
I like pod, but kwid is just like it
ingy without the braces there are "cuddling" rules
luqui only better, it seems
no complex table formatting, right? 07:52
and no automatic doc generation, right?
ingy kwid is just cleaner pod
luqui great
ingy well .table
buu er wait 07:53
What?
luqui but I'm just making sure we're not falling into the html trap
ingy well there is no current plans for .table 07:54
I'm just saying you could do it
buu So wait: `foo *bold* baz`; has no bolding, but *`foo`* has bold and code? 07:55
ingy buu: yes 07:56
buu @
!!!
ingy buu: how else would you do it? 07:57
buu Er, `*foo*` would be bold and code, *`foo`* would be the same
luqui how would you represent the latest syntax
*foo*
buu Er, what?
luqui *foo* = 3, for instance
buu I'll answer 07:58
luqui a new type of variable with stars on either side
wait for it....
buu I was losing in war3 =[
luqui huh?
buu 'huh?' what? 07:59
luqui i don't understand your last utterance
buu I was losing in my game.
Warcraft3.
ingy a bold phrase is subparsed. a code phrase is not. I think the makes the most sense
luqui war3, right
exactly
that's how it is 08:00
`code` *bold* *`boldcode`* `*neither*`
ingy wwpd?
buu heh 08:01
ingy C<<foo B<bold> bar>>
buu When you put it like that, it makes more sense
luqui & reboot
ingy I think that means `foo B<bold> bar` 08:02
buu ingy: Duh, you'd say: \*foo* = bar 08:04
ingy you mean `\*foo* = bar` 08:05
buu if necessary
ingy that is a possibility
I'm not convinced that having bold inside code is what people want, but perhaps 08:06
I can see the possibility anyway
to be determined... 08:07
not going to decide tonight
it might actually be nice to italicize parts of code 08:09
buu Well, yeah. I was approaching it from more of a consistency stand point
ingy I'm not so concerned about that
buu I noticed
ingy there are not 1000 phrase formats, there are about 6 08:10
bold italic code asis link named
luqui asis? 08:11
ingy asis is definitely not subparsed
{ *i* *like* *stars* }
starting tag is '{ ' 08:12
ending is ' }'
luqui does it go in tt font
ingy no
it just means, leave me alone, I don't have any formatting codes, even though it looks like I do 08:13
asis
named is {file: /etc/passwd} 08:14
luqui and that does...
ingy it is the equivalent of whatever pod's F<...> is 08:15
luqui if only I knew what that was
crysflame interpolate slurp file
ingy kwid needs to support all of pod
crysflame: really??????!!!! 08:16
crysflame what?
ingy interpolate slurp file?
I doubt it
crysflame oh, wrong though path. i thought F<> was interpolation of content, not "this is a filename". sorry.
can't say i've ever used it 08:17
ingy ah
I think it is total pod cruft
luqui we could petition (read: mention to) Larry to remove it altogether
ingy that's why I support it with the ugly {file: ...}
crysflame semantic distinction between <b> <em> <strong> reminds of of C<> F<> 08:18
luqui why not {{FILENAME: }
ingy heh
luqui to stay with perl 6 vogue of making things you're not supposed to use big and ugly
ingy {THIS_IS_A_FILE_NAME_DUDE: ...}
luqui something that long is guaranteed to remove your hard drive if the file doesn't exist 08:19
ingy compared to the rest of kwid, {file: } is ugly
hmmm, tempted to support #comment# inline 08:20
nah, yagni
crysflame {# ... #}
ingy crysflame: aye. I started with that, then reasoned that I could also support the simpler form 08:21
but it is yagni
ok, I have a BIG RACE tommorrow 08:22
wish me luck
luqui what kind of race?
ingy 72 mile bike race
with the fastest people in the NW
luqui sweeet
good luck
ingy thanks
I'll probably be dead this time tommorrow 08:23
at least I'll feel like it
knock on iBook
crysflame nice 08:24
good luck
ingy have fun y'all
luqui hiya metaperl 08:26
metaperl hi luqui
luqui wish ingy luck on his 72 mile bike race tomorrow 08:27
metaperl where does he live?
luqui ummmmmm
somewhere in the northwest?
metaperl he used to work at activestate . i guess he quit... oh well, I will chat with him sometime and see 08:28
cls_bsd ossf is back? 08:58
ping alive
luqui ossf? 09:02
cls_bsd well, tha't means svn is back 09:03
metaperl openfoundry is back?
luqui openfoundry doesn't seem to be back
scw But I still cannot pull.. 09:04
metaperl what's the difference between say and print?
scw end-of-line?
luqui one newline
scw ossf response to my ping all day but I cannot touch the reposetry..
luqui openfoundry.org says that it will be back on 3/6 09:05
you
're using the repo at svn.perl.org, right?
scw now, yes.
metaperl we need a pastebot here again 09:06
scw Who can commit to svn.perl.org?
luqui me
metaperl does anyone know the invite syntax 09:08
Somni it's not working
Khisanth /invite nick channel
luqui what is not working somni?
Somni oh, you may have to be an op
metaperl oh
Somni /invite pasteling #perl6
luqui hmm... we have no ops 09:09
Khisanth you do ..
hrm see if any oper is alive?
buu hrm
what?
I can put a pastebot in here
metaperl luqui, in this code you wrote: nomorepasting.com/paste.php?pasteID=33434 09:10
can you explain what &?SUB does
luqui it refers to the innermost lexical sub
in this case, &foo 09:11
it is necessary in case that sub is lexical (but it only applies to subs that were declared with "sub", or "method", etc. not parameterized blocks)
metaperl can a lexical sub be anonymous?
luqui s/lexical/anonymous
metaperl oh, yes. can an anonymous sub be lexical? I think the answer is yes 09:12
luqui yeah
buu hrm
weird
There it is 09:13
metaperl: Your pastebot =]
erxz_paste "buu" at 24.206.236.187 pasted "testing" (1 line) at erxz.com/pb/503 09:14
metaperl thanks buu
buu Np.
If it goes down, blame dngor and or poe.
oh wait
sigh 09:15
metaperl buu, so you play some sort of computer game?
buu metaperl: Yes.
metaperl which one?
I play myth
buu I've been playing war3 recently.
metaperl warcraft?
buu yes.
metaperl hmm. I hate having to build all those things
the controls seemed complicated but I never tried war3 09:16
is there a demo online somewhere?
buu Probably.
metaperl yeah, the demo is at download.com
www.download.com/3120-20_4-0.html?q...ch=+Go%21+
buu Amazing.
But it's sleepy time for buu. 09:17
Nighty night.
metaperl wait a minute there are a lot of war3
which one should I get?
buu uh
What?
metaperl look at that URL
buu sigh
perlbot shorten it
perlbot Shortened URL: xrl.us/fcbp
metaperl there is LOTR, chaos,
buu WAR3
weird
Get reign of chaos.
metaperl ok 09:18
thanks
buu I can't believe you've never heard of war3 =/
metaperl good night
heard of it
heard of starcraft, etc
buu It's by freaking blizzard.
Juerd Who wants free keyboards?
metaperl never could get into them
buu Who made the best game ever.
Juerd: Why do you have free keyboards?
Juerd IBM, black, brand new, PS/2, dutch layout
metaperl best game ever?
buu metaperl: Starcraft dude. 09:19
metaperl I am a Myth player
Juerd buu: Because nobody on this planet uses the dutch layout
buu Juerd: Reasonable.
Juerd That's no problem if you touch type
buu Yes, but can it program perl6 ?!
Heh, did you see that link to the giant keyboard
Juerd It can, actually 09:20
If you use the Dutch layout, you even have buttons for ōæ½xAB and ōæ½xBB.
No
buu Oh. It had like 800 keys.
Juerd Is it a Chinese keyboard?
buu No actually I think it was a musical keyboard.
scw Still not enough for unicode user!!!
luqui Mario 1 you guys 09:21
metaperl Mario 1?
09:21 buu is now known as buu_NOTHEREANYMO
luqui best game ever..? 09:22
metaperl games: another reason to have a PC instead of Linux :)
what is Mario 1? that's your favroite game?
luqui you mean windows?
Juerd "PC instead of Linux"?
metaperl I liked Escape from Monkey Island
Juerd I have a PC that runs Linux!
luqui no, it's the best game ever
metaperl yes, I have windows
luqui i didn't say it was my favorite
metaperl Mario Brothers? 09:23
is there a PC demo of it?
luqui and I think that is "_the_ reason to have windows instead of linux"
you could probably find a nintendo emulator somewhere
Juerd Why instead of?
metaperl you can always run vmware
luqui if you're going to pick one
Juerd They can very well be installed on the same pc at the same time.
metaperl double boot is passe'
vmware r0x
Juerd luqui: Then there's still vmware, cooperative linux, wine, etc
metaperl cooperative linux? never heard of it 09:24
luqui doesn't like vmware or wine
Juerd I run MSIE under Linux; works very well, except that it's still MSIE, of course.
metaperl any URL to Mario 1?
Khisanth scw: you should try that keyboard with 72 keys :)
luqui but has dual boot, and (painstakingly) reboots
scw Khisanth: Which one?
Juerd metaperl: coop. linux is linux in windows
luqui only played mario on the original nintendo
Juerd a linux kernel that runs as a windows service
It's efficient
luqui oh, neat
feber luqui: colabti.de/irclogger/irclogger_log/...&sel=3 09:25
Khisanth scw: www.engadget.com/entry/1234000870034519/ this one 09:26
luqui feber, I was talking about the index page 09:27
of all channels logged
but it appears to be fixed
Khisanth luqui: www.zophar.net/unix/nes.html NES emulators :) 09:29
feber luqui: nothing was broken. :)
luqui hmmm
okay i'll believe you, I can't recreate my mental state
that keyboard is frightening
metaperl is there really no cleaner way in Perl6 to read from STDIN thatn my $line = =$*IN 09:30
luqui man if I could get fast at that, that would rock. but I've been playing the "regular" piano for long enough that I'd rather not learn a new instrument...
metaperl how is that dirty?
metaperl compare with my $line = <STDIN>;
luqui (you can also s/\*//)
how is that clean?
metaperl two equals signs looks weird 09:31
why are there 2 equals signs?
luqui one for the assignment, the other for the "iterate" operator
which looks like two lines
but please, if you have other suggestions, shoot
that's why we're redoing perl: to fix our mistakes 09:32
metaperl luqui, I think you will like Ehren Starks: magnatune.com/artists/ehren
scw Khisanth: "But, like decimal time, and the Dvorak computer keyboard, ..." yes, I'm the one use dvorak but may not try this :p
metaperl that is some of my favorite piano music these days
Khisanth scw: but it has enough keys!
luqui I'm kinda a classical guy, but I like the occasional venture into modern music. I'll have a listen 09:33
09:33 feber is now known as feb
scw Khisanth: Er, no. You still have to combine several keys together to type a Chinese charactor :) 09:33
Khisanth hmm true 09:34
luqui man she's a good cellist
Khisanth but you would have to start growing extra fingers too
metaperl why did ? : becomes ?? :: 09:35
luqui because ? : were too short for what they were doing
they just blended in with their surroundings when they should be standing out
that and we wanted ? and : for other things
Khisanth is ? : being used for something else?
metaperl how about $IN.next to iterate an iterator? just brainstomring
luqui : is used in every context you can think of
? isn't anything in operator position yet 09:36
metaperl oh in the new Perl it is . I see
Khisanth that non blending thing actually makes a lot of sense ...
metaperl I thought it would be wild to have a different sigil for junctions:
luqui I've been arguing that for some time
metaperl my ?password = none(qw(a b c)) :
$password eq ?password; 09:37
luqui ahh, but ? does mean something in term position like that
metaperl oh really?
luqui yeah, it's the "boolify" context operator
along with unary + and unary ~
metaperl boolify? lol
Khisanth "no, you can't have a junction sigil because there aren't enough keys on the keyboard!" :p
metaperl i've been wondering what +@array_name does
luqui it's just an explicit way to get the length. puts it in numeric context 09:38
metaperl scalar @array_name
@array_name.elems
luqui Khisanth: hehe. I've actually been arguing a secondary sigil
metaperl that'll do it even more
Khisanth like $$ ?
luqui like $~ 09:39
it looks like a "tie" of sorts.
lightstep lazy binding?
luqui I've been arguing it for junctions and other tied things
lightstep: hmm?
cls_bsd ossf--
lightstep in haskell, ~(a:b) binds a and b lazily
Khisanth metaperl: shouldn't scalar @array; return a ref to the array?
luqui oh... 09:40
metaperl a ref?
luqui doesn't know that part of haskell yet
lightstep (in pattern position_
luqui isn't everything lazy in haskell by default?
metaperl is ~(a:b) haskell or perl?
lightstep it's a pattern matching thingy (in haskell)
luqui haskell
metaperl what does the tilde do? 09:41
lightstep it makes the compiler stop checking for patterns after that obe
metaperl : is the cons operator
lightstep so if it doesn't fit, it's an error
luqui hmm
Khisanth metaperl: maybe it has changed but the last time I was reading through one of the exegeses, it was $array_ref = @array;
metaperl oh I see 09:42
$array_ref = \@array
luqui scalar @array does it too
metaperl was just fine
why is everything changing?
luqui hehe, the newcomer attack
Khisanth ask Mr. Wall :)
metaperl no, this is a community rewrite of Perl 09:43
luqui I'm part of the community
I love where it's going
Khisanth ask Mr. RFC?
metaperl there are going to be many distraught people where I work... I can see it now
luqui basically, Perl 6 is a different language from Perl 5, it just feels the same and is based on the same principles
metaperl people at my company are going to code in Perl5 style in Perl6. it's going to force me to switch jobs 09:44
Khisanth it's not like p5 will disappear instantly
luqui we're trying pretty hard to make the migration process as easy as possible
metaperl my boss is a php-weenie who programs in Perl like PHP
Juerd luqui: But don't say that when Damian's listering :)
metaperl and he tries to dumb me down to his level
Juerd luqui: Damian keeps trying to convince people that it's not a new language :)
luqui Juerd: well it is.
those of us on the design team can think otherwise 09:45
Khisanth well more like a mutant p5 :)
luqui because we've been learning it for three years
metaperl it's an effort in perfectionism
instead of gradual improvement
Juerd metaperl: It's gradual, but without the software.
metaperl lol
lightstep what is the context thingy? are expressions evaluated with thought about their context? how is this reflected in pugs? 09:46
Juerd It's not as if Perl 6 will just leap into existence
metaperl how did Autrijus manage to singlehandedly outstrip the whole ponie/parrot team ?
Juerd lightstep: Context is the same thingy as it always was, but more complex and detailed when you need that.
nothingmuch morning
metaperl he has done so much in under 1 month
lightstep Juerd, i never learned perl
Juerd metaperl: Some other kind of sugar cubes
luqui he's been working constantly, and we all have other jobs, and we're all lazy slackers
nothingmuch luqui: are you human and not some cabbage or something? 09:47
Juerd lightstep: Then now is a good moment to start learning it :)
luqui nothingmuch: what?
Juerd lightstep: Don't worry about learning something you won't be able to use for long, as it'll be years until we have Perl 6
luqui er, What?
nothingmuch you should sleep occasionally
luqui (deserves a capital)
nothingmuch oooh, we're playing with the transcript... Let me get my principia
luqui oh. I just go on a six day week 09:48
lightstep is there a good intro, from the language perspective? (in contrast to "let's get to make some example work")
Juerd lightstep: tnx.nl/intro.bp
eh
lightstep: tnx.nl/intro,bp
nothingmuch do ye wish to better thyself? 09:49
Aankhen`` juerd: Are you the same Juerd as on PerlMonks?
Juerd isn't that does thou?
Aankhen``: Yes. 09:50
metaperl no, he's princepawn on perlmonks
Aankhen`` Might I say thanks for the section on PHP vs. Perl on your site? :-)
Juerd metaperl: I should kill you for that.
metaperl LOL
Juerd I am NOT metaperl/princepawn
metaperl: You know how I think about princepawn, and that I like metaperl a lot better.
Aankhen``: You might :)
Aankhen`` A million thanks! 09:51
It's been very useful indeed.
Juerd You're welcome
And so are patches
Aankhen`` For?
Juerd tnx.nl/php :)
Aankhen`` Ah, for the document itself.
Juerd It's still far from complete, but I lack both time and motivation to study PHP better. 09:52
Many hours went into that document already
I spent a whole day just finding the info for the table of matching operators
Aankhen`` Yikes.
Juerd (Mostly by trying and from PHP's source)
nothingmuch is not nothingmuch on perlmons 09:53
Juerd afk 09:57
Aankhen`` Could anyone suggest a good regex tutorial for a complete beginner? 10:16
luqui doesn't even remember how he learned 10:17
perlretut
?
Aankhen`` I need something available on the 'Net. 10:18
sorje perldoc.com ;-)
Aankhen`` Good point.
I wub you.
nothingmuch Aankhen``: i'd actually suggest perlre alongside perlretut
If you feel you are 'getting' it, but are more curios about how to do xyz, it's a reference 10:19
Aankhen`` I'll tell him to go through perlretut to get the gist of the thing.
After that he can look up perlre.
nothingmuch perlretut is an intro, but doesn't cover a lot
Aankhen`` Only problem is, perldoc.com never opens for me.
nothingmuch yah, it sucks that way
search.cpan prolly has it, since it has perl
and there is also perldoc.perldrunks.org, i think
Aankhen`` Ah, that's perfect. 10:20
Thank you very much. :-)
nothingmuch search.cpan.org/~nwclark/perl-5.8.6...lretut.pod
Aankhen`` I'm looking at the one on perldrunks.org.
nothingmuch or rather, the way i typed it: search.cpan.org/perldoc?perlretut
theorbtwo perldoc.perldrunks.org 10:26
luqui & bed 10:44
Khisanth strange junction 10:45
luqui bed & luqui 10:46
what about a strange junction?
oh, 10:47
duh
I'll say it again... and mean it
luqui & bed
crysflame &
theorbtwo Hm, pugs is mentioned extensively in the newest perl.com newsletter. 11:03
autrijus_ greetings 14:42
Khisanth yo~ 14:43
autrijus_ luqui++ # multival for() and .kv
I've fixed most of the problems luqui documented in his commit log 14:44
r572
scw autrijus_: oops, I made a patch for it but try to fix another problem w.csie.org/~scw/pugs/kv-on-ref-and-list.diff 14:48
autrijus_: I'll re-generate it... wait a minute
autrijus_ cool
mm, anyone got a copy of perl.com newletter?
<- was subscriber, but somehow not receiving things anymore 14:49
scw Er, I'm comfused by svk.. 14:57
I'll use original patch format..
autrijus_: same url, svk patch format 14:58
autrijus_ cool 15:03
scw autrijus_: But I found it really ugly that kv, values and pick all dereference themselves... 15:04
autrijus_ w.csie.org? 15:05
can't fetch it
scw Oh? then www.csie.ntu.edu.tw/~b92103/pugs/kv...-list.diff 15:07
scw always lazy on typing longer, official one zzz
... connection refused 15:08
autrijus_ also blocked.
scw Seems been DDoSed XD 15:09
autrijus_ ....
nopaste?
mail to p6c?
scw autrijus_: mailed to your email ZZZ 15:13
another server also down, and failed to mail from department. Today is not my day :/ 15:14
autrijus__ it is not.
nopaste it? 15:15
erxz_paste "scw" at 218.174.159.24 pasted "patch for .kv to work on ref and list" (23 lines) at erxz.com/pb/504 15:17
autrijus__ I agree 15:18
there should be only one single cast
to a hash
I'll drop your patch and do that. 15:19
scw ok :)
rooneg loves that all he had to do was post a test for %hash.kv and when he wakes up in the morning it's been implemented 15:20
autrijus__ :) 15:24
15:24 autrijus__ is now known as autrijus_
clkao hi rooneg 15:24
autrijus_ it's the "oil for food"
I mean "test for feature"
program
rooneg clkao: nice to see so much heavy svk use on this project, good publicity for you ;-) 15:25
autrijus_: yeah, i'll have to find more features i want to play with that don't have tests yet
autrijus_ scw: so it's just
clkao and the svn server prompts went down to promote svk
autrijus_ op1 "kv" = \v -> return . VList $ concat [ [key, val] | (key, val) <- vCast v ] 15:26
clkao s/prompts/promptly
autrijus_ keep it simple, indeed
nothingmuch morning
autrijus_ actually, to rewrite in point-free style: 15:27
op1 "kv" = return . VList . concatMap (\(k, v) -> [k, v]) . vCast
scw and make vCast accept VRef ?
autrijus_ vCast already accepts VRef.
see AST.hs.
scw autrijus_: Fall to the default one in ``class Value n''? 15:28
autrijus_ Gah
nothingmuch autrijus_: funny bug... see t/op/assign.t
autrijus_ I misremembered.
my plane is actually ~7hr from now.
autrijus_ panicks
well, there goes sleep
nothingmuch =)
i think it's a percedence thingy
theorbtwo Where are you going?
rooneg sleep on the plane
autrijus_ is there a t/op/assign.t? 15:29
theorbtwo: back to .tw
nothingmuch i sent a patch, and i think luke applied
autrijus_ did he.
nothingmuch i dunno
autrijus_ not seeing it.
I'll proc when I wake up
need to totally sleep now :-/
scw nothingmuch: Hi, I don't have commit permission on svn.perl.org, but would like to know what's the bug you found 15:36
nothingmuch: could you paste the assign.t somewhere? it's not in tree yet.
nothingmuch one minute 15:39
(sorry, my dad needed help with his cello)
ok, i put up a darcs repo on nothingmuch.woobling.org/pugs 15:41
t/op/assign.t should be in there
(once rsync finishes) 15:42
ingy hola
nothingmuch hola
growl.info/files/Growl-0.6-Bindings.tar.gz
nonono!
ugh, how did that happen
nothingmuch.woobling.org/pugs/t/op/assign.t
scw? 15:44
scw Yes, I've seen them 15:45
But what's the difference between the first and the fifth? 15:46
nothingmuch my $t; $t =
scw oh, assignment or default value
nothingmuch vs. my $t =
if you look at the parse tree, it appears that under non declarative assignment $t is getting the value of the cond's boolean result 15:48
and the ?? :: is not getting executed at all
autrijus_ nothingmuch: assign.t is committed
nothingmuch jolly
autrijus_ and yes, your diagnosis is correct.
clsung++ # "Hacking Pugs at such an early stage is a lot of fun... I feel like I'm playing the Pricess Maker game from GAINAX." 15:49
scw and the ?? :: part is... awful, I think
how gcc or perl5 implements "? :" ? 15:50
autrijus_ I don't know, honestly.
You'd be my hero
if you can figure out how to deal with ??::.
I mean, pmichaud punted, and I punted
so if you can get it working, that'd be super cool 15:51
scw I think I can make it work -- in some very, very dirty way. 15:52
So referencing to others implementation should be better. 15:53
autrijus_ you are welcome to check them :)
nothingmuch autrijus_: remember my earlier promise to learn haskell?
autrijus_ nothingmuch: yes?
nothingmuch well, yesterday i spent my first 30 minutes or so
i got as far as the basic IO
autrijus_ YAHT? 15:54
nothingmuch yes
now i'm having a bit of a hard time getting the number munging program compiling properly
yesterday i was too tired
but i'm hopeful for today
autrijus_ cool! nice!
nothingmuch++
nothingmuch excercise 3.10
nothingmuch dances
scw I am blocked by Manod... 15:55
nothingmuch tr/ao/oa/
i om blacked by monad
that makes sense, huh?
beh. Now i've got to go to jerusalem to return a bass... Ciao!
autrijus_ be glad you are not blocked by Manos
nothingmuch que Manos?
autrijus_ en.wikipedia.org/wiki/Manos 15:56
nothingmuch ah 15:57
one of those
see my first or second patch to pugs ;-)
autrijus_ yeah I saw that. it was cute
nothingmuch i remembered =) 15:58
i was more noting to scw
anywho, i best be off, i'm already going to be late
malaire "All about monads" seems to be a lot better introduction to Monads than YAHT chapter 9. (I'm reading it after YAHT chapter 9 seemed to be too hard to understand) 16:00
scw hmm... then I'd better try it :)
malaire It does some basic understanding of Haskell, but YAHT (chapters 1-8) seems to be enough 16:01
scw autrijus_: You didn't fix the reference problem!!!
malaire s/does/does require/
scw autrijus_: But go to sleep first :) 16:02
rooneg is it just me, or does t/op/string_interpolation.t not match how S02 says it should work? S02 says arrays are only interpolated when they have empty brackets after them, but the last test there has a bare "@a" being interpolated... 16:05
shouldn't it really be "@a[]"?
scw rooneg: How about this? <quote>In string contexts these container references automatically dereference to appropriate (white-space separated) string values. In numeric contexts, the number of elements in the container is returned. In boolean contexts, a true value is returned if and only if there are any elements in the container.</quote> 16:10
rooneg i'm looking at the part "In order to interpolate an entire array, it's necessary now to subscript with empty brackets:" 16:11
autrijus_ p6l! p6l! 16:12
(luqui wrote that part; he is a cabal member; however he still might be wrong.)
joodie good evening
autrijus_ (I mean, he wrote the interpolation part of pugs)
so p6l?
greetings joodie.
rooneg damn, i was hoping i could avoid subscribing to that list ;-)
rooneg goes to subscribe to p6l
autrijus_ well you can certainly post to p6c :)
but it's more ontopic of p6l.
scw hmm... seems what I found is unrelated to this one 16:13
rooneg yeah, i'll cc p6c, since it's a question about the pugs impl vs the spec
scw and pugs should not make @a=(1,2);"@a" into "1 2", but "@a[]" into "1 2".. 16:14
rooneg scw: if the spec is right then yes, that's exactly what i'm saying. i've got tests for it if someone wants to commit them, i just wanted to confirm that it was right.
scw rooneg: I'll keep my eyes on p6l :) 16:16
rooneg writing the email now
autrijus_ rooneg: can you attach a request at end of it that calls someone to update synopses to mention unary =? :) 16:18
prefix:<=>
if not, that's okay too, but I don't want to move io.t to t/unspecced/
rooneg doesn't know enough about what unary = should do to write a useful request ;-) 16:19
autrijus_ oh ok :)
I can do that later :)
unary = is readline.
rooneg ahh, right, the confusing syntax i can't wrap my brain around ;-) 16:20
autrijus_ exactly!
rooneg notes how remarkably easy it is to see what's wrong with a diff right after you send it to a public mailing list 16:24
scw so that commits usually appear pairing :p 16:25
rooneg i wonder what percentage of commit messages worldwide contain the word oops ;-) 16:26
joodie I've been writing some tests on subroutine prototypes (mostly to try out the syntax for myself) and I found that sub swap(*@_ is rw) is not supported yet. 16:27
in fact, it doesn't compile :-) 16:28
autrijus_ yup. write a test 16:30
the internal support is there
just need parsing for "is rw"
and add a isLValue boolean. 16:31
that is a simple task for people interesting on Parser.hs hacking.
in Param parsing bits.
s/isLValue boolean/isLValue = True/
joodie I've got a test. Taking a look at Parser.hs - haskell looks funny to me 16:33
scw autrijus_: I wonder that are Rule.hs and Rule/*.hs copied from Parsec? 16:35
autrijus_: If so, why do you do that instead of simple use it?
autrijus_: To be easier to hack it?
autrijus_ yeah because.
that
rooneg notes a lack of tests for things like $?LINE and $?FILE. any suggestions as to a place to put them? 16:36
autrijus_ Rule needs to be in "Eval ..."
so I need to hack Parsec's internals a lot
and syntaxninja also cleaned it up to be run under -Wall
once all these is settled I'll push upstream.
and I noted that it's from parsec; see Rule/LICENSE and README 16:37
scw autrijus_: So you are doing two projects at once :p
autrijus_ uhm.
more than 100, in fact.
it's just some are more active than others.
journal up; I'll sleep now.
scw Er, ok, that true..
autrijus_ rooneg: I think t/base/ ? 16:38
oh wait
t/magicals/
is the place.
rooneg autrijus_: ok
autrijus_ zzz& 16:39
ingy seen buu? 16:53
perlbot: seen buu?
perlbot: smell you later
buu: '-' lists should be rendered with dl/dt/dd 16:54
buu: see www.perlpod.com/5.9.1/pod/perlpod.html 16:56
&
this sucks: svn.openfoundry.org/pugs/ 16:58
gone for real& 16:59
lukhnos ingy: openfoundry.org has been that for *weeks* 18:27
18:58 edgewalk1r is now known as edgewalker
qmole ;/ 19:06
20:14 buu_NOTHEREANYMO is now known as buu
lightstep how can i tell the makefile to build pugs with profiling? 21:13
theorbtwo lightstep, first make sure you have the approps files installed. Then at the bottom of Makefile.PL, after the stanza beginning $pugs, add another stanza: 21:31
pugs-prof : src/Config.hs @srfiles $version_h (newline tab) ghc --make -prof -auto-all -Wall -fno-warn-missing-signatures -fno-warn-name-shadowing -o pugs-prof src/Main.hs -isrc 21:33
You might also try -caf-all.
lightstep without a dollar? 21:34
theorbtwo Dollar where? 21:41
Oh, at the beginning. No; otherwise on windows it'd expand to pugs.exe-prof
lightstep yes. i got it. but ghc here doesn't like -caf-all 21:42
theorbtwo The recipie I gave is valid for unix, but not win32; adjust to taste.
Neither does mine.
lightstep life.p6 seems unbelievably slow 21:43
theorbtwo I wonder if my Int $foo; is useful.
Khisanth lightstep: get things working first, optimize later :) 21:44
lightstep exactly what i want to do with it
is hanoi.p6 broken? 21:57
it gives me AS AS SS AB SS SB SB, and on closer inspection, $s turns out to always be 'S' 21:59
luqui arrrggggh why doesn't 'try' (in parsec) work like it's supposed to! 22:51

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!