6.2.3 is airborne! | pugscode.org <Overview Journal Logs> | pugs.kwiki.org
Set by autrijus on 12 May 2005.
autrijus like ovid said, you have to really try to do that :) 00:00
obra clkao++
clkao++ # might do charts on the rt front page.
autrijus but now, really sleep. 8am. :) &
clkao it looks prettyuy! 00:01
crysflame mm, sleeep 00:02
svnbot6 r3134 (eric256) -- Added tests to show junction bug
clkao run the new bot and do karma 00:03
eric256 adios all. feels good doing my little part. ;) 00:04
revdiablo autrijus: thanks, I guess I should have thought to check the README :) 00:08
drbean pugs requires parrot for rules, but parrot-0.2.0 has no Makefile.PL 00:16
It appears I have to run perl Configure.pl first 00:18
clkao is there Makefile.PL in perl ? 00:20
you have to run Configure.sh, not even .pl 00:21
drbean In parrot, you mean? 00:23
I tried to install parrot with CPANPLUS, but failed 00:24
revdiablo running Configure.pl by hand works for me, drbean 00:25
drbean Yeh, it seems to be being built, running by hand. 00:27
merlyn builds the PUGS release in another window
hey - is this s aknown problem - ld: warning multiple definitions of symbol _BC 00:29
and if so, can I ignore it?
that's for _BC and _UP and _PC
Juerd I've not seen it before.
merlyn looks like it's curses related
drbean clkao: I don't see any Configure.sh, just Configure.PL
s/PL/pl/
merlyn 6.2.3 built and installed! 00:30
Juerd merlyn: Great!
merlyn: Now you can start writing Learning Perl 6 ;) 00:31
revdiablo pugs it up, baby.
merlyn Eventually sure
I gotta finish alpaca 23d first
2ed
due out in november
Juerd still has to read it
merlyn probably start writing in june or july
Juerd But I promised myself not to read any book until I finish HOP
I never finish books :( 00:32
And this time I really want to.
merlyn: As long as it's in stores by christmas (when Perl 6.0.0 is released), any schedule suffices :) 00:33
Odin-LAP Ever the optimist...
merlyn of course, which christmas. 00:34
Juerd Odin-LAP: I didn't say which christmas... :)
merlyn correct. :)
mugwump it will be all our christmasses coming at once!
Odin-LAP Juerd: 2106 sound reasonable?
Juerd Odin-LAP: No.
merlyn 2106 in what base? 00:35
Juerd Odin-LAP: The world ends in 2037, so it must be before then or all our efforts were meaningless
Odin-LAP merlyn: Hexadecimal, of course.
merlyn the world ends dec 23, 2013
so say the mayans!
and look where they are... dead!
mugwump 2012 actually
merlyn right
sorry
Juerd merlyn: The mayans obviously don't use unix time stamps. 00:36
merlyn apparently not.
Juerd They end january 2038
merlyn for the last few months, I've been sending out x-mayan-date: headers. :)
Juerd And we humans always try to solve problems before they occur
So we'll have to kill the planet before 2038
merlyn thansk to gnu emacs. :)
mugwump "significant" astronomical events on or around that date include: Sol's passing through the equatorial plane of the galaxy, and the precession of the earth's Equinox into Aquarius 00:37
merlyn o/~ the dawning of the age of aquarius ... o/~
Juerd I expect perl 6.0.0 by christmas 2006 now
But I have been too optimistic in the past :) 00:38
merlyn it's also merely a rollover in their calendar too
they believed 13 major cycles was enough
obra Juerd: which xmas?
merlyn and that' weve been through 5 of those long sets of 13
or something like that
wikipedia knows more than I do
Juerd obra: 2006 00:39
merlyn baktuns - there it is
obra Juerd: which xmas 2006? Orthodox or Catholic?
Juerd obra: Mine, and I'm neither.
perlmonks.org/?node_id=204115
I guessed 2003-07-05 before :) 00:40
Odin-LAP obra: The distinction can be traced to julian/gregorian calendars. So it's not just the catholics, but also protestant churches... 00:41
obra Odin-LAP sure.
merlyn sorry. we're in the 4th long day. the 5th begins dec 21 2012, or maybe dec 23 2012. 00:43
hard to tell which
thanks to the catholic church burning all the writings of the priests who visited the mayans
(catholic church)-- continually wanting us to live in darkness
Odin-LAP merlyn: Hmm. Confirmed burned, or possibly in the papal archives? 00:44
merlyn well, same net effect.
but from what I understand, confirmed destroyed
what little we still now have is from the memory of the priests who wrote the originasl
Odin-LAP Hm. That sucks.
merlyn at least, that's what I recall from visiting central america a few months ago 00:45
when I started adding my mayan date. :)
Odin-LAP (As does the catholic church and most of organised religion, but that's another story.)
Juerd It'd really suck if they turn out to be right, though.
merlyn just like the catholic church really trying to supress copernicus
etc.
it's all about control bay-bee.
just like the republican party right now.
turnign the US into their own playground 00:46
Odin-LAP Bah.
Juerd All politics and religion are about control.
ajs_home I'm about to send out a follow-up to the "we'll take S29" mail, letting peopel know to send mail to me with suggestions, etc. Anything anyone wants in that message?
mugwump right now? it's been going for a LONG time merlyn :)
merlyn true. dems have troubles on the other side.
getting the populus depedent on government
Odin-LAP Western "democracy" is a sham. A ploy to hide the powerplaying by various people.
merlyn Yeah, the closest I label myself is a libertarian, and even that has troubles 00:47
mugwump it's a great way to excuse the need for consensus on any issue
Juerd Democracy can work
Odin-LAP merlyn: Libertarianism is yucky too, IMO.
ajs_home Ok, sending
Juerd merlyn: It's funny how america's left wing is considered right wing in the netherlands
we go lefter! :)
merlyn someone put it nicely. you want to have a debate that gets about 49% of the people on each side.
so you slide the argument left and right until that happens.
then that aargument can be a lightning rod to keep people from asking about real issues
Odin-LAP The U.S. only has a right wing. It's got an extremist right, and a far-extremist right. 00:48
mugwump that's 49% of the people who vote, too
merlyn apparenly, the current axis debates are on gay marriage and such
that's apparently a 49% split issue
like abortion was a few years ago
Juerd It quite schocked me when a few days again I learned that in the US, 1 out of every 150 or so people is in prison. 00:49
merlyn and a child is beaten every 14 seconds. poor kid. :)
why wouldn't you believe that 1 of every 150 people are so below average (6 sigma deviants) that they deserve to be locked up? :") 00:50
that makes sense to us. :)
Odin-LAP Nobody deserves to be locked up. 00:51
Juerd merlyn: Is that 6? I'd guess around 4.
Odin-LAP Have you heard of the Stanford prison experiment?
Juerd Probably less, 3.1?
merlyn: Yeah - that it makes sense is the scary part. It's normal for you.
merlyn we convict on a curve in the US. :) 00:52
mugwump prisons are just as bad as western hospitals, functionally. ie, they work sometimes but generally are going about it all wrong..
merlyn there's no absolute wrong, just relatively bad enough to look really wrong by comparison
Odin-LAP merlyn: I'm sure that's of comfort to ODESSA.
merlyn well - I figure it's how I got convicted. 00:53
mugwump if they released the 75% of people in prison for drug crimes and started selling them through pharmacies to undermine the gangs that would be a good start
Juerd wouldn't want to live in, or visit, the US
merlyn I was an easy target. "looking whre the light's good"
legallizing most vice crimes would proabbly eliminate 60% of the prisons
but that'll never happen here
Odin-LAP merlyn: Of course. If they were to start looking where there's no light, they might have to actually spend some effort!
Juerd "soft" drugs will eventually be legal here
mugwump shit, I thought this was perl.org #perl for a moment.
Juerd It has to go that way.
Odin-LAP Juerd: Hm. For the whole country, then? 00:54
Juerd mugwump: Why does that matter?
merlyn I'm still trying to figure out where I'd rather live though.
Juerd Odin-LAP: It can't be legal for only a part of it.
merlyn We still enjoy relatively more freedom than almost any other country.
everythign taken as a whole.
Juerd Odin-LAP: We have a single law system for the entire (small) country.
Odin-LAP Juerd: I understand the Amsterdam police are quite liberal with their interpretation..?
merlyn it is indeed time to garbage-collect the case-law of the US 00:55
Juerd Odin-LAP: All are, all are.
merlyn as in hit the big reset button
Juerd Odin-LAP: Amsterdam just likes to benefit from the extra tourism :)
And Amsterdam is the only place where a tourist can easily find the "coffee" shops.
merlyn I suspect I'll marry a mexican, and move there.
that seems like the easiest path out.
Odin-LAP Juerd: Hmm. I see.
Juerd (To a dutchman, a "coffee shop" is not a place where you get coffee. Remember this if you ever visit our country.) 00:56
merlyn oddly enough, they want to marry me to get to the US. :)
Odin-LAP merlyn: Sure. Just so long as NAFTA doesn't screw them too much...
merlyn maybe I'll meet a great girl in brazil.
gonna be there for two weeks
Odin-LAP Juerd: Heh.
Juerd Too many foreigners enter these pot houses thinking they can get a cup of coffee. It's not even funny anymore.
merlyn got my visa a few days ago, being on edge the entire time.
but done deal. 00:57
time to go finish some more beer.
& beer 30
mugwump huh? I've bought coffee in an Amsterdam coffeeshop before :) 00:58
with a piece of space cake on the side
Juerd www.xs4all.nl/~tluif/Humor_bestande...nders.html 00:59
svnbot6 r3135 (ninereasons) -- convert force_todos to per-line :todo, and fix a broken example 01:00
Juerd Most of it is true.
obra amsterdam++ #Good eetcafes
Juerd obra: You're not picky enough then :P
obra Juerd: no. I spent three days looking and finally found a couple very nice ones
Juerd Okay :) 01:03
Juerd never spent three days in amsterdam 01:04
I don't like that city.
01:10 ninereasons is now known as mkmcconn
drbean Netherlands seems to be a great place for perl, but although Guido 01:13
is also from there, is python as big there?
Juerd Nah 01:17
Though python attracts more new programmers than perl does 01:18
nite! 01:19
afk
f0rth version_h.pl hates svk ._. 01:44
because my svk installation speaks Chinese 01:47
svnbot6 r3136 (eric256) -- Fixed up some, actualy does damage and ends. 02:08
eric256_ realizes that commit comment doesn't make much sense out of context 02:09
mugwump lists.utsl.gen.nz/pipermail/pugs-co...00467.html makes sense, though :) 02:12
eric256_ that is spefiy. the bot should provide that link ;)... to bad IRC doesn't realy have a link sending protocol. 02:19
meppl gute nacht 02:40
svnbot6 r3137 (jhorwitz) -- * return proper exit codes from system() 02:55
r3138 (Stevan) -- Tree - changed from Tree::Simple to Tree; it works too 04:55
r3139 (Stevan) -- Tree - moved Tree::Simple completely to Tree
mugwump & # weekend 05:19
Khisanth hcchien: did you notice the misspelling of MultiDaysEvent? :) 05:46
svnbot6 r3140 (mugwump) -- Add union/different/etc operators to Set
hcchien Khisanth: maybe not? can you give me a hint? :) 05:52
oh, I saw that. :/ 05:53
Khisanth was just looking for some calendar generating module 05:54
what exactly is a pugs live cd?
svnbot6 r3141 (Stevan) -- Tree - More methods and tests added/converted 06:15
06:45 chady_ is now known as chady
mj morning 07:07
svnbot6 r3142 (scook0) -- Haddocks for Monads.hs 07:21
r3143 (scook0) -- Tests for slurpy hash params
Aankhen`` Yey, no more notices... :-) 07:44
svnbot6 r3144 (scook0) -- Updated Haddocks to reflect Autrijus's answers to my questions 08:16
r3145 (scook0) -- Fixed & added Haddocks for Bind.hs 08:22
tphyahoo Hello. This is my first time using irc. I hang around perlmonks, am interested in pugs, and just thought I would check this out. 08:25
clear 08:27
All right, I've got to go. Will check back in later. :)
f0rth is it ok to replace "grep /Mirrored From/" with "grep /https?/" and "/Rev\. (\d+)/" with "/ (\d+)$/" in util/version_h.pl to work with non-english svk? 09:23
clkao f0rth: reset the LC_* env before invoking svk 09:24
f0rth and windows? 09:25
having tried that morning on win32 but not working 09:26
clkao ah. bah 09:27
gaal has anyone had any luck building the latest snapshot parrot on either msys or cygwin? 09:50
Daniel_Nee Dear Autrijus: could you please check your email, I need your help on Monday or Tuesday afternoon, Thanks a lot. 10:37
Aankhen`` Assuming the Parrot source has been checked out into C:\Development\parrot -- and built there, of course -- what would PARROT_PATH need to be set to? 10:53
Just setting it to C:\Development\parrot doesn't seem to work. It complains about a lot of .lib files.
(I verified that it only happens when I set PUGS_EMBED to "parrot")
Aankhen`` wanders off and hopes he'll come back to an answer. 10:55
iblech Khisanth: The Pugs Live CD is a small bootable cd which boots directly into Pugs 11:11
11:12 chady is now known as chady_
iblech Khisanth: I plan to add Parrot to it, and some example programs, and networking support, etc. 11:15
autrijus f0rth: yes, please fix the version_h 11:25
f0rth: you're a committer now, btw. welcome aboard!
nothingmuch gugod: ping 11:59
Limbic_Region perlmonks.org/index.pl?node_id=456696 # pugs examples being used as a reference source at the Monastery 12:47
svnbot6 r3146 (autrijus) -- * Partially revert jhorwitz's change on `system()`.
r3146 (autrijus) -- Perl 6's `system()` actually returns true on success and
r3146 (autrijus) -- false on failure. We can, though, set $! as the exit code.
r3147 (autrijus) -- * "make haddock" should clean up after itself.
r3148 (autrijus) -- * minor cleanup.
autrijus Forth: I've implemeted your suggestion. You may wish to add your chinese name to AUTHORS
svnbot6 r3149 (autrijus) -- * fix pugs_version.h generation for svk users, as suggested
r3149 (autrijus) -- by Forth.
r3150 (autrijus) -- * add Forth to AUTHORS.
Limbic_Region holy cow svnbot going nuts
www.perlmonks.org/index.pl?node_id=456587 # P6 wizard game by our newewst convert eric256 12:48
Juerd What's with this www. thing? 12:51
Do you hate your hands?
Limbic_Region Juerd - heh, it is called a bookmark 12:54
1 click beats all those key strokes
hey eric256 13:02
eric256 hey 13:03
Limbic_Region is glad to see you were easily swayed to the dark side
muhahahahaha
eric256 lol
so when %.weapons (where weapons is a hash) should pass if $_ is a key in %.weapons? that seems to be what juerd is saying 13:04
nothingmuch ingy: spoon revision r2148 seems to have broken Kwiki::Installer 13:07
kungfuftr wonders if there is a nice way of doing chained accessors in perl6 13:11
eric256 what do you mean? 13:12
iblech kungfuftr: given $obj { .foo = 3; say .bar; .baz = 42 } (assuming the current ". without explicit invocant" semantics)
kungfuftr given My::Class->new() { .bar("hehe"); .foo("haha"); .baz("hoho"); } ? 13:14
eric256 My::Class->new(:bar<hehe>, :foo<haha>, :baz<hoho>) would also work (i'm just playing with the quoteing ;) 13:15
iblech if you s/My::Class->new/My::Class.new/, then, yes
eric256 that -> to . is going to get me forever....the . to ~ works just fine for me though, ~ is actualy a lot nicer. ;) 13:17
stevan eric256++ # nice game :) 13:19
Juerd Limbic_Region: You probably typed it once in your life at least. 13:20
stevan Limbic_Region++ # for getting eric256 to come over
Juerd: I always do www :)
kungfuftr eric256: does that do modification on the attribute, or actually call the mutator/accessor?
Juerd eric256: Re Any~~Hash, yes, that does %Hash.exists($Any)
eric256 hey juerd...that doesn't seem to be working in Pugs yet then 13:21
iblech kungfuftr: I *assume* the accessors/mutators are called. I'm actually pretty sure of thart
Juerd eric256: t/ :)
eric256 and the givens work. but I currently have to put $_. instead of just . 13:22
Juerd Which is short for: "it would be really nice&great of you if you would want to write a test for that, and commit it under a sensible name somewhere in pugs' t/ directory hierarchy."
eric256 yea i guessed as much...me and when tests seem to be friends already anyhow 13:23
lol
Juerd That's terrific :)
eric256 not sure how to test for the given $object { .method } though....so if someone else would like to right those tests! ;)
Juerd Write a simple class, with one method that returns a true value 13:24
Then given $o { ok(.method, "...") }
Khisanth iblech: btw, mklivecd.pl seems to have slight bug, --help prints the help text twice
Juerd If .method is a parse error, use eval
given class { method foo { 1 } } { ok(.foo, "...") } 13:25
But perhaps that's stretching it too far :)
iblech Khisanth: Ah, thanks... I'm not at my development computer ATM, but will fix it when I get home 13:26
eric256: FWIW, I like that I've to put "$_." in front of it, because then we can leave off the $self in methods
Khisanth at first I thought I was having a memory problem :)
iblech eric256: But Larry's decision of this week is that ".method" always calls $_.method 13:27
(and not $?SELF.method)
eric256 ahh but pugs is still looking for $?SELF ...the error makes much more sense now..
jhorwitz autrijus: ping 13:28
Juerd iblech: Current .default is $_.default 13:29
Oh, you said so yoursellf to
Something's wrong with ths keyboard
iblech ?quit Shutting down svnbot6 for a minute to integrate the karma thing of today
Juerd s/yoursellf to/yourself too/
iblech: Please make it understand all bracketed forms 13:30
iblech: ''++ ""++ <>++ []++ {}++ ()++
iblech: Perhaps even q//++ :)
autrijus jhorwitz: pong
Juerd iblech: Or .++, and find a matching . somewhere on its left 13:31
ajs so what's the karma thing? I've seen people putting plusplus after names, but had no idea why
Juerd ajs: It's hard to explain
autrijus Juerd: as implemented, .default in pugs is $?SELF.default, btw.
Juerd: I forgot what the consensus was.
jhorwitz autrijus: thx for p6-izing my system() change. i often forget that p6 != p5. ;-)
Juerd autrijus: That's currently wrong then 13:32
autrijus jhorwitz: that's fine :)
Juerd ajs: (something I like)++ (something I dislike)--
autrijus Juerd: so what was the resolution vs .default and $.foo ?
Juerd ajs: (something I like)++ # reason for liking it
iblech Juerd: No, the thing I referred to was that the author name is always displayed as "nick++". This was because of a svn checking earlier today
Juerd autrijus: $_.default, $.foo works on $self
iblech: oh
autrijus and how do I access to the implicit self? 13:33
I used $?SELF because I need something like that
iblech autrijus: And &.method always calls $?SELF.method, IIRC
ajs Juerd: oh ok, so it's just a social convention, and has nothing to do with anything automatic?
Juerd autrijus: IIRC, $?SELF.method and ^method
autrijus iblech: you need the ()
Juerd ajs: It's a social convention, that can exist because it's actually counted in some channels
autrijus Juerd: is the ^ accepted?
Juerd ajs: But the counters are meaningless
autrijus: I'm not sure
iblech autrijus: er, yes
Juerd autrijus: I think so
iblech ^ is not in the synopsis, last I checked 13:34
Juerd synopses lag too much :(
iblech yes...
Juerd karma for autrijus
jabbot Juerd: for autrijus has neutral karma
Juerd Either the bot is broken, or the bot is broken.
osfameron karma autrijus 13:35
jabbot osfameron: autrijus has karma of 146
Juerd ORRR... it thinks the whitespace was significant
Which would really suck :)
osfameron (for autrijus)++
karma for autrijus
jabbot osfameron: for autrijus has karma of 1
Juerd karma for autrijus
jabbot Juerd: for autrijus has karma of 1
Juerd Hm
That's un#perlish!
eric256 committed another test ;)
Juerd is used to purl
eric256++
ajs speaking of synopses, I've finished colating everything outstanding for S29 except the string stuff that Rod mentioned having somewhere. I have to check to see where he said that was. Next I need to review how all of these colated things relate 13:36
autrijus ajs++
Juerd: write tests for .foo and ^foo? 13:37
Juerd autrijus: Will do if there are none by tonight. Just got a call and now have to go fix a server
(If someone else writes them, let me know, so we don't do double work)
autrijus eric256: so your
.attack($enemy, $.weapons{$_});
Juerd afk
autrijus needs to say
^attack($enemy, $.weapons{$_});
now 13:38
13:38 nnunley_ is now known as nnunley
autrijus because .attack would mean $choice.attack 13:38
in the given block
I fixed it for you
welcome to the quicksand of perl 6 syntax changes
eric256 lol 13:39
so ^ always means $?SELF. while . can be taken differently depending on its context but probably means $_.
autrijus yeah.
bbiab 13:40
svnbot6 r3151 (eric256++) -- Added test to make sure .method calls default to $_ not $?SELF
r3152 (autrijus++) -- * From Juerd:
r3152 (autrijus++) -- .method # $_.method
r3152 (autrijus++) -- ^method # $?SELF.method
r3152 (autrijus++) -- * Also fix wizard.p6 to reflect this change.
iblech I think I've to get used to this... (liked the . w/o explicit invocant meaning $?SELF.method better)
osfameron ooo, you get a ++ every time you commit !
Juerd osfameron: No, for every line in the commit message. So flood away! 13:42
osfameron heh
Juerd afk # wish me luck
svnbot6 r3153 (iblech++) -- Fixed mklivecd.pl to print the help only once when started with --help. 14:11
eric256 peeks at the Pugs source and then remembers he doesn't speak/read/comprehend haskel and runs off 14:13
14:34 Aankh|Clone is now known as Aankhen``
wolverian eric256: no, .foo() always means $_.foo() 14:50
(by the way, ^foo and $^foo clash somewhat, to me. I wonder if that is a problem.)
pmichaud actually, isn't this still under discussion?
or did that get decided?
eric256 begins to wonder if anything is truly decided yet. ;) 14:51
wolverian pmichaud: well, yes, but if we're going with ^, that's the only useful behaviour. (that was re: autrijus and eric about ^ vs .)
eric256 what would $^ bee?
PerlJam eric256: an error
:-)
eric256 i meant in his $^foo example.
PerlJam eric256: it's one of the implicit parameters
wolverian eric256: @array.sort:{ $^a <=> $^b }; # $^a is @_[0], $^b is @_[1] (alphabetic order) 14:52
eric256 and soo $^foo would be???? what?
lol
PerlJam @reds = map { $^a ~~ 'red' } @colors is the same as if you'd used $_ instead of $^a
eric256 ohh... /me slaps head 14:53
PerlJam wolverian: s/alphabetic/unibetic/
wolverian PerlJam: right. :)
also, is $_ aliased to @_[0], and if so, when?
eric256 you just mean $^{something or other}. it seems $^a and ^a are different enough not to conflict
wolverian I guess. it just seems $.foo might be then a property on $_, and $^foo a property on $?SELF :) 14:54
PerlJam pmichaud: And I think that Larry had a strongish leaning in one direction regarding the meaning of .foo() but I don't remember what that direction was (I think it was a wolverian said though) 14:56
wolverian I think the last message from larry in that thread was that .foo() would always mean $_.foo() and we should find some other character to mean $SELF 14:57
I don't remember him picking one.
pmichaud I don't think one has been picked yet
wolverian that's good.
I don't like ^.
PerlJam I know! Let's use colon! :-) 14:58
or * :)
wolverian or #
PerlJam or `
pmichaud pj: actually, damian proposed colon :-)
pj: but it completely destroyed the option-pair syntax 14:59
wolverian and calling a private method on self would be '::foo'
which is a type
drbean Miyazawa notes pugs and parrot are hooked up at blog.bulknews.net/mt/archives/001696.html 15:00
PerlJam I wonder if we could use make one of these work (to mean $SELF.foo()) : ,foo() ;foo() 'foo() 15:01
pmichaud , is bad
semi is bad
PerlJam They're all bad!
pmichaud $a + ,foo()
or even
[ 1, 2, ,foo(), 3 ]
PerlJam But without going unicode I don't see much good.
pmichaud at some point it just makes better sense to declare invocants as $o and then use $o.foo() 15:02
PerlJam personally, I'm probably going to go with making $o always be $SELF or something $o.foo() isn't a hardship. 15:03
wow
great minds *do* think alike
:-)
pm: besides that's what I tend to do in perl5 now. $o is the object, $c is the class
pmichaud well, larry has proposed o.foo() and c.foo() a couple of times, so $o and $c aren't a stretch 15:04
PerlJam I'm not so sure o.foo() would interpolate in double quotish strings, while $o.foo() certainly would 15:06
Those sigils are quite useful 15:07
pmichaud argggh, luke is misreading my rule also ... :-| 15:08
eric256 could do ..foo() ;) 15:11
pmichaud looks a lot like 1..foo() :-)
PerlJam eric256: that would work too. we'd need a whitespace rule to make sure that it didn't get confused with rnages
pmichaud I dunno if a whitespace rule is needed there, since ..foo() occurs when a term is expected while .. occurs where an operator is expected 15:12
eric256 .-foo ? two character codes give a lot more options .= .- 15:13
pmichaud unfortunately, .. reminds me of the parent directory in filesystem paths. So, how about ./foo() for the invocant? ;-)
Juerd rehi!
eric256 was that a name?
Juerd Has consensus re ^foo and .foo changed since 2 hours ago?
PerlJam eric256: two chars is fine but if we have to shift we might as well just use $o.foo() 15:14
Juerd pmichaud: Not bad. I mean that.
PerlJam so that limits the chars a bit
pmichaud points out that $o.foo() is three chars.
Juerd pmichaud: ./pugs is pugs in the current directory -- classes can be seen as directories
pmichaud: But then ../method must call in the super class. This may just work.
pmichaud I wasn't here two hours ago, what was the consensus on ^foo and .foo ?
Juerd pmichaud: .foo is $_.foo, ^foo is $?SELF.foo 15:15
I promised autrijus to write tests
pmichaud do we run into a conflict with &infix:<^> ?
Juerd But I saw you discussing things alike, so I thought I'd inquire before writing them
PerlJam pmichaud: another whitespace rule :)
Juerd pmichaud: I wouldn't know how
pmichaud i dunno either, I'm just askin :-) 15:16
Juerd PerlJam: termies don't clash with oppies
PerlJam $foo^^bar() looks a wee odd
Juerd PerlJam: That's $foo xor bar anyway.
pmichaud pj: that would be $foo ^^ bar()
PerlJam oh right.
$foo^^^bar() then :) 15:17
Juerd PerlJam: Now $foo^^^bar() looks odd. But how often do you expect that'll happen? :)
Not to mention $foo^^+^^bar()
Which is $foo ^^ +^ ^bar()
I see new opportunities for the obfu people.
And no practical limitation whatsoever. 15:18
PerlJam Juerd: add hypers and we're all line noisy
Juerd Okay, then I now write tests
After updating and recompiling.
PerlJam I think I'll just stick with $o.foo() 15:19
Odin-LAP What was that six-character operator someone used ... [»+^«]?
eric256 wasn't the concern with ^foo that attribute access would then be $^foo which is taken? 15:20
PerlJam eric256: attributes of the current object will still be $.foo and $:foo 15:21
Juerd Odin-LAP: If you count >> as 2 characters, it's more :)
PerlJam Juerd: I call that operator "reduction to madness" 15:22
Odin-LAP Juerd: Well, I don't. ;)
wolverian eric256: my concern is that they look similar, that's all.
eric256 thats confusing. .foo is $_.foo but $.foo is $SELF.foo
Juerd Odin-LAP: I have no idea how to interpret a reducing >><< though.
Odin-LAP: Unless it's used on a list of arrays. Hm, funny.
Still, too confusing :) 15:23
Odin-LAP It's possible ... but a bit mindboggling.
Juerd eric256: Yes, but $ is not .
eric256: :foo is also not $?SELF:foo or $:foo
eric256 is now very confused. lol 15:24
Juerd eric256: $:foo is a private attribute
eric256: $.foo is public
eric256: :foo is a pair
eric256: .foo is $_.foo
Don't look for half consistencies 15:25
Odin-LAP Ooooo. Learning Perl6 will be a true joy for newbies. ;)
eric256 okay.i think i'm gonna need a book ;)
Juerd Hm 15:26
Perhaps it's time for a series of Perl 6 articles
"All uses of ... the colon"
wolverian and .:foo is a private method call
Juerd "All uses of ... the dot"
"All uses of ... the circumflex"
eric256 someone should start a quick reference for Perl6
Juerd eric256: I'll try to whip something up
Odin-LAP Juerd: Didn't you have one at some point?
Juerd eric256: It'll be a hell of a lot more complex than perlcheat though :)
Odin-LAP: Several 15:27
Odin-LAP Heeh. A Perl6 "cheat sheet" would fit on ... say, A0?
Maybe A1. If we're lucky.
pmichaud with small enough type it'll fit anywhere :-)
15:27 knewt_ is now known as knewt
pmichaud I bet I can fit it on a DVD :-) 15:28
Odin-LAP pmichaud: Sure. I'm thinking of something legible, of course.
Juerd Shall we write a quickref set together?
In pugs' svn repo?
wiki-style?
(but plain text!)
All in the form of: 15:29
code example # what it does
like:
.method # shorthand for $_.method
eric256 sounds like a plan juerd 15:34
might need a shortcut reference
pmichaud btw, juerd, that is one mean nasty rule expression you posted yesterday :-) 15:36
eric256 and then a type reference one, and a class reference one....pretty soon you'll need a quick reference quick reference to find the right quick reference. ;)
Juerd pmichaud: It's very straight forward! No embedded code, even! :) 15:37
pmichaud Juerd: well, the problem is that it's not getting past the rule compiler
Juerd pmichaud: It works very well in Perl 5
pmichaud: Ruby even refuses to compile a regex this big :)
pmichaud i'm not sure why yet -- I'm going to have to break it down and take a look at it piecemeal to find out where it's failing
Juerd pmichaud: You can play with matrix.p6 to generate your own 15:38
pmichaud: Look at the outer build_re call. Isolate that and call it once instead of 25 times, and you get one part instead of 25... :)
pmichaud: Reduce the todo argument, and you get a much shorter (exponentially) rule as a result 15:39
pmichaud: This will allow you to build up to the critical point
It's probably smarter to use matrix.p6 to generate the rule though
That'll save you hours :)
Eh
matrix.p5
PerlJam What does matrix.p[56] do exactly? 15:40
(besides generate a pge-breaking re)
Juerd Use the rule against words, to find matching candidates 15:43
Then sort those in order of score, based on a simple scoring table
There's now pugs/docs/quickref 15:59
svnbot6 r3154 (Juerd++) -- First steps for a set of cheat sheets. It's a wiki, people, so contribute! :)
r3155 (Juerd++) -- First document, far from finished. Feel free to add or correct stuff.
broquaint Hey Corion. 16:10
Corion Hi broquaint
broquaint Have you successfully adopted S29? 16:11
Corion broquaint: Luckily not :) ajs has adopted it :) 16:12
(or that's what I think we ended up last night)
... with ...
I get to snipe from the sides
broquaint Nice one :)
BTW, has caller() et al been implemented yet? 16:13
Corion broquaint: Yep. $?CALLER
broquaint: I think under t/magicals
broquaint But the function caller() hasn't, right?
kolibrie don't think caller() is around yet 16:14
Corion Will there be the function caller() ? I thought that $?CALLER and $?CALLER::* and $CALLER_CONTINUATION took care of it all? 16:15
(I may be wrong)
broquaint There should be.
Just for a simpler way to access contextual information about the sub and its callers.
Corion broquaint: Hmmm - aaah - yeah, a dynamic way to walk up the caller chain would be useful, true. 16:16
kolibrie S06 "The caller function"
Corion broquaint: You could implement it in Perl6 with a loop that walks up $?CALLER:: :)
(or steal the code implementing it)
broquaint I'd certainly abuse ... er ... use it :)
Corion ... but I welcome tests for it :)
broquaint Do we have any perl6 functions implemented in perl6? 16:17
Corion Having something like caller() is important. Even JS has it! :)
broquaint Most of the 'functions' are in Eval.hs, right?
Corion broquaint: Unfortunately not. Well, one half-implemented "function" - the split-on-whitespace behaviour of %hash<...> is implemented as AST
(which was taken from dumping a Perl6 expression)
(or was it the split-on-whitespace behaviour of @array = <...> ?) 16:18
broquaint looks into implementing caller() ... 16:19
Corion Yay! ;)
broquaint Although there is a test for it so maybe someone got there ahead of me ... 16:22
eric256 is there a timeline/schedule somehwere? i'm curious about when we will see class inheritance in Pugs... ;)
integral "soon"...
eric256 soon is always good 16:23
Corion eric256: There is the Pugs development plan, detailed in one of autrijus' early journals
... which explains the version number convergence, together with the features
eric256 wanders off to search the journal then
integral finds it amazing that he can go for 3 hours and find 10 commits
svnbot6 r3156 (ninereasons++) -- give clearer mesg to proclaim() in 'isnt' 16:24
Corion ... no timing is mentioned. But as #perl6 warps space-time, you can stick around and wait :)
eric256 what is zip? in the quick ref you have Y # zip ?? 16:25
integral take N arrays. Take the first element from array 1, ..., take the first element from array N, take the 2nd element from array 1, ... etc
broquaint It "zips" lists together i.e (1,2,3) Y (a,b,c) becomes (1,a,2,b,c,3) 16:26
integral hmm, is Y associative?
broquaint How do you mean "associative"?
integral @a Y @b Y @c == zip(@a, @b, @c) or zip(zip(@a,@b), @c); #or the same but for Right assoc 16:27
Juerd integral: It's chaining
integral: former
integral ah, neato :-)
Juerd integral: ==, eq, <, etcetera are chaining too now.
3 < $x < 5
broquaint Everything will chain and it will be good! 16:28
Corion r3155 - datenzoo.de/pugs/win2k.html - 4909 ok, 20 failed, 695 todo # win2k, so something is weird with the newly introduced fails
broquaint Except, the things that don't.
integral yeah, I knew about hose ones. Chainign is a really neat feature
I suppose = could be chaining too: $x = 5 = $y; :-P
Juerd integral: What would that mean?
integral it sets all the non-constant things to the single constant thing, or the right-most non-constant thing 16:29
Juerd Then simple $foo = $bar fails
No, let's not change =
eric256 chaining sounds cool
integral no, that's covered by the bit after the comma juerd
Corion Can somebody confirm the unexpected success in t/statements/given.t (last test) ? 16:30
broquaint Sure, just let recompile ...
Juerd integral: Oh, right, but still, = can be used as chainingly as you want already 16:31
= being right associative and returning an lvalue are both very useful.
integral sure, but you can't say: 3 = $foo;
Juerd You shouldn't want to.
integral Just think of the obfuscation value! :-P
sheesh, it's not a serious idea
broquaint People need to be careful with casedness in filenames as Motd.pm just collided with MOTD.pm on this windows box with case-insensitive FS wickedness. 16:32
Juerd Can't you even mount something case sensitively when you want to in win32? 16:33
broquaint I believe so, but I'm no windows guru.
Corion Juerd: No. Win32 / FAT32 / NTFS is case-preserving but case-insensitive
broquaint points to Corion's reply ^ 16:34
Juerd Corion: Yes, but if you mount any of those in linux, the result is case INsensitive
Corion: Can't windows mount ext2 or another more sane fs?
Corion Juerd: Not natively, but there are drivers for that.
eric256 corion that test fails here (rev3135)
Juerd It couldn't do this in 2000 when I stopped using it
Wonderful, how Microsoft just won't cooperator or invent 16:35
Corion Juerd: But I'd expect the drivers to map to case-preserving / case-insensitive as well, that is, to refuse to create/open foo.pl if there is foo.PL
(why should they cooperate with a hippie OS?)
Juerd Because it's the only way to get rid of the hippies. 16:36
It's in MS' best interest if they would support ext2. It would cause people to dual boot more often than they switch completely
Corion Juerd: What should they dual boot to? (you seem to be missing the point - anyone who dual-boots is already lost anyway) 16:37
Juerd No
They're not lost 16:38
Well, idioligically, they're lost
But not commercially.
Corion eric256: Weird ... I'll update and rebuild then - what platform do you run? With parrot or without? 16:39
eric256 WinXP Pro no parrot ....hmmmm i've checked out rev3135, but it appears to be running 1241 agian....arg. i'm running an old one...arg 16:40
Juerd 1241? That's prehistoric.
eric256 no kidding...and i don't know where it came from, yestarday i was running on a recent build..i've got something screwy going on
Corion eric256: No, that's the svk info 16:41
The svk and svn rXXXX are wildly out of sync
(and currently, the svk info gets picked up while the svn info only displays in Test::TAP::HTMLMatrix I think 16:42
eric256 hmmm...so how can i make sure that the executable is the newest version?
Corion svn up && nmake clean && perl Makefile.PL && nmake 16:43
:)
eric256 well i just set out on a holy crusade to delete all pugs.exe's (of which there were several)...now lets get this right...rebuilding.
nmake clean? would that have been screwing me up? because i've never done that. ;) 16:44
Corion eric256: It should wipe all the leftovers. But if you need a make clean, most likely deleting src/ is better anyway.
(and then check out again)
eric256 corion it passes that test now... looks like someone fixed it ;) 16:47
Corion eric256: Heh ;) 16:48
Juerd just discovered that he does want a symbolic alias for 'scalar' 16:49
Is there a special reason we can't just use $ for that?
eric256 shall i remove the :todo<bug> on that test then? since they seem to have fixed it 16:50
Juerd eric256: If something's fixed, it's no longer todo :)
Corion eric256: Yes please 16:51
eric256 done...there is a test being run....but i don't see it in the code...and it has no label 16:52
Corion A heisentest ?
eric256 ehh? 16:53
Corion A test that gets run only if you don't look at it :) 16:54
eric256 lol
hmm... test passes and yet my game code still failes.... 16:57
svnbot6 r3157 (eric256++) -- un-todod a test. '.' shortcut in given works now
broquaint The last test of t/statements/given.t succeeds here, Corion, although a bunch of others fail. 17:00
Corion broquaint: Yeah. I wonder what's up with the breakage, as "svn blame" didn't tell (me)
broquaint Although all the failed tests are marked as TODO (so why aren't they called as TODO tests?), so I guess that makes sense. 17:02
eric256 if <code>$_.pass<a> = 1</code>works inside a given then shouldn't <code>.pass<a> = 1</code> work as well? 17:03
hmmm code tags arn't realy usefull here. lol
pmichaud only if $_.pass is rw, perhaps?
eric256 it is rw. 17:04
looks like every class definition is creating a test...? 17:06
lol. i'd called my temp attribute pass...which i'm guessing is a test sub? it was in double qoutes and getting interpolated... 17:07
svnbot6 r3158 (Stevan++) -- Perl::MetaModel - added document about Meta-Meta-Classes in an attempt to clarify the subject for myself, ... and purge these demons from my mindsvn status! :)
PerlJam pm: PGE doesn't yet do <<alpha>+<digit>> does it? 17:09
pmichaud no, not yet, and based on Larry's last post it may end up being <+alpha+digit>
(at least I hope it is)
PerlJam Excellent.
pmichaud but you can always do \w for now if you don't mind underscores :)
PerlJam Is the leading <[-+]> required? 17:10
pmichaud yes, except with <[xyz]>
PerlJam so all character classy things start with <[, <- or <+
pmichaud well, yes and now
now
no 17:11
(argh)
<alpha> will still exist, but it's a capturing subrule
so <alpha>+ can be a bit expensive on the captures :)
PerlJam right, the CC is only on indirect
pmichaud <+alpha> will be calling a subrule as if it's a CC 17:12
PerlJam Hrm.
pmichaud and for things like alpha, digit, upper, lower, etc., PGE can optimize them whenever there's no overriding subrule in the scope
PerlJam How does that play with rule foo { a*b } ... m/<+alpha+foo>/ ?
pmichaud see Larry's post: "But then it's not entirely clear how character class set theory works. Another thing we have to work out." 17:13
also: 17:14
cognominal how do I access the trait of a function?
pmichaud Hmm, well, positive matches can be defined to traverse whatever the
longest sequence matched is, even if it's actually multiple characters
by some reckoning or other.
(from Larry's post)
cognominal do you know the API to get at the value of a trait? 17:19
svnbot6 r3159 (eric256++) -- added test for .attribute access to hash
eric256 wanders off to find food 17:21
Limbic_Region needs to put his computer together as he is getting highspeed internet tomorrow (along with VoIP phone and digital cable) 17:45
svnbot6 r3160 (ninereasons++) -- add highlighting for ATTN FIXME, etc, in POD and comments.
Limbic_Region (consolidated services and bills) plus plus
cognominal I ask my question again because I have not yet found an answer in the synopsis: is there an API for trait/property getter/setter? 17:52
stevan autrijus: if you are around,.. I found (what I hope is) a small bug
pugs -e 'class Foo::Bar { method baz returns Foo::Bar { Foo::Bar.new() } }'
it seems to choke on the C<returns Foo::Bar> part 17:53
ajs I don't think it groks the :: notation there. Make the class Foo, and it works 17:55
stevan ajs: yes, but I need the :: notation :)
ajs cognominal: not sure
stevan: oh, well then, that's broken :-) 17:56
cognominal: I never fully grokked traits in A/S12... 17:58
cognominal sent a mail to p6l ml... 18:00
ajs cognominal: Ok, here's my theory: A trait is a role, applied by the use of either a property or trait verb. Thus, you should be able to do normal "role stuff" to sense its taint 18:01
cognominal: that includes "$obj ~~ sometrait" 18:02
cognominal: no, I'm wrong. That includes "?($obj does sometrait)
"
I think... ask p6l ;-) 18:03
Ah, but your p6l mail is a different question. You can apply traits without an eval. Just remove the eval and quotes from your example and it should work fine. 18:05
svnbot6 r3161 (Juerd++) -- quickref/op: near completion; please sanity-check
revdiablo anybody have any idea why examples/phonewords.p6 is so slow? it takes 25 minutes to complete, compared to 5 seconds in perl5... is that just standard pugs slowness, or is something weird going on? 18:09
Juerd revdiablo: Because pugs is just very slow, and perl 5 is just very fast 18:10
autrijus that's right!
and if you write .{} codegen for parrot it will be faster :)
Juerd revdiablo: I have a script that finishes within a second in perl 5, but takes half an hour in pugs. It's something thatfor now, we must learn to live with
svnbot6 r3162 (ninereasons++) -- add highlighting for '=for COMMENT' region
r3163 (autrijus++) -- * wikification! :)
autrijus Juerd++ # damn good idea, that quickref 18:11
Juerd Thank you
Not entirely my idea
Well, not today
I had parts of op around already though
revdiablo Juerd: yeah, what I figured. but thought I'd check to make sure. maybe there's something that sticks out as slowing it down a lot 18:12
autrijus Juerd: I kwidified it; the good thing is that it still needs no escaping :) 18:13
the indent makes it easier on my eyes
Juerd I'm not entirely happy with that
But okay.
autrijus thanks for your understanding. I don't quite care about how to mark the headings... =head2 NAME or [NAME] also works for me 18:14
I just want it somewhat indented :)
Juerd That's the thing that bothers me
autrijus how does it bother you?
Juerd But now that I see this, I do find it easier to read. 18:15
autrijus oh. :)
Juerd It bothers me because it eats away 4 characters of my terminal
I'm un-kwidding it though
autrijus that's fine
svnbot6 r3164 (autrijus++) -- * oops, leftover
Juerd The risk of non-verbatim paragraphs is too high
autrijus just kill the "== "
stevan: fixing your bug 18:16
ajs I'm concerned about type "Char"... I think type conversion overhead is going to be stunningly awful 18:17
autrijus where is this "Char" type? 18:23
svnbot6 r3165 (autrijus++) -- * UI cleanup on wizard.p6
r3166 (Juerd++) -- unkwid (s/^== //, .=lcfirst for @headings)
r3167 (autrijus++) -- * Stevan reported that `returns Foo::Bar` didn't work. Fixed.
r3168 (ninereasons++) -- better regexp for '=for COMMENT'
r3169 (Juerd++) -- add reduce meta-op to quickref/op
r3170 (Juerd++) -- update README
gaal is $_ supposed to be ro in map? 18:26
./pugs -e 'my @a = <123>;my @b = map {$_+=1} @a'
Error: Can't modify constant item: VNum 124.0
(funny error message, too.)
i meant: @a = <1 2 3> :-) 18:27
eric256 you can't modify the original @a array so i would assume it is r and not rw 18:30
pugs -e "my @a = <1 2 3>;my @b = map {$_+1} @a;print @b.perl"
behaves well
gaal compare with for (@a) { $_ ++ }
"autoextracted" $_ in for is rw by default. 18:31
i think it sohuld be in map too
or at least, is copy.
there has to be a nice way to avoid the map { s/// } p5 pitfall... 18:32
eric256 hmm . dunno. when would you use map like that though? i mean most times you want to modify wouldn't you use for instead?
autrijus sides with eric256 :)
gaal is copy at least could be nice. if there's a way to make that COW... that's grand. 18:33
autrijus I've just heard that in parrot, cloning a string is guaranteed a O(1) operation :) 18:34
eric256 i see. copy would be slow but copy on write would only slow it down if they actualy modify it? what was the map { s///} pitfall you where refering too
copy can take longer than O(1)? 18:35
gaal @basenames = { s,.*/,, } @pathnames
svnbot6 r3171 (ninereasons++) -- clear screen
gaal that's a bug in p5
because you wanted to return the modified $_, not the return value of the s///, whcih is 1 or 0 18:36
ajs hmmm... the more I think about it, the more I like the idea of a gcc back-end ;) 18:39
ninereasons (although karma (both + and -) messes with my head and I try to ignore it) someone might want to know that svnbot is not granting karma as she seems to think she is
eric256 that isn't a bug. that is just a bad side effect of s/// returning the number of values replaced. 18:40
gaal it's not a bug in perl5, of course. it's a user pitfall.
ah, punctuation. above i meant to write: "that's a bug, in p5" :) 18:43
eric256 either way that is a problem you have with the semantics of s/// not map, unless you want map to pass in a copy and place the final value of that copy in the new array, which i don't think is going to happen. ;) 18:45
svnbot6 r3172 (gaal++) -- cleanup temporary files in tests; add some more nonces
gaal assuming for a moment that s/// returns the substituted value, why not? 18:49
eric256 why not what? if s/// returns the substituted value then map does what you want now. i think we lost each other somewhere 18:50
gaal you can't s/// on $_ now. 18:53
(in a map that is.) 18:54
svnbot6 r3173 (ninereasons++) -- sp 18:56
eric256 ohh. gotcha. i was confused because that is unrelated to your s/// 'bug' in p5....lines got crossed 18:58
Juerd gaal: I have asked for non-mutating s/// several times. Never really got a useful answer. 19:01
gaal btw: map { s/// } is not something i advocate or anything. it's just that from observing people for a while i know it does tend to happen a lot. s/// is very often better done as m// anyway.
svnbot6 r3174 (iblech++) -- * EOL at EOF added to unlink.t.
r3174 (iblech++) -- * Usual svn properties added to docs/quickref, ext/Tree, and examples/.
Juerd gaal: s/// has little to do with boolean stuff, so I'd say break with ~~, and make s a method, from which flow . for non-mutating and .= for mutating
eric256 well the current ro $_ makes s/// an error. ;) 19:02
gaal Juerd, what do you think about 'is copy' map?
Juerd eric256: map -> $_ is rw { ... }, @foo
gaal: possible but inefficient.
eric256 is map functioning correctly in Pugs? and is it suppose to flatten to a string in some cases? 19:03
Juerd Does anyone know what the correct behaviour is? Are there tests?
Methods and functions need specification badly
gaal hmm, "old" note by larry on this. perlmonks.org/index.pl?node_id=435596 19:04
i don't understand the mechanism of mutator methods mentioned there though. 19:05
eric256 me likes the ability to do (1..10).map:{$_+1}.join(",");
now that is OO DWIM goodness
gaal (it's a nice perlmonks thread btw, if you don't know it.)
imagining it would work, wouldn't (1..10).map:{$_++}.join(",") look as natural? 19:07
cognominal does someone remember how to get french quotes in emacs?
gaal (is it ruby that doesn't have ++ for ideological reasons?)
cog, ^k >> in vi, sorry :/ 19:08
Limbic_Region stevan ping 19:09
cognominal arf, that was ~< in iso-accent-mode. 19:12
gaal can't 'given' be an rvalue? $res = given $x { ... } 19:15
PerlJam gaal: it should be able to function as an rvalue. 19:16
gaal this parsefails. ./pugs -e 'my $x = given 1 { when /1/ { "asdfadsf" }; when 2 {"aa"}; }' 19:17
PerlJam oh, pugs has the correct behavior. 19:19
from S04 "Statement-level constructs may start only where the parser is expecting the start of a statement. To embed a statement in an expression you must use something like do {...} or try {...}"
and given is a statement-level construct
gaal hmm.
thanks,
i'll go with that.
do isn't implemented yet though. :) 19:20
and rvalue try doesn't seem to work.
Juerd sighs at perlmonks 19:21
I made the same mistake the what, third time now?
autrijus Juerd: hm?
Limbic_Region which mistake is that?
Juerd assumed open mindedness
I assumed that a bare idea, without implementation details, could lead to a useful discussion and perhaps implementation
PerlJam Juerd: surely the opposite assumption is the safe one for humans in general
(and perlmonks is a nice collection of general humans) 19:22
Juerd As L~R said in his pugs advocacy, planting the seed can often be fruitful
But not on PM, not when you "touch" the site. Not when you suggest a feature there. It just can't happen. PM is too much like p5p in that respect.
Limbic_Region Juerd - a term I have used with management is "jedi mind trick", making someone believe what you want to happen is their idea 19:23
Juerd I'm not good at mind tricks
I absolutely hate social politics, especially when people activate unnecessary force fields 19:24
PerlJam Limbic_Region++ that's exactly how it has to happen IMHO
svnbot6 r3175 (ninereasons++) -- move motd
Juerd p6l is a positive contrast, as is this channel 19:25
Here, an idea can enter at any time, and it is discussed and often quickly specced and sometimes even implemented
Limbic_Region mostly agrees with Juerd's sentiment
WRT site changes at the Monastery 19:26
PerlJam Juerd: time to build perl6monks where discussions like that can happen! ;)
Juerd I'm too upset right now (upset because I was wrong believing things could have had changed), so I'm going to take a walk
afk
nmcfarl /msg NickServ IDENTIFY nmcfarl 19:32
eric256 opps 19:33
you might want to change that password now
nmcfarl yeppers
it needed to go any how
gaal has there been a decision on how qx// is spelled now? 19:36
19:38 broquaint_ is now known as broquaint
svnbot6 r3176 (iblech++) -- Usual svn properties added to examples/advocacy, too. 19:38
eric256 curses at his keyboard...then realizes he set a book on a second keyboard attatched to the same system....teach me to screw around in the future 19:40
gaal ouch! getting many undef references (s2gb5_slow) with a simple pugscc attempt on linux. parrot-0.2.0, but i'm actually using the pugs backend.
Corion gaal: I think it's rx:P5// 19:41
err - I was thinking of qr// obviously
It's still spelled system<> :)
gaal :)
eval_perl5 ( use Test::Output; output_from($cmd) }, yeah. 19:43
oh, the parrot backend works ok. 19:44
Limbic_Region wonders what preverse thing gaal is using the backend of a parrot for 19:45
Limbic_Region ducks
gaal wow, i've been outsillied!
this calls for a drink :)
Limbic_Region it is Friday afterall 19:46
gaal yeah
eric256 so by compliing parrot i get rules support?
kolibrie yeppers 19:47
eric256 but there is no Makefile.PL .... /me searches for readme
Limbic_Region with a couple of options - linking against the lib or the external executable
gaal perl Configure.pl 19:48
Limbic_Region eric256 - perl Configure.pl <options>; make; make test; etc
gaal was faster on the keys
gaal if rules work well i'll take a stab at porting this, heh heh. www.forum2.org/gaal/lj/fool-pretty.pl
fagzal hail
Limbic_Region salutations
eric256 what options would i want?
Corion Hmmm - taking a script from stdin seems broken currently... 19:49
kolibrie eric256: I didn't use any options, and mine works great...
fagzal Erm... anybody could tell me where I can find some info on "what's going to be implemented in Pugs approximately when" :-) Actually I tried my $a = "a" _ "b"; and it did not work out ;-)
gaal Corion, as does interactive mode on windows
it's because of that terminal thing, remember?
Corion gaal: Hmmm. Why that?
What terminal thing?
eric256 string concatenation is now ~ not _
Limbic_Region eric256 - you might want some of the optimization options but for now just get it to work 19:50
was string concatenation ever _ ?
fagzal eric256: ouch... I was looking at some old docs then :-/ Sorry 'bout that.
eric256 it was in the docs. it still is in one set of docs
Limbic_Region fagzal - see dev.perl.org/perl6
gaal Corion: hIsTerminalDevice lies, on msys at least.
Limbic_Region or better yet - install Perl6::Bible
Corion gaal: Ah. Ugh.
gaal Corion, so Main.hs:88 doesn't do the right thing. 19:51
eric256 limbic those installed ones wont stay up to date though
Corion So we'll have to look if we can ask the Win32 API instead.
Limbic_Region eri256 - assuming you keep them up to date they will
gaal personlally i'm not sure interactive mode shoudl start w/o params.... but that's just me.
fagzal Thanx! I'll RTFM now :-)
Limbic_Region ingy makes a new release everytime one changes (or so he lead me to believe)
eric256 wonders if all the warnings going by should worry him.. 19:52
gaal brb
Corion gaal: But GHCi does that, so Pugs does too.
gaal so offer a pugsi
brb really
eric256 beigns make test and sees yet more warnings... 19:53
Limbic_Region eric256 - I wouldn't be worried too much - parrot compiles quite noisily
hmmm - I don't remember make test being to squaky though
eric256 soo its done....but i don't see an exe or anything...how do i test it? 19:54
Corion gaal: Pugsi sounds cute
Limbic_Region eric256 - you compiled Parrot on windows using which compiler?
svnbot6 r3177 (iblech++) -- Added a (sucessful :)) test testing state() inside the substitution part of 19:55
r3177 (iblech++) -- regular expressions.
Limbic_Region only ever did Parrot on Cygwin for Win32
eric256 by compiled....i just did perl Config.pl....then nmake, then nmake test
gaal you can call it pigs too, if you like.
eric256 i haven't done anything else besides download it....and i have VC++ installed.
stevan autrijus++ # for fixing my returns Foo::Bar problem :) 19:56
Limbic_Region *shrug* - I would have expected a parrot.exe to end up in the directory you did the nmake in
stevan Limbic_Region: pong
gaal eric, cool! it didn't workforme earlier today :(
eric256 not sure thats a reason to celebrate! lol.
gaal well i didn't use msvc 19:57
eric256 just once i would like these things to say "Completed succefully" or even something spelled right
Limbic_Region gaal - mingw?
eric256 i dunno if i used it or not. it never asked
autrijus pugs> say do { if 1 { 3 } else { 4 } }
3
gaal l~r, yes, also tried multi
autrijus (r3178)
gaal s/multi/cygwin/
eric, it figured it all out from whatever perl you used.
eric256 well i have activestate perl...so 19:58
ninereasons compilation seems to be using src/Pugs/pugs_version.h; which hardcodes the svn version to 1241
autrijus ninereasons: that manes util/version_h.pl isn't doing what it should be doing 19:59
ninereasons ah
Limbic_Region gaal - I can tell you 101 things that could go wrong with compiling Parrot on Cygwin
autrijus try rm pugs_version.h and make again
Limbic_Region not sure if anyone picked up the torch as resident Parrot on Cygwin weenie after I dropped it late November
ninereasons will do, autrijus ; thanks 20:00
svnbot6 r3178 (autrijus++) -- * do {...} literal.
eric256 guesses he doesn't need rules right now. ;)
Limbic_Region gaal - what were the Cygwin problems (if you remember) 20:01
gaal L~R, maybe you should tell them to p6-c, so some <ESC>dd aah, you said *parrot*.
undefined rerferences to _assert and stuff.
s/_/__/; # make that extra magic
eric, try make install! :) 20:02
autrijus eric256: parrot.exe needs to be in your path
that's all
ninereasons do wee need to say 'PUGS_EMBED=parrot' anymore, now that parrot is more or less required for a fully working pugs ? 20:06
svnbot6 r3179 (Stevan++) -- Set - making things work again, slight issue with some methods
eric256 but i need to have a parrot.exe
20:09 Qiang_ is now known as Qiang
svnbot6 r3180 (iblech++) -- Added some more submethod tests. 20:12
r3181 (ninereasons++) -- oops (take out a debug line)
Limbic_Region oh - so every commit results in a karma upvote? 20:13
where is the karma whore bot?
ninereasons it doesn't seem to work, though, Limbic_Region
eric256 is the svnbot programmed in perl6?
ninereasons or someone has mercifully exempted me from being karmalized 20:14
Limbic_Region perlbot karma for ninereaons
perlbot ninereaons doesn't have any karma
Limbic_Region perlbot karma for ninereasons 20:15
perlbot Karma for ninereasons: 2
Corion Does anyone know where in Perl5 the -X tests are implemented?
eric256 read taht caramelized. :)
perlbot karma for me
perlbot Karma for me: -1
Limbic_Region ninereasons - it is probably because of the parens
Corion (and/or what the name of -T is)
eric256 perlbot karma for eric256
perlbot Karma for eric256: 2
ninereasons prob'ly so, Limbic_Region
eric256 eric256--
perlbot karma for eric256
perlbot Karma for eric256: 1
eric256 ohh ;(
lol
eric256++++ 20:16
perlbot What kind of idiot karmas himself? Your kind of idiot!
Limbic_Region when something with a space needs to be karma'd you put it in inside the parens and the upvotes on the outside
eric256 lol
eric256++
perlbot What kind of idiot karmas himself? Your kind of idiot!
gaal Corion: op/filetest.t
eric256 hehe
Limbic_Region that will result in a karma loss too
perlbot karma for eric256
perlbot Karma for eric256: -1
eric256 i'm not too concerned
ajs hmmm ....
ajs--
Limbic_Region prefers to have neutral karma
ajs doh!
Corion gaal: That are the tests. I need the C side.,
ninereasons karma does funny things to my mind
jabbot ninereasons: does funny things to my mind has neutral karma
ninereasons he 20:17
Corion gaal: But "filetest" was the good word to search on, thank!
s
svnbot6 r3182 (iblech++) -- Added a small additional test to t/builtins/fail.t. 20:19
ninereasons it's like rep in Perl Monks. It made me worry about things that don't matter (e.g. that my rep is too low) and in the end demotivated me from participation. I must be a psycho.
eric256 i like sharing my game...but someone else changed it..lol. i'm not sure i like that..... 20:23
ninereasons that slightest whiff of anarchy, eric256 20:25
eric256 it was you! lol
ninereasons :-)
Limbic_Region eric256 - I really like the changes people have made to my examples 20:26
eric256 at least make clear screen a sub instead of putting that ugly $OS stuff around. ;)
i'm still learning to share Limbic...its scary coming back to your code after a couple hours and having it be different
ninereasons as you wish, eric256 ; it will take a minute 20:27
autrijus hey, anyone have parrot 0.2.0-release? 20:28
I'd like to know if rules still works with the latest pugs
eric256 i just did it down' worry
actualy i must admit i kind of like the way it works
frankly i was just shocked to see a system command in there and it took me a minute to realize what had happened 20:29
Limbic_Region oh wow
it is officially the weekend 20:30
eric256 not here
Limbic_Region and I might even be connected to the net sometime tomorrow
PerlJam autrijus: works in what fashion? I just did a little test and it doesn't appear to die :)
svnbot6 r3183 (autrijus++) -- * trans_charset.
r3184 (ninereasons++) -- attempt to unless the karma
r3185 (ninereasons++) -- put cls in a sub
Limbic_Region if not - see everyone on Monday
TTFN
autrijus PerlJam: ok, that's good enough
PerlJam++
Limbic_Region: see ya
ninereasons s/unless/unleash/
PerlJam well, that's was with r3184 ... no telling *what* happens with 3185 ;-) 20:31
autrijus heh. the change that matters is r3183
eric256 whats trans_charset? 20:32
autrijus eric256: it's a parrot op code that translates a string from a character set to another
I've been hacking proper utf8 support to parrot 20:33
eric256 ahh
can i put my classes in a seperate file? do i need package? does it work in pugs yet?
autrijus you can put it in separate files, yes. 20:34
currently there's a small bug in that classes defined in external files must be addressed with ::
like
use Foo;
::Foo.new()
eric256 ohhh.
autrijus this is expected to be fixed soon
eric256 that explains it
ninereasons so much perl6 goodness does work, though.
autrijus (but that syntax is still legal)
svnbot6 r3186 (iblech++) -- Cleaned up Net::IRC a bit and added some IRC commands to it. 20:35
iblech ninereasons: I think we should test what the karma bot accepts... 20:36
ninereasons yes
iblech r1234, somedev++: description
karma somedev
jabbot iblech: somedev has neutral karma
iblech r1234, somedev++ description
karma somedev
jabbot iblech: somedev has karma of 1
iblech r1234, somedev++ --> description
ninereasons yay!
iblech karma somedev
jabbot iblech: somedev has karma of 2
iblech ok :)
I'll fix svnbot
?quit Restarting svnbot 20:37
PerlJam autrijus
<+uri> PerlJam: i doubt energy will be wasted on both. the goal of pugs is to bootstrap p6 compiling itself. once that is done, pugs should fade away 20:38
what say you to that?
autrijus that is correct.
pugs either _becomes_ perl6
or is replaced by real perl6.
PerlJam Then it's a good thing I didn't bet uri the other way ;-)
autrijus either way, pugs as we know it will fade away. 20:39
that is why I can go with this crazy version scheme.
you don't think we'll stay with 6.283185.99 for long, right?
PerlJam well, "long" is the word I'm having problems with. Long relative to what? 20:40
eric256 can you do @array.any.pick() by any chance? ;)
autrijus isn't it @array.pick?
(it's already there) 20:41
PerlJam any isn't a method of any object is it/
autrijus @array.any.pick, incidentally, works too.
ninereasons see examples/hangman.p6 eric256
autrijus but only because that gets invoked as any(@array).pick.
iblech hm... but [1,2,3].any.pick, does not
eric256 .pick on a regular array picks what? 20:42
autrijus iblech: it works for me
iblech a random element
autrijus it's the same as any([1,2,3]).pick
because any is not a method on array
pugs> [1,2,3].pick
2
but .pick is.
iblech yes, but [1,2,3].any.pick always returns "(1, 2, 3)" here
autrijus and that is as expected
because any([1,2,3]) is just [1,2,3].
svnbot6 r3187, iblech++ --> Fixed svnbot to output the committer's names in a karma-bot compatible way.
iblech oh yes
of course
autrijus :) 20:43
ew.
can I change --> to | ?
(in svnbot)
iblech of course. :)
?quit Shutting down svnbot
autrijus done 20:44
iblech ok
ajs autrijus: any([1,2,3]) is [1,2,3] because the array is not flattened? 20:51
svnbot6 r3189, ninereasons++ | any(@ar).pick to @ar.pick
ninereasons although I deleted src/Pugs/pugs_version.h and ran make again, my final pugs is still Version: 6.2.3 (r1241)
autrijus ajs: array refs are never flatteneed
ninereasons: does pugs_version.h say 1241 still? 20:52
ninereasons yep
ajs Ok thanks.... though is a junction of one not a junction? Could that ever end up causing problems?
ninereasons I wonder where it got that from. 20:53
PerlJam ninereasons: I just noticed the same thing 20:54
ajs: a junction is a junction is a junction.
ninereasons must depart && /quit bye 20:55
ajs PerlJam, then isn't any([1,2,3]) actually NOT [1,2,3] ? 20:56
PerlJam depends on the flattening (or lack there of)
svnbot6 r3190, iblech++ | Delete empty directory t/kwid as its contents have moved (in r374!).
r3191, Stevan++ | Test - adding tests for isa_ok() with classes
ajs PerlJam: see context above. autrijus was saying that it doesn't flatten therefore, any([1,2,3]) is same as [1,2,3], but I'm saying that a single array ref is not the same as the any-junction of that one array ref... I'm not 100% sure about that, though 20:58
PerlJam ajs: no, they are not the same. One is a Junction, the other is an Array 21:00
eric256 cool.ios there a descructive pick? or a .shuffle.pop ;)
Corion Hah. GetFileType in the Win32 API should be a reliable way to implement -t 21:02
eric256 ajs i think he measn like $this = 1; any($this) = 1; $this = [1,2,3]; any($this) = [1,2,3]; because $this is realy a ref to [1,2,3]
no shuffle.pop :( .... can we program new .method's for arrays and stuff? 21:05
PerlJam does anyone happen to know the URL for the dbdi repos on svn.perl.org? 21:06
autrijus PerlJam: svn.perl.org/parrot-modules/dbdi/ 21:14
it's empty.
PerlJam thanks. 21:17
cognominal I thought that synopsis where the reference but I realize that for object the synopsis scratches only the surface. 21:23
for my question on trait/properties setters/getters, indeed one must read A12 to get an answer 21:25
Corion Anyone know if there is a way to get from a Haskell (file) Handle to the file handle number? 21:31
I want/need to pass it on to a Win32 API function, that wants an Int
21:32 wilx` is now known as wilx
autrijus System.Posix.IO.handleToFd 21:33
not sure if you have it
stevan autrijus: does C<has Type $.attr;> work yet? 21:37
svnbot6 r3192, Stevan++ | Set - adding some tests for class inserting and removing
autrijus stevan: it's parsed
but has no effect yet
stevan ok
thats explains it :)
I was trying to do $.attr.method() and it was complaining 21:38
hmm, any thoughts on how to get around that?
autrijus is it tested? 21:39
stevan i think so, but if not, i will write one
especially now that I know what is happening and I am not just going crazy :) 21:40
Juerd re 21:41
Driving in reverse sucks
Driving in reverse in the dark sucks
Driving a kilometer in reverse in the dark sucks
Or, put differently: fallen trees suck.
Corion Too bad - System.Posix.IO.handleToFd doesn't exist on Win32 21:44
eric256 wonders how a fallen tree translates into driving a mile in reverse...why didn't you jsut turn around? 21:46
autrijus stevan: got a test?
stevan autrijus: working on it 21:47
several of the tests in t/oo/attributes.t work, but eval_ok() does not :) 21:48
lives_ok {} gives better diagnostics as well 21:50
so I am fixing them first
then I will add the test
wow 21:56
a lot of these tests work
just eval_ok sucks!
autrijus: test added (t/oo/attributes/methods_on_typed_attrs.t) 22:03
svnbot6 r3193, Stevan++ | adding test for methods called from typed attributes
stevan I am going back to cleaning up t/oo/attributes.t :)
submethod BUILD($.bar, $.baz, $:hidden) {} works 22:06
nice!
autrijus++
_metaperl stevan, ping? 22:11
autrijus stevan: the test is flawed
stevan _metaperl: pong
autrijus: whats wrong?
autrijus my $foo = Foo.new(:bar($bar));
should be that, no?
otherwise you don't have an obj
stevan oh 22:12
:)
sorry, yes that is correct
autrijus fixed
and implemented.
stevan nice
BTW - about 20 more tests pass in t/oo/attributes.t now
autrijus cool... 2 more :)
stevan eval_ok should be officially discouraged
oh yes 22:13
autrijus wizards.p6 fixed to take advantage of this
stevan I use a longer lives_ok {} and is() combination which gives better diagnostics too
autrijus cool
stevan autrijus: did you see my meta-meta-class document? 22:14
autrijus no, sounds scary 22:15
looking now
stevan it is scary
svnbot6 r3194, Stevan++ | fixing several t/oo/attributes.t tests; eval_ok() does not work very well, I think we should stop using it; ~20 more tests pass now
r3195, autrijus++ | * MMD now deals with many more cases of expressions; in particular,
r3195, autrijus++ | $.chained.attr.methods works.
r3196, autrijus++ | * change wizards.p6 now typed method dispatch works
stevan it may not make much sense
but a sanity check would be appreciated 22:16
stevan realizes he is asking for a sanity check from autrijus, that may not be sane :)
autrijus stevan: have you seen my wikipedia user page? 22:17
stevan not in a while
autrijus it has been like that for months
# en.wikipedia.org/wiki/User:Autrijus
stevan nice 22:18
sanity is very overrated, and not much fun IMHO :)
autrijus stevan: looks good 22:23
stevan cool
svnbot6 r3197, Stevan++ | adding some more submethod BUILD tests to t/oo/attributes.t 22:33
Juerd eric256: Well, one doesn't turn on a narrow dyke with just enough room for a car to pass, and a fence on both sides. 22:34
nmcfarl /quit 22:56

Powered by Cro and the Raku® Programming Language.

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

Thank you!