pugscode.org <<Overview Journal Logs>> | r1734: ITypes *Test.pm currently broken* | pugs.kwiki.org
Set by autrijus on 10 April 2005.
jabbot pugs - 1751 - --Incorporated Juerd's notes (mostly, st 00:33
mdiep in r1868, I have 10 subtests unexpectedly succeeding 00:34
is that, uh, expected? 00:38
stevan Ovid: just sent your an email back 00:45
mdiep: r1868?? from which repository, I am only on r1751 00:46
mdiep URL: svn.perl.org/perl6/pugs/trunk 00:47
now r1871
stevan mdiep: oh,.. well a few tests succeeding is sometimes normal 00:48
mdiep should they get hunted down and get marked so that they're expected to pass?
Ovid_ Thanks for the feedback stevan. I'll be heading to the theater soon, so I can't really work on it much now anyway :)
stevan Ovid: I would be happy to help out on it more tomorrow (or whenever the ITypes refactor is done) 00:49
mdiep: I think those may have been fixed already
Ovid_ Thanks.
Any word on when classes are likely to be ready?
stevan Ovid: after the 6.2.0 release that is the major focus 00:50
Ovid_ OK. (whee!)
stevan given the current rate of development I would guess we would have primative objects in a week or two
and full on Perl6 OO maybe in month and a half?
(although I may be optimistic here) 00:51
mdiep: the tests are all currently broken though (Test.pm is not working) due to a major Type system refactor
Alias_ Maybe just a little :)
stevan basically Pugs has a broken nose
Ovid_ For the time being, even basic support would be a godsend. I don't even need inheritance right now. Just give me some polymorphic lovin' and I'll be quite happy.
stevan Ovid_: me too :)
Alias_: the IType refactor is a first step 00:52
and mugwump has been hacking away AFAIK
Alias_: would you have thought this much as possible in 69 days? 00:53
Alias_ How many developers are there now? 50? :)
stevan Alias_: yeah about that
Alias_ Then why not :) 00:54
stevan Alias_: thats the spirit :P 00:55
mugwump Objects! 01:09
jabbot pugs - 1752 - more tests on ** with special values 01:33
mugwump Has anyone here looked at any of the preliminary objects banter in the pugs tree? 01:52
01:58 luors is now known as LCamel 02:05 khisanth_ is now known as Khisanth
mugwump I'm trying to define a quick hack operator that turns one string into another... 04:27
I'm getting : Couldn't match `Val' against `t -> t1'
that's with:
OpName: VStr "foo" = VStr "bar"
s/:// 04:28
I tried:
Opname "foo" = "bar"
But then I got an error trying to match [Char] to [Val]
autrijus OpName (VStr "foo") = return (VStr "bar") 04:52
something like that
hmm, shapr did fix Test.pm. woot! 04:53
rehi lambdacamels, btw. :)
masak hi :)
mugwump thx autrijus, but still Couldn't match `Val' against `m a'
autrijus is this a op1? 04:54
mugwump nods
autrijus op1 "foo" = \x -> return (VStr (vCast x)) 04:55
op1 "foo" = opName
err
op1 "foo" = return . opName . vCast
opName "foo" = VStr "bar"
try that? 04:56
mugwump tries
autrijus backlogeth... 04:57
hmm, s/pleac/cookbook/ makes sense to me 04:58
but Ovid is not here
do you folks think examples/cookbook/ reads better than examples/pleac/ ?
mugwump what does pleac mean?
autrijus "Programming Language Examples Alike Cookbook."
mugwump was just playin'
q[acme] given that we're perl, i think it makes sense
autrijus pleac.sf.net that is 04:59
ok. well in the spirit of wikidev...I'll just go ahead and rename it.
mugwump That looks happier, now it appears to be moaning that I'm not providing a "default" case
autrijus mugwump: right, because pugs is now developed with super fascist -Wall -Werror 05:00
mugwump heh
autrijus mugwump: so you need to say
opName _ = VStr "undef"
or
opName _ = VUndef
or some such
mugwump cool! It compiles! 05:01
autrijus it compiles! it must be correct! 05:02
mugwump This proves that "clkao"? == "www.flickr.com/photos/obra/7893068/"
autrijus heh 05:03
jabbot pugs - 1753 - * rename examples/pleac/ to examples/coo
mugwump hmm, unexpected "\9832" ... 05:04
mugwump fiddles some more
autrijus you can't write nonlatin1 literals in code.
have to escape them 05:05
that's a ghc bug, not a pugs bug
(and it's being fixed I think)
mugwump I notice a yen operator too... so how would you escape that? 05:10
autrijus mugwump: the yen operator is within latin1 05:11
to get the hotspring operator 05:12
write
"\xE2\x99\xA8"
i.e. its octets in utf8 05:13
castaway_ inline haskel support diabled .. hmm, I thought I had the hsplugins..
jabbot pugs - 1754 -
mugwump ah, cool, thanks... so that ends up as 3 octets in the source, or the literal string you typed?
mugwump tries
autrijus I think you need that 3 escapted octets. 05:14
i.e. literal
mugwump Actually, this doesn't matter for my talk, I'll just get the ascii version working :) 05:15
dvergin FWIW 1754 is some work this evening on docs/S28draft.pod
Gotta remember to put in the comment with the commit
autrijus cool 05:16
dvergin More to come. But it's time for bed now.
autrijus nite dvergin!
dvergin 'nite.
autrijus mugwump: err, wow, you are actually going to commit the hotspring operator? 05:17
maybe doing it as a howto writeup on pugs.kwiki.org will be more fitting... 05:18
(but you're certainly free to commit... it's just I think it will get removed before 6.2.0)
mugwump ah, just flicked back to this after the commit 05:19
I'll un-do it before Sunday ;) 05:20
autrijus you didn't push, though, so that's fine :D
mugwump heh, true 05:23
ok I won't put that in the repos, then. 05:24
autrijus thanks :D
mugwump++
mugwump svk++
autrijus there are other primitives in S29 awaiting your work... :)
mugwump alright, off to give a talk to wellington.pm.org 05:25
later ...
autrijus see ya
gaal what's a hotspring operator? 06:12
Khisanth a person that operates a hotsprint? :) 06:21
Ovid Er, who moved examples/pleac to examples/cookbook? I think it's a good change, but I was rather curious about how this came about. autrijus asked for it to be in examples/pleac. 06:33
06:34 castaway_ is now known as castaway
castaway it was autrijus iirc, try asking svn blame ,) 06:34
Ovid I tried, but it doesn't work on directories and the files don't show the change :) 06:35
I'm glad autrijus did it, though.
castaway 06:58 < autrijus> hmm, s/pleac/cookbook/ makes sense to me
06:58 < autrijus> but Ovid is not here
Ovid Thanks! 06:36
castaway (presumably someone suggested it earlier, he was reading the backlog)
Ovid I had, actually.
castaway CAN YOU SEE WHO DID !/%ōæ½xA7 ōæ½xDF 8i HAVE NO KNOWLEDGE OF SVN9
oops! 06:37
castaway turns the capslock back off.. sorry
1753, that was meant to be ,)
Ovid Er, how do I see who did a particular commit by number? (I don't know svn that well)
castaway me either :) 06:38
Ovid :)
castaway looks in the redbean book
Ovid Is there a particular file you want to know the change for? I can do that. 06:39
svn log filename should do the trick. 06:40
castaway ah
svn log -r 1753
that was the pleac->cookbook one
Ovid pugs $ svn log -r 1753
------------------------------------------------------------------------
r1753 | autrijus | 2005-04-10 21:59:53 -0700 (Sun, 10 Apr 2005) | 1 line
* rename examples/pleac/ to examples/cookbook/.
------------------------------------------------------------------------
castaway yup
Ovid Thanks. I didn't know you could do a log that way. 06:41
castaway svnbook.red-bean.com/en/1.0/ch03s06...3-sect-5.1
now we're all more knowledged.. 06:42
castaway boggles 06:44
nothingmuch castaway: you know a bit of RPM.... --relocate is like PREFIX for Makefile.PL?
castaway We have Paris Hilton under contract?
In theory, nm, yes.. but only with rpms that are built to support it, iirc
nothingmuch ook, it's not installing even with --force 06:45
i just want the one binary out of the file
grr....
castaway aua.. what is it?
nothingmuch gcc 3.3 06:46
i want to try to make gcov with with gcc 3.2
and if it doesn't work adhoc, i'll just install a newer gcc along side 3.2
the 3.2 gcov is completely insane though
castaway hmm,m my suse 9.1 has gcc 3.3.3 per default, it seems
nothingmuch rhel as 3 here
horrible horrible platform 06:47
nothingmuch loaths redhat
rgs 2 </aol>
castaway hmm, could give you an account here, if its just to try something
nothingmuch no, i need to run gcov 06:48
i have other gccs, they are better
castaway ah
nothingmuch we have rhel 4, fc 3, rhel 2, etc
castaway mm, fun
oh, wait, you just want a binary?
which?
nothingmuch gcov itself =)
hah, why didn't i think of that 06:49
castaway you could just unpack the rpm.. rpm2cpio or somesuch..
but I can give you it
nothingmuch ah, i'll do that
castaway alien++
nothingmuch rpm2cpio | cpio <extract> ?
castaway something like that
I usually use alien
nothingmuch cpio -id worked 06:50
thanks
castaway :)
mine is on d-i just in case (in /misc) 06:51
tell me when you get a chance to start SEE?
nothingmuch i'll try now 06:53
i started it again yesterday
i wonder why it died
oh 06:54
ping woobling.org
=(
damn them to hell, my ISP
i hate pptp
and i hate l2tp
and i hate anything that is so flakey it doesn't stay connected for more than a week
*sob*
castaway hugs nothingmuch
nothingmuch you are very touchy-feely lately... =) 06:55
castaway I always am ,)
nothingmuch found a long-lost brother or something?
ah =)
castaway nah, its just my default mode ,) 06:56
nothingmuch wtf?!
i think it's a route problem
traceroute goes 3 hops,
and then stops
could you try to traceroute, see where it gives up?
castaway 8 netvision.ge4-0.3.cr02.nyc01.pccwbtn.net (63.218.9.2) 148.368 ms 167.602 ms 173.179 ms 06:57
9 * * *
10 * * *
11 * * *
nothingmuch arses
that is before even the backbone
castaway eep
nothingmuch from there it has to get to haifa/rehovot
and from there to my house 06:58
i get something entirely different:
3 gi1-0.core1.pt.nv.net.il (212.143.10.1) 10.272 ms !H * 7.246 ms !H
ā€¢ [syeeda:~] nothingmuch %
that's where it stops
this router is our upstream one
castaway hmm, odd 06:59
nothingmuch well, gcov dies with 'abort'
nothingmuch will use a hash table and move all the gcovs before every run 07:00
it's such a stupid tool
castaway whats it for?
nothingmuch it talks about ../foo/File.h, for example
and writes ./File.h.gcov
Devel::Cover for C/C++
castaway ah
8 netvision.ge4-0.3.cr02.nyc01.pccwbtn.net (63.218.9.2) 224.805 ms 244.667 ms * 07:01
9 pos9-0.core1.hfa.nv.net.il (212.143.12.21)(H!) 381.808 ms * *
(and then stopped, that time)
castaway tries to remember what H! is
nothingmuch damn, now this means I can't run gcov in parallel either 07:11
nothingmuch used to run 12 procs at a time to make it mui fast
*sigh*
castaway ,) 07:12
castaway pulls both corners of nothingmuch's mouth up.. smile! 07:13
nothingmuch bad day
but thanks, =)
castaway better 07:14
its only 10:15 ..
nothingmuch right
castaway plenty of time to improve ,) 07:15
nothingmuch i'll try ;-) 07:16
castaway gets to make excel templates, what fun
nothingmuch oi oi oi
i just write tsv's
castaway luckily I got to automate most of it a while back
nothingmuch wants ternary assignment 07:34
$lvaluse ??= "foo" :: "bar";
useful for $hash{$key} ??= die "duplicate" :: $value; 07:35
castaway :)
suggest it
nothingmuch i wonder if i could just implement it in p6 07:36
castaway probably
nothingmuch sub &ternafoo<??=::> ($lvalue is rw, $expr, $expr)
but that won't be short circuiting 07:37
castaway looks like gobbeldygook to me ,)
nothingmuch i'll mail p6l about thunks as parameters 07:39
damn, can't reach pasta 07:40
no mutt
fuck fuck fuck! 07:43
i hate gcov
castaway aww..
nothingmuch who the hell was stupid enough to think that basenames are a good enough distinction?!
castaway use something else?
nothingmuch i have two files in two dirs named common.h
they can'
t coexist, as far as it's concerned
it will just overwrite one with the other
castaway some idiot, no doubt
nothingmuch phooey!
gcov is the only tool 07:44
and it has gotten better
castaway really?
nothingmuch but i need gcov for gcc 3.2
which still sucks
castaway sounds like you'll need to build some ghastly framework around the calls 07:45
nothingmuch no
i'll just skip cvg of those files for now
they are probably unimportant
castaway (uhoh.. )
nothingmuch i'll warn about them now 07:46
if it's just common.h in 3 places, i don't care
they are 40 lines of unexecutable code, prolly
i could hack gcov itself 07:47
and recompile it
2.2
3.2, that is
castaway true
nothingmuch maybe that's the way to go
but i don't want these scripts failing if gcov suddenly gets swiped under them
because someone ran up2date
castaway any gcov/gcc groups you could whine on? 07:48
nothingmuch i've wined on #c
but, i dunno... it's a futile war
gcov was fixed
it's my problem that i can't upgrade
castaway oh, I see
:(
nothingmuch i love pbcopy and pbpaste 07:54
pbpaste | awk '{print $3}' | xargs -n 1 basename | pbcopy
castaway wossat? 07:56
nothingmuch take clipboard
munge it a bit
castaway oh, copy to clipboard temporarily? 07:57
nothingmuch to make paths in output table a list of basenames
and then put that back in the clipboard
castaway sneaky.. would generally just use a variable ,)
nothingmuch well, this is going between shells
between machines, actually
i would too, if it's the same shell 07:58
since it's more recoverable
castaway ah, useful 07:59
08:00 gugod_ is now known as gugod
castaway allo gugod 08:01
gugod hi 08:02
castaway seen theorbtwo
jabbot castaway: theorbtwo was seen on Mon Apr 11 03:20:32 2005
castaway is the jabbot yours? 08:03
gugod castaway: yes
castaway would you mind adding some timezone info or something similar? 08:04
gugod oh I was going to
castaway nifty
gugod but I forgot :p
I should do it now 08:05
castaway would be much appreciated
jabbot: time 08:07
jabbot castaway: Please go on.
gugod jabbot: seen gugod? 08:10
jabbot gugod: gugod was seen on Mon Apr 11 08:09:52 2005 GMT
gugod there. :p
castaway lovely :)
seen theorbtwo
jabbot castaway: theorbtwo was seen on Sun Apr 10 19:20:32 2005 GMT
castaway yay :)
gugod hmm, correct :) 08:11
I should have use gmtime() \
castaway hmm?
nothingmuch is there any way to set the pipe size? 08:25
that is, how many unread bytes before a the writer starts blocking?
ulimit -p can only display, btw, so don't look there 08:26
castaway has no idea 08:27
nothingmuch i feel so witty 08:37
cprint("$rel ... $.") unless $i++ % 7 08:38
cprint is just 'print \r, print @_, print lots of spaces'
i love prime numbers
lexical scopes + perl -n are scary 09:01
i just had 'my %hash; sub { }' 09:02
where sub plays with hash
since it's a perl -n script, though, %hash is different for every loop, and sub is a closure
beh
elmex ;) 09:06
lol
nothingmuch i don't know how newbies manage to deal with this stuff
elmex nothingmuch: they give up 09:07
nothingmuch i didn't
and i don't remember sufferring too much either
elmex it's just, that newbies don't know about -n 09:08
just a guess guess
oops
nothingmuch that's my theory too
elmex well, also that newbies ask on #perl or a friend i guess. or they just don't write code thats too complicated 09:09
i guess (programming) newbies also don't know much about closoures
shapr awakens 09:12
I'm desperately curious what the general purpose fix was for vCast
elmex vCast?
shapr Yesterday's IType upgrade added fromVal and I think my fix for Test.pm was likely a hack 09:13
Alias_ Answer, nobody uses perl's flags :) 09:14
elmex jafa?
err...java? 09:15
nothingmuch Alias_: i use them all the time =)
but that wasn't my point
autrijus \camels -> greeting !
shapr autrijus: I didn't fix all the vCast stuff, lots of the tests were failing when I fell asleep.
nothingmuch hola autrijus 09:16
there's lots of weird behavior like that all about p5
autrijus shapr: that's fine, I'm up and fixing them now
shapr yay
nothingmuch recompiles with -fno-inline 09:18
i thought what I did would be enough, but nooooo
theorbtwo Hm, the CHANGELOG is perhaps a bit overzelous about what I did. 09:32
IE it still doesn't actually work, but AFAICT the problem is in code at a higher level then mine.
Also, it still needs fixing to work for anything but SHA1 and anywhere but the paths I'm using. 09:33
autrijus right. would you tackle the latter? 09:46
I'll deal with the former.
theorbtwo Nod.
Turns out it's also broken by the ITypes changes, and I'm not clear on why. 09:47
autrijus really. hrm
let me install SHA1 and try out
a sec though, I'm fixing 09:48
() = (1..Inf)
theorbtwo That may have been a misdiagnosis.
castaway snuggles theorbtwo 09:50
(hmm, where'd he go?) 09:52
nothingmuch gah 10:20
meeting on coding standards in 40 minutes 10:21
and i want food
conflict
theorbtwo I suppose you can't eat at the meeting. 10:22
nothingmuch i would like a big fat sandwich
autrijus yay!
nothingmuch there's a guy that makes them by the univ 10:23
dirt cheap
very good
anyway, they always end up more like soup, by the time 10 minutes have passed
autrijus $ ./pugs -e '%ENV.ref.say'
Hash::Env
I win :D
nothingmuch because i'm a sucker for the sauces
autrijus++
yay!
theorbtwo Dammit, I want a good sub now.
Cool!
Autrijus++
nothingmuch death to gcov! death to gcc! 10:25
eek! 10:26
if i say no-default-inline i get to keep explicitly inlined functions
if i say no-inline, they all go away
if i say keep-inline-functions the used ones stay, but i want the dead code too
and some used ones are still inlined and thrown away
theorbtwo -fkeep-inline-functions' 10:27
nothingmuch tries -fno-default-inline, but with -Dinline="/* */"
theorbtwo: i tried that, no go 10:28
i should note:
with -fno-default-inline it seems behave correctly
that is, code is generated for the functions
but it is never executed
theorbtwo Try declaring "extern inline"? 10:29
nothingmuch will lookup exactly what that means
theorbtwo isn't quite sure. 10:30
nothingmuch ah, i see 10:31
it basically is supposed to make inline functions
but also keep a regular version, for external linkage, right?
ok, that doesn't help
i need these functions to be not inlined at all
like i hoped -g would do
so that they are actually executed as normal functions
theorbtwo Oh! 10:32
theorbtwo was thinking you wanted something completely different.
nothingmuch is trying to get the most debuggy, simplified code
that way coverage is accurate
on a per statement basis
theorbtwo Ah.
nothingmuch if optimizations get into the picture, the meaning of the code -> instruction conversion is sometimes vague 10:33
theorbtwo nods.
castaway mm, food, good plan
theorbtwo Just pass a -O0 (that's oh, zero) option?
nothingmuch '-g' 10:34
i'm already passing that
autrijus hm.
@_ = (1..Inf); @_[3] = 4;
this is surprisingly difficult.
nothingmuch is it generative?
autrijus guess I'll need to implement full COW.
nothingmuch full cow for the element?
autrijus full cow for elements 10:35
nothingmuch what does it do now?
autrijus it tries to create a IORef for each rhs
and dies
nothingmuch is curious
each rhs being 1 .. Inf?
autrijus yeah
nothingmuch why does that happen? 10:36
autrijus because each array element is potentially a fully tieable scalar.
nothingmuch needs a -fnever-ever-inline-you-stupid-twit option
autrijus i.e. you can do
tie($ENV{USER}, 'Tie::Google')
nothingmuch right
but isn't that supposed to be figured out per element
when the generator returns it?
autrijus I'm implementing that "supposed to". ;) 10:37
nothingmuch [3] doesn't exist yet, when @_[3] is assigned
autrijus yup.
nothingmuch so it's genererated
to see if it's tied
and then replaced
ok
again /me is just curious as to why that would have happenned
autrijus praises "lazy evaluation by default"
theorbtwo Can you generate @_[0..2], assign them to where they should be, and leave [4...] as the inf bit?
autrijus yes.
that's the idea
theorbtwo Ah, OK. 10:38
nothingmuch why do you need 2?
are the generators always iterative?
theorbtwo Because then you can have it so the left hand "half" is each-element-is-known, with a generator to the right.
And not a strange mix of generated and known. 10:39
nothingmuch right, but these are not linked lists
i don't want to wait forever if i have
@a = 0 .. Inf
@a[2**64] = 1;
autrijus you don't need to wait forever :)
theorbtwo I think they are linked lists...
nothingmuch perhaps generators should get something akin to slice notation
theorbtwo: in haskell ;-) 10:40
autrijus they are linked lists, but slicing will be automagic.
nothingmuch in parrot/perl6 i hope they will be sparse arrays
autrijus in parrot it's already sparse arrays
and we can do that in haskell too
just need to get around to it :)
theorbtwo leaves the heavy lifting to autrijus and nm, tries to figure out how to dump a hi file. 10:41
nothingmuch peh
huraaah! Mutex.h:109: multiple definition of `std::memchr(void*, int, unsigned int)' 10:42
now i've really messed it up
shapr wadler mentioned pugs and higher order perl on his blog, spiffy 10:46
theorbtwo Hmmpf. 10:49
ghc -package hi dumphifile.hs && ./a.out
a.out: magic number mismatch: old/corrupt interface file?
cognominal is there a file with the siganture of the implemented predefined function? 10:50
theorbtwo Oh. 10:51
cognominal s/siganture/signatures/; s/function/functions/
theorbtwo cognorminal, ghci, then :t whatever
Or check the specification: haskell.org/onlinereport/ 10:52
cognominal heu, are we speaking perl6 here?
theorbtwo Oh. 10:53
I thought you meant haskell
cognominal sorry for changing context without saing
theorbtwo There's a table at the end of src/Prim.hs 10:54
shapr yay - wadler.blogspot.com/2005/04/perl-6-...-pugs.html
autrijus ok. I triggered this again 10:59
src/AST.hs:1020:12:
My brain just exploded.
I can't handle pattern bindings for existentially-quantified constructors.
I wonder why I tend to write brain-explosion code.
nothingmuch autrijus: diet coke 11:00
autrijus I'm not on diet coke anymore :p
nothingmuch that's good news =)
theorbtwo autrijus: That's an actual compiler error message: "my brain just exploded"?
autrijus aye it is
rumour has it that it was spj's brain that has exploded when he worked on that part in the compiler. 11:02
nothingmuch spj?
autrijus research.microsoft.com/%7Esimonpj/
nothingmuch ah 11:03
autrijus father of GHC and inventor/coinventor of lots of nifty features we use
shapr and obsessive research paper author
autrijus that too.
sad, because with bindings for ex.q. constructors, we'll have full binding power for GADTs 11:04
and persumably making Show etc derivable
shapr you could always ask SPJ
osfameron "the implementation people seem to do
nothing but tweak their VM" # fair ?
nothingmuch whose brain exploded, btw? 11:05
autrijus nothingmuch: simon pj's
nothingmuch oh oh oh
i get it
autrijus osfameron: well, there's python, not intercal
nothingmuch that's the compiler output
i thought it was a comment 11:06
autrijus it's a compiler output :)
castaway Anyone speak/read italian? /me looks at larsen 11:08
osfameron and me
(not as well as larsen, certo)
castaway trying to decipher: cgi.ebay.de/ws/eBayISAPI.dll?ViewIt...eName=WD2V 11:09
osfameron In German?
Oh, I see, the description is Italian 11:10
castaway does that say it has RoutePlanner, *and* CityMaps on it, for EPOC systems?
oops, yeah, was earching on .de
osfameron Looks like there are 2 products? 11:11
castaway there are
but Ive not seen them on the same CD before, nor v2004, since the last Ive found is 2003 11:12
osfameron You've noted that he's only selling the software, not the GPS.
castaway yup, I have a GPS :)
just looking for software upgrades, the barstewards renamed it, changed the company name, and then stopped selling/giving map updates :( 11:13
osfameron Ah
castaway (mine is StreetPlanner Millenium from 1999 ,)
osfameron well he talks about both applications (sorry, haven't got the time to read in depth at mo')
castaway Ok 11:14
and running on EPOC?
(idiots only do their new stuff on Pocekt PC)
osfameron All he says about EPOC is that the RoutePlanner version supports additional languages 11:15
which implies that it does run on Epoc..
castaway indeed
ah well, 10 days to go on it
osfameron He suggests you contact him, telling him your OS 11:16
castaway thanks 11:18
autrijus hm 11:21
perl -e 'my @a = (1..10000000); @a[0]'
this is slow.
I wonder why ;)
theorbtwo Not Lazy enough for your level of Impatience?
autrijus likely so. 11:22
shapr That's a neat quote
Entwicklung is perl6 expected to be released this year ?
shapr gutentag Entwicklung, wie geht es?
Entwicklung shapr: no im not german, i just like the nick 11:23
:)
theorbtwo Thanks, shapr.
shapr That's okay, I'm from Selma, Alabama.
Robin_Kin why not $a[0] but @a[0] ?
pjcj aber hier sprechen wir deutsche
shapr Entwicklung: jump in and help! Then pugs will be further sooner!
Entwicklung i know nothing about compilers 11:24
shapr This is a great opportunity to learn!
theorbtwo Well, considering how far we are after 2 months, releasing after a year might be possible.
Ent, neither do I.
Entwicklung lol, by the time i know enough to contribute, perl6 will be ready
theorbtwo Writing tests is a good place to start, generally.
shapr There's a lot that can be done with straight-up perl skills.
Entwicklung i was under the impression that there were enough of people working on it already 11:25
theorbtwo pugscode.org <<Overview Journal Logs>> | r1734: ITypes *Test.pm currently somewhat broken* | pugs.kwiki.org 11:25
Robin_Kin it's an exciting message 11:25
11:25 Aankh|Clone is now known as Aankhen``
theorbtwo Ent: too many cooks spoiling the soup is because the soup doesn't have a sufficient version control system. 11:26
If you put in too much garlic, we can roll some back.
shapr laughs
Aankhen`` LMAO.
shapr likes spicy food.
theorbtwo Yeah, well, I'm not sure there is such a thing as too much garlic.
autrijus Robin_Kin: @a[0] is because my finger is used to that
Aankhen`` Er... I'm sure there is; does that count, theorbtwo?
Entwicklung hehe
autrijus same symptom if you use $a[0]
Entwicklung: fancy working on the cookbook? :) 11:27
Robin_Kin like spicy food to
theorbtwo OTOH, for(1..1000000) doesn't have that symptom, at least in sufficently new p5s.
Entwicklung lol
shapr Ah, the cookbook is under example/cookbook 11:28
autrijus theorbtwo: right, but that's peephole optimisation
shapr I might actually learn how to write Perl.
autrijus $ time ./pugs -e 'my @a = (1..Inf); @a[16384].say' 11:29
16385
sys 0m0.000s
infinite lists, constant time
pjcj where constant = zero!
autrijus O(0)!
shapr Entwicklung: jump on in, the code is warm and inviting!
Aankhen`` O_O
theorbtwo boggles, as autrijus warps the structure of space and time just for the fun of it. 11:30
shapr speaking of quoteworthy...
pjcj theorbtwo: are you suggesting autrijus has put on weight? 11:31
theorbtwo laughs aloud.
autrijus lol
jabbot pugs - 1757 - * Lazy IArray structures: "Infinite list 11:33
pugs - 1756 - * make ref() work on ITypes.
pugs - 1755 - * iType interface to let ITypes have dis
Aankhen`` Hrm... 11:49
I just came across this in examples/cookbook/01strings/01-00introduction.p6: 11:50
$string = q'Jon 'Maddog' Orwant'; # literal single quotes
Does that work?
How does that work, even?
castaway use something else as the quote on the outside?
q(Jon 'Maddog' Orwant); ?
Aankhen`` It's with a bunch of other ways to quote the string. 11:51
svn.openfoundry.org/pugs/examples/c...duction.p6
Take a look at the portion under "# Alternative operators can be used in the place of // delimeters"
castaway ah.. 11:52
It doesnt look like it would to me, but what do I know? :)
Aankhen`` I was just wondering whether that was put in accidentally, or it actually works.
BRB. 11:53
theorbtwo Hm, I'd expect that to be an accident; it works with bracketing chars, but you can't expect a ' to come in matched pairs.
jabbot pugs - 1758 - Added merlyn Soup with Crackers recipe
osfameron I didn't know // was the canonical single-quoting delimiter
theorbtwo (Witness "can't".)
It's not. It is, however, the cannonical character to use with anyquotes, no matter what the verb is on them.) 11:54
castaway thinks someone cut&pasted all the comments ("literal single quotes") 11:55
autrijus yeah :)
Aankhen`` Ah.
Anyway.
Aankhen`` BBL.
autrijus however, hash is strict. 11:58
i.e. an infinite sized hash does not exist yet 11:59
I think it's a feature.
castaway looks at some of the holes in the cookbook directory and hmms
12:02 integral is now known as aardvax
theorbtwo So do I. 12:02
12:03 aardvax is now known as integral
theorbtwo OTOH, you could create an infinite hash well enough if you don't mind creating a new IType. (In a theoretical future where you can create ITypes.) 12:03
jabbot pugs - 1759 - * unified hash and array fetching.
theorbtwo Of course, that's syntatic sugar for a function with a single argument... but what isn't?
shapr grins 12:04
autrijus lol :) 12:07
ok, @a[1..10] = (1..Inf) now works too. 12:08
castaway blinks.
theorbtwo Oh, that could be a nice idiom.
shapr Is that like "take 10 [1..]" ?
autrijus it is. 12:09
theorbtwo Equiv to @a[1..10] = [1..10], but without having to specify that 10 twice.
autrijus and I bet it's implemented like that underneath
yup.
autark-jp what happens with @a[1..10] = reverse (1 .. Inf)? If that is even legal that is.
autrijus ok... now let me repair quicksort and friends.
kungfuftr ullo all 12:10
autrijus autark-jp: I imagine it will go on forever.
theorbtwo Allo, kungfuftr.
autrijus either that, or incorrectly fill all 10 slots with Infinity.
I prefer that first
autark-jp ie, reverse is non-lazy?
kungfuftr odd... pugs actually builds for me now 12:11
theorbtwo Why is that odd?
kungfuftr because i couldn't over the weekend
theorbtwo For a few hours, hs-plugins was a required dependency; it's now optional.
(You should get a message at Makefile.PL time if you don't have it.) 12:12
kungfuftr complained about memory exhaustion during linking
theorbtwo Oh.
theorbtwo has no explanition for that.
shapr Your memory is tired, it needs a nap.
kungfuftr boomboom 12:13
shapr snickers
With useful content like that, who needs distractions?
"shapr, now with 52% more silliness!" 12:14
shapr sucks down some anti-silly Ritalin
castaway grins
silly++
kungfuftr "shapr, now with missing kneecaps"
shapr laughs
Silly is great if you don't have a deadline.
castaway indeed 12:15
kungfuftr (brad pitt)++ # impressive rendition of a northern irish accent
shapr kungfuftr: I'd have to see a picture of you to know whether I'd fight like a man, or run like a supremely confident man.
theorbtwo IIRC, 6.2.0's deadline is in about -6 hours.
shapr w00
kungfuftr shapr: northern irish... run!
shapr yipes!
kungfuftr hehe 12:16
theorbtwo tries to remember what he was doing.
shapr I have a good friend who's from Dublin, run with confidence is the best choice.
kungfuftr shapr: man with a plan, etc. 12:19
castaway something about frying fish ,) 12:20
shapr Panama!
kungfuftr wonders if you could use a crusoe chip to run parrot on top
castaway grins
cognominal how one would enumerate the keys of the global hash? 12:21
theorbtwo keys %*ENV 12:22
rgs %*ENV.keys 12:23
cognominal hum, that is the environment
jabbot pugs - 1760 - Add $*UID, $*EUID, $*GID, and $*EGID. N
autrijus say
shapr says
theorbtwo Yeah, stick whatever name you want in there.
"\n"
autrijus { say $_ }.(@a, @b, @c);
cognominal for example I want to list the name of all the known function 12:24
autrijus this prints the first element of the first nonnull array, right?
theorbtwo I'm not sure "all the known functions" are in a hash (yet).
cognominal so I need a way to enumerate the keys of the global hash
or that is what I think
kungfuftr cognominal: you want to walk the symbol table? 12:25
autrijus oh, you want the symbol table?
sure, I can add it in ~1min
cognominal yea
autrijus just give me the reference :)
%:: ?
cognominal does not work in pugs
autrijus is that still %:: ?
sure, I'm asking in the bible.
theorbtwo %MY:: -- L<S02/Names /%MY::/> 12:26
cognominal I like intropectable environments
kungfuftr "And abraham said, all shall be good if you walk in the way of the symbol table"
theorbtwo Also, %MyPackage::
(Just below.)
kungfuftr oh yay... smoke tests just decided to work again 12:28
theorbtwo I don't see any explicit mention of %*::, but I'm guessing that'd be the really global namespace.
kungfuftr lots of parse failures
cognominal "relative" names are conceptually searched in that order? %MY, %Mypackage, %:: 12:30
autrijus that looks sane.
please submit tests :)
cognominal correction: "relative" names are conceptually searched in that order? %MY, %Mypackage, %*:: , %::
how can I ask for commit privileges? 12:31
autrijus easy; give me your email addr :) 12:32
cognominal [email@hidden.address] 12:33
rgs cognominal, do you read chinese ? 12:35
cognominal not yet
autrijus cognominal: invitation sent. 12:36
you will not see chinese :)
rgs do you know how to say "first born child" in chinese ?
cognominal thx
rgs ooh, autrijus, it's translated now :)
autrijus yeah :)
cognominal I hope that echo, who handles my mail does not block mails from taiwan 12:37
castaway wonders idly if GeoIP gets taiwain & china lumped in together 12:38
autrijus yeah :) 12:42
oops. :)
castaway it does? 12:43
jabbot pugs - 1761 - * quicksort (slurp param) fixed
cognominal autrijus: I don't see anything, can you try to send the invitation to [email@hidden.address] ?
gaal does anyone have a pdf or a ps of Two Dozen Short Lessons in Haskell, that interleaves the Q and A pages? 12:44
Limbic_Region chip about? 12:46
seen chip?
jabbot Limbic_Region: chip was seen on Fri Apr 8 19:06:05 2005 GMT
theorbtwo castaway: Nope.
nothingmuch Limbic_Region: try irc.perl.org?
theorbtwo 2 days, 15 hours ago. 12:47
Limbic_Region nothingmuch - I will just /msg him
his response to a question at the Monastery has be a bit worried
castaway nope to?
autrijus cognominal: sent 12:48
theorbtwo no, geoip gets TW for taiwan, not CN.
cognominal autrijus: thx 12:49
castaway oh, good
oh, could have checked that myself, so it does 12:50
theorbtwo Autrijus, your last mail to p6c looks like it was written with the full-width latin chars, at least on my thunderbird. 12:54
autrijus it is not. maybe your big5 encoding font is borken.
maybe it uses the ugly MingLiU font.
theorbtwo Quite possible. 12:57
jabbot pugs - 1763 - * s:perl5:g restored; writeMVal excised 13:03
pugs - 1762 - * condition really works now, with fromV
autrijus hangman restored. 13:04
Test.pm fully unbroken. 13:05
theorbtwo Yey!
castaway yay!
shapr yay!
autrijus someone fix topic please
that's as of r1764
I'm restoring primitives now.
kungfuftr is the kwiki down? 13:10
autrijus yeah
kungfuftr ah... k
jabbot pugs - 1765 - * revive Compile.Parrot 13:13
pugs - 1764 - * IHandle and IRule; hangman restored.
kungfuftr autrijus: will you be attending YAPC::EU at all? 13:16
xern just added a naive p5 to p6 convertor utilizing regexps only.... 13:17
it needs much enhancement
autrijus kungfuftr: not sure, prolly not
oh wow!
kungfuftr xern: PRD goodness?
xern is planning to use PPI later. 13:18
kungfuftr: ?
autrijus may I rename it to p5_to_p6.pl?
"p526" reads like "page 526"
castaway :)
xern sure.... i was worried about the naming too
autrijus cool. 13:19
fantastic work.
now, dinner first. :) &
xern but it's quite silly in some way
i'll improvei t
it
kungfuftr: PRD?
autrijus rgs mentioned mandrakesoft has a GPL version of OCaml perl5 parser 13:20
if we can port that to Haskell
then maybe src/Parser/Perl5.hs :D
but, dinner really. &
kungfuftr xern: parse recdescent 13:21
xern kungfuftr: i considered it, but i'm wondering if PPI is much better than that. 13:22
kungfuftr xern: not sure, as Alias_ 13:23
ask*
jabbot pugs - 1767 - * rename.
pugs - 1766 - A super-naive p5-to-p6 convertor
xern kungfuftr: see here: ali.as/CPAN/PPI/index.html i guess it's better for normal perl5 scripts without Filter::Simple interference. 13:24
err... seems like a bad example 13:25
kungfuftr =0)
xern sorry, wrong url
kungfuftr i'm sure that it's well doable
xern ok i see
i'll try to use PPI for p5_to_p6.pl first 13:26
perhaps, PRD is good for regexp conversions?
xern is away for a while & 13:27
obra OCaml perl5 parser? 13:28
obra blinks
how close is it?
nothingmuch damnit 13:30
theorbtwo Hmm?
nothingmuch the tree huggers' website only has info about a campaign they're doing
no contact info for anything else
castaway hugs a tree. 13:31
theorbtwo hugs a tree-hugger.
nothingmuch castorbway: pasta is back up 13:32
lightstep which org?
nothingmuch not that it was down... =( 13:33
www.teva.org.il
lightstep i asked 144 once
nothingmuch i'm too lazy to get up
cognominal erl-MDK-Common
oops
nothingmuch but there's d.co.il
i'm also too lazy to pick up the phone to just call them
lightstep try clicking ליצי×Ø×Ŗ קש×Ø
castaway yay, so it is! 13:34
nothingmuch lightstep: only for that campaign
castaway hopes it stays up an hour or two til she gets home
nothingmuch there's the automated answering machine's number at the bottom of the page though
cognominal autrijus: it is a perl-checker that parse not too exotic perl5. It is part of perl-MDK-common source rpm package of mandriva 13:35
madriva being the new mandrakesoft name
castaway nudges theorbtwo
theorbtwo Huh? 13:36
castaway points to the /msg window
stevan pugscode.org <<Overview Journal Logs>> | r1764: *Test.pm fully unbroken* (details to follow) | pugs.kwiki.org 13:39
obra Whoa. parrot switched to svn? rockin 13:39
theorbtwo When is the release? 13:43
castaway WIR? :) 13:53
mj41 WinXP, nmake failed ... Compiling Compat ( src/Compat.hs, src/Compat.o ) 14:00
src/Compat.hs:86:20: Not in scope: type constructor or class `UserID'
src/Compat.hs:89:25: Not in scope: type constructor or class `UserID'
src/Compat.hs:92:21: Not in scope: type constructor or class `GroupID'
src/Compat.hs:95:26: Not in scope: type constructor or class `GroupID'
jabbot pugs - 1769 - s:g/%ENV/%*ENV/ 14:03
pugs - 1768 - Added some debugging statements; Tracked
nothingmuch is there a well accepted heap 14:06
like structure
that allows you to insert cheaply
get the top element cheaply 14:07
i know who it is since i have a ref to it
but also knows dynamic scoring?
kungfuftr nothingmuch: yes... a compost heap
nothingmuch that is, i'd like to yank out an arbitrary element
and then reinsert it
was that a joke?
kungfuftr um... yes 14:08
kungfuftr hides
nothingmuch beh
in a binary heap i could do exactly what I want
i think
kungfuftr more context required i thinks
nothingmuch well, i have a queue 14:09
i'm pushing into it 14:10
and popping out of it
sorry
and shifting out of it
i don't really care what order
but i can gain a lot if the thing i take out of the queue is good
because it means i can skip some things in the queue later
so i need a queue that:
i can pull out arbitrary things i know of 14:11
that sorts, based on a simple comparison
and can also be notified when an element needs to be reassessed 14:12
jabbot pugs - 1771 - * change comma to period in japh 14:13
pugs - 1770 - * "pop" and "shift" restored.
autrijus hi mj41 14:14
kungfuftr hhhmmm... no idea off hand
mj41 hi autrijus 14:15
nothingmuch anyway, this is an optimization for later
autrijus mj41: fixed in trunk r1773 14:17
please retest
wolverian er, isn't the "official" JAPH supposed to have a comma?
autrijus er, is it? 14:18
wow it is.
wolverian :)
autrijus my bad.
mj41 autrijus: seems ok
theorbtwo s/official/original/ # Who cares?
Whoever was looking for chip, he's about #parrot now. 14:20
autrijus now working thru failing tets...
castaway points at Limbic_Region
jabbot pugs - 1774 - * oops. wolverian points out comma is ri 14:23
pugs - 1773 - * stub Win32's UID/GID for now
pugs - 1772 - * env is no longer readonly
autrijus unshift av vals = liftIO $ modifyIORef av (map lazyScalar vals ++) 14:24
autrijus praises superconcise coding
theorbtwo What's that ++ operator doing? 14:26
lightstep concatenating strings 14:27
theorbtwo What strings -- I thought ++ was an infix operator (unless parened). 14:28
lightstep i meant lists
but strings are lists of chars
theorbtwo What lists -- I though ++ was an infix operator (unless parened). 14:29
lightstep yes
l1 ++ l2 = concat [l1, l2]
theorbtwo And map lazyScalar vals ++
lightstep ? 14:30
theorbtwo unshift av vals = liftIO $ modifyIORef av (map lazyScalar vals ++) 14:31
I don't think I understand that code.
autrijus ("Hello " ++) 14:32
is a function that takes a string and returns a string.
lightstep liftIO $ <-- from here on, i specify an IO action to be exewcuted in Eval
theorbtwo Oh!
Thank you, autrijus!
So that's equiv to... 14:33
Er, nevermind, I'm not quite sure how to say that expanded.
But it's that pointless... er, point-free style.
...and partial application of ++. 14:34
scw hi, `return' in pugs always returns a list which makes examples/games/animal.p6 break :( 14:35
autrijus scw: ok, fixing 14:36
scw and sorry for that I cannot spend much time on pugs these several weeks. Midterm now.. 14:39
stevan are tests still breaking? 14:40
autrijus stevan: many are recovered 14:41
stevan autrijus: ok,.. I am rebuilding now
autrijus: are you fixing tests or pugs? 14:42
autrijus pugs
stevan autrijus: when you are done with Pugs I want to do some more test cleaning 14:43
I found a few duplicated tests, and I think we are ready for a good cleaning
autrijus cool 14:45
castaway hands stevan the broom.
kungfuftr yay! my wuto-smoker works again 14:47
ninereasons autrijus, hi; thanks for renaming pleac to cookbook. more immediately recognizable, IMHO 14:49
dvergin so THATS what happened to pleac!! 14:50
I suggest a bookmark file called examples/pleac for poor fools like me. I was going nuts trying to find it just now
castaway dvergin: you should read the logs ,) 14:51
ninereasons it's exhausting to read the logs. (but good advice)
obra dvergin: it was only there for what, a week?
dvergin Still... 14:52
obra and did it exist in more than one snapshot?
pugs is a pretty fast moving target right now. and leaving around lots of clutter is going to hurt us pretty badly
dvergin Yes. I try to keep on the logs. But even grepping for pleac was a bit daunting
shapr agrees
castaway ninereasons: sure, but its easy to grep them for mention of something you cant find :)
dvergin agreed on the clutter issue
ninereasons to grep i must know what to grep for
theorbtwo Sadly, svn log rather sucks. 14:53
jabbot pugs - 1775 - * use modifyIORef to speed up arrays
shapr darcs.cgi is quite sexy, I could keep up a darcs mirror to show pretty changes.
dvergin mirrors the irc log as raw text. Works nice.
obra shapr: example URL/
?
castaway hmm, svn log worked fine for me this morning :) 14:54
shapr look at scribble - www.scannedinavian.org/cgi-bin/darc.../?c=browse
click on patches
dvergin Aha. We're talking about searching the svn log. That makes more sense.
castaway indeed 14:55
sorry, didnt mean the irc one
ninereasons ah. yes, i was talking about irc re: "exhausting" 14:56
dvergin tires of the mechanics sometimes
Can't find the svn log. I know I've seen it.
castaway type "svn log" :) 14:57
dvergin <giggle>
shapr obra: interesting for interactive visualization of changes? 14:58
obra isn't seeing anything that special
shapr ok
obra rt.openfoundry.org/Foundry/Project/.../pugs/log/ may be useful to me
obra may be missing where to look
shapr obra: the sexy darcs.cgi changes look like this - www.scannedinavian.org/cgi-bin/darc...c=annotate 15:03
jabbot pugs - 1777 - Forget to s/%ENV/%*ENV/ in the test desc
pugs - 1776 - Added very preliminary tests for macros.
shapr And you can click on any line to jump to the patch where it was committed, etc 15:04
obra oh. an annotate mode 15:05
ninereasons that's pretty impressive, shapr 15:06
shapr I haven't used svn much, is there something equivalent?
integral svn blame possibly
clkao the deleted one is the last modification only i presume? 15:07
ninereasons I like how every line is a link - the url is too long to read, is a history of changes retained on each revision, and linked?
shapr clkao: yes, but you can jump to any patch to see what happened there. 15:08
ninereasons: yes, pretty much.
ninereasons combined with wiki editing, darcs would make things go very wikiwiki! 15:09
shapr Yeah, I think the rss link is cute too. 15:10
Is util/testgraph.pl expected to generate the linked pages at some point? 15:13
theorbtwo shapr: util/categorize_tests is for that.
jabbot pugs - 1779 - Added test for paren-less 'defined'.
pugs - 1778 - Added some more tests for %*ENV.
shapr I really like the generated pugs-smoke.html, it is way spiffy. 15:23
I've been wishing for something like pugs-smoke that works for a wiki. I'd like to be able to specify a regex match or something as a 'unit test' and have the cells in a table turn red or green depending on whether it works. Sort of like fit/fitnesse. 15:24
Where do I get Perl6-Bible for catalog_tests? 15:25
castaway htp://search.cpan.org
oops
or just use your "cpan" installer
mj41 WinXP, r1773, 587/4256 subtests failed, 86.21% okay 15:26
autrijus hangman now hangs.
r1780
ninereasons in r1775: my @a = 1..2; # *** Error: cannot writeRef at {ref:<Array>}
theorbtwo pugscode.org <<Overview Journal Logs>> | r1773/winxp: 587/4256 | pugs.kwiki.org 15:26
Corion lights another pug. It's so hard to make good smoke from pugs ... 15:27
autrijus 587/4256 failed!?
oh my.
theorbtwo It could be worse. 587/4256 could have *succeeded*. 15:28
autrijus heh.
Corion Oh. "not a lvalue: Val( VStr "Hello World" )
autrijus what where eh? 15:29
Corion autrijus: In t/builtins/uc.t and t/builtins/lc.t
shapr that looks like it needs fromVal
shapr uses cpan for the first time 15:31
Corion The first hit is always free!
shapr tokes on a lambda 15:32
Corion Oy vey - the -I tests fail :) 15:33
jabbot pugs - 1780 - * return() and &?CALLER_CONTINUATION() a
mj41 wiki.kn.vutbr.cz/mj/index.cgi?diff_...ision=1.17
ninereasons re: the "cannot writeRef" error (above)? could that be the result of hs-plugins not being installed on this machine? 15:35
theorbtwo Shouldn't.
Makefile.PL now detects if hs-plugins is available or not.
Also, the functionality that hs-plugins allows for is untested. 15:36
ninereasons the build completed. but the completed pugs produces that error in r1775
autrijus just a sec 15:37
Corion ... maybe it's now time to rip out compatibility functionality out of P5, to make setEnv actually work on Win32 ;)))
autrijus Corion: look at your ghc's docs/libraries/ 15:38
err doc/html/libraries/
you should see System.Win32.*
including a System.Win32.DLL that can let you call arbitary win32 api
Corion autrijus: w00t ! That'll get me started with porting $*PID 15:39
dvergin Is there a searchable archive of perl.perl6.language? Google groups search seems to work but then cannot find the hits it reports. 15:41
I can't find anywhere to just download the whole mess. 15:42
jabbot pugs - 1781 - Cleanup. 15:43
ninereasons maybe you need to set loose a bot on www.nntp.perl.org, dvergin, to slurp up the perl.perl6.language directory ? 15:45
Corion Or just send Ask an email, maybe :)
ninereasons maybe :)
dvergin ninereasons: thought of that. Was looking for something pre-existing 15:46
autrijus %ENV.delete(<USER PATH>)
woot, works.
dvergin Corion: what is "Ask"?
theorbtwo Ask Bjorn Hansen.
ninereasons "There will be more nifty features like threading, searching and better navigation later. "
Corion Ask Bjoern Hansen. That's his name. His first name is "Ask".
ninereasons "Please send comments to Ask (address at the bottom of the page)" 15:47
www.nntp.perl.org/group/
theorbtwo BjĆørn, really.
dvergin goes off to investigate
Corion is too lazy :) 15:48
The RED ! It burns my eyes! datenzoo.de/pugs/win2k.html r1780 15:50
autrijus it's actually not too bad :p 15:51
theorbtwo pugscode.org <<Overview Journal Logs>> | r1773/winxp: 587/4256 smoke/win2k (occasional): datenzoo.de/pugs/win2k.html | pugs.kwiki.org 15:52
theorbtwo pugscode.org <<Overview Journal Logs>> | r1773/winxp: 587/4256 smoke/win2k (occasional): xrl.us/fqum | pugs.kwiki.org 15:53
jabbot pugs - 1782 - * delete() and exists(). 15:53
autrijus argh 15:56
my @array = <a b c d>;
is ~@array.delete(2), "c",
is ~@array, "a b d", "deletion of an array element";
who wrote this thing? :)
shapr runs svn blame
jabbot has commit access?? 15:58
autrijus what? 15:59
shapr is joking
svn blame actually shows iblech 16:00
Is it good or bad code? 16:01
autrijus it's mistaking an array for a hash :) 16:02
shapr ah
autrijus deleting things from an array should not shorten it.
shapr ok, how do I install Perl6-Bible with cpan? 16:04
theorbtwo perl -MCPAN -eshell
install Perl6-Bible
autrijus may need root permission 16:05
shapr maybe Perl6::Bible ?
aha!
theorbtwo Either -- Perl6-Bible is the distribution, which includes a module named Perl6::Bible 16:06
shapr yay
iblech autrijus: [S29] Deleted elements at the end of an Array shorten the length of the Array
autrijus iblech: it's not at the end.
iblech autrijus: Oh, ok, sorry -- fixing 16:07
autrijus err, no
I've fixed locally
let me commit first
iblech ok
(I unTODOed and unEVALed exists.t and delete.t too, btw -- already checked in) 16:08
autrijus delete.t all passes. please sanity check.
also, +%hash is right
no need to say .keys
I'm working on exists.t now
need to make exists() work on infinite lists 16:10
I mean arrays
gaal is a release planned today? (hi all) 16:11
autrijus it is.
depends on whether I can get the fixes in time.
theorbtwo Allo, gaal.
gaal i hope i can be of use, then. 16:12
iblech autrijus: [delete.t] looks sane, and I added a comment explaining the double spaces
autrijus audit tests :)
ninereasons having trouble installing hs-plugins on freebsd
"/src/altdata doesn't exist"
autrijus try gmake? 16:13
it works on my fbsd 5.4pre
jabbot pugs - 1785 - Added a comment explaining the double sp
pugs - 1784 - * .delete all works.
pugs - 1783 - * UnTODOed and unEVALed the tests for ex
ninereasons gmake fails. I'll investigate further; (freebsd 4.11) 16:14
ah. it assumes chroot "/src/altdata" leading slash 16:15
theorbtwo Odd...
jabbot pugs - 1786 - * exists.t now passes. 16:23
16:32 andras_ is now known as andras
kcwu gugod: how about let jabbot tell 10 more characters of log message? 16:32
autrijus and how about let it not reverse? 16:34
cognominal hum, I don't quite grok svk yet 16:38
svk push -C /home/stef/svk/pugs/ # prints the modified files 16:39
svk push /home/stef/svk/pugs # says "Empty merge"
stevan autrijus: am I doing something wrong? I am still getting "cannot writeRef Val (VRef <Scalar>)" on all the tests 16:41
(BTW - I am doing 'make optimized')
ninereasons 'make optimized' maybe that's the issue (same here) 16:43
stevan tries plain 'make'
cognominal with the -l option it did work! 16:44
stevan -l ?
iblech FYI, pugs compiled with plain "make" works 16:45
ninereasons thanks iblech. I'll stop wrestling with plugins for the present, then.
stevan its alive!!!! 16:46
theorbtwo Make optimized causes -dcore-lint to barf, meaning that there are probably bugs lurking there.
iblech (But a plain make <-> make optimized should make no difference -- Unless GHC has bugs?!)
cognominal stevan: I had to do a "svk push -l" to get my changes commited 16:47
autrijus hrmph hrmph
which means maybe I should try to kill .hs-boot
16:47 _metaperl is now known as metaperl
autrijus because that seems to be triggering all sorts of ghc bugs 16:47
ninereasons same bug, partially squished, as yesterday's 'cor-lint' bug? 16:48
autrijus oh well. so much for cutting edge treatment of mutual recursive modules
ninereasons autrijus, will a negative index work no an infinite array, eventually? 16:51
or is that crazy talk?
autrijus ninereasons: I very much doubt so.
stevan ninereasons: that *is* crazy talk 16:52
ninereasons :)
iblech I remember this question on p6l and Larry saying something like "I suppose yes"
16:52 Boogie is now known as boogie
ninereasons I have to train myself not to think "Inf eq end-unknown" 16:52
iblech i.e. my @array = (1,2, 3..Inf, 10,11); say @array[-1]; # ==> 11 16:53
stevan given enough time and memory,.. anything is possible :)
jabbot pugs - 1788 - * keys_values.t passes.
pugs - 1787 - r14@stefp: stef | 2005-04-11 17:46:23
stevan autrijus: so the current task is test auditing right?
autrijus stevan: yes. 16:55
clkao sb goto 00:37 16:56
stevan looks for some good test auditing music .... 16:57
t/oo/proxy has some unexpected succeeds :)
iblech BTW, autrijus, would it be hard to implement next() inside of for loops? 16:58
autrijus iblech: probably not hard.
it's just my mind's in IType right now :)
iblech :) np
autrijus it's basically just creating a lexical next()
that invokes the CC
that's code for return() in Monads.hs 16:59
theorbtwo Shouldn't it be more like ret() and less like return()? 17:02
autrijus what is this ret()?
jabbot pugs - 1789 - unTODOed tests which pass. 17:03
cognominal pugs> print keys %ENV 17:05
Segmentation fault
theorbtwo Hmm, can't find it now.
iblech plain '%*ENV;' segfaults in the interactive pugs here, too
but it works in "pugs -e '...'" or in test scripts
autrijus same here. 17:06
maybe readline? maybe something else?
ninereasons gah - segfault and now pugs-interactive is hung
iblech (FYI, r1790 -- unTODOed all tests which pass) 17:07
cognominal iblech: same here 17:08
kcwu same here 17:13
jabbot pugs - 1791 - Ported a Perl 5 solution to QoTW regular
pugs - 1790 - unTODOed two other tests which pass.
stevan ninereasons: do you mind if I remove fail_is() from Test.pm? 17:24
I think test functions like that are confusing, and ok(not($a == $b)) or something similar is easier to read 17:25
ninereasons: even better,.. I will rename it to isnt() 17:26
ninereasons certainly
stevan ninereasons: I just dont like the word fail, becuase it implies the test will fail (like the fail() function does),.. but that is not what you are actually after 17:27
ninereasons i never liked it much, stevan; and used it for lack of imagination :)
"fail if match" is what I wanted 17:28
stevan ninereasons: I changed it to isnt() and adding a unlike(), and I stole them from Test::More so no imagination here :)
autrijus hew, cleaned up builtins/keys_and_values/ 17:29
test driven++
autrijus looks for a uptodate smoke chart
ninereasons you guys are awesome fast. 17:30
autrijus stevan: can you sanitize builtins/io/?
if there's no too appalling failures there, I'd like you to unTODO/TODO them before release
I don't think 6.2.0 can deal with them
ninereasons retires to the couch 17:31
stevan autrijus: sure
gaal autrijus, util/run-smoke.pl?
autrijus oh, there's that?
thanks
I'm so behind on util/ technologies
castaway :)
gaal nothingmuch has somthing he's using, too, in an svn up ; get_a_smoke_test loop. 17:32
but it's proprietary technology and closed source! :)
castaway notes more things unexpectedly suceeding
jabbot pugs - 1794 - * cleanup redundant eval 17:33
pugs - 1793 - * pick.t passes.
pugs - 1792 - * pairs.t passes.
autrijus must be doing something right then
castaway indeed (this is one from 2 hrs ago) 17:34
gaal hmm, i should talk with kff about removign the dep on Shell since it isn't being used consistently there anyway
stevan autrijus: it seems that isa_ok() is broken 17:35
(it was a hack anyway)
autrijus broken how?
iblech stevan and others: Just as a remainder, you don't have to TODO failing tests in the test files, but you can simply edit t/force_todo
castaway Corion ?
gaal stevan: can i help with anything?
stevan iblech: yes, actually can you help getting started on that,. I am not sure I get it
autrijus right. but don't blindly force_todo
first check for sanity :) 17:36
iblech stevan: Of course
castaway oh, iblech, you speak german too, right?
autrijus iblech: so, please start from builtins/io/
castaway Would "telenummer" mean "telefonnummer" to you, or something else?
stevan autrijus: isa_ok() just checked the ref(), which is returning 'Scalar::Const' at the moment
iblech castaway: Yes. And yes, it'd mean "telefonnummer" for me 17:37
autrijus: Ok
castaway ta.. (silly finanzamt)
gaal i'm getting "Val (VRef <Scalar>)" crashes after each test on msys
.cannot writeRef 17:38
ninereasons force_todo won't work on fatal parsefails, though, will it ? or is that just what it's for?
autrijus stevan: please use index()
stevan autrijus: index() where?
autrijus Scalar::Const isa Scalar
ninereasons gall, don't "make optimized"
autrijus because index() test holds true
hack hack ;) 17:39
ninereasons gaa, don't "make optimized" (sorry)
stevan autrijus: It returns that for hashes too :)
autrijus aww.
oh well. forget I said that.
stevan and arrays and everything
autrijus clearly we need a true "isa" primitive.
stevan either I can hack it for a bit or just make it TODO all the tests for now
ninereasons i give up on getting your nick right, gaal :(
autrijus stevan: I'll hack on a isa. sec
stevan autrijus: even better :) 17:40
gaal ah, thanks 9r (don't worry about it, i keep mistyping it too)
(and it's my name, not just my nick :)
ninereasons even worse for me :( 17:41
theorbtwo BTW, how do you pronounce it?
gaal hard g, flat ah
shapr This is niner niner zelda comin atcha 17:42
autrijus oh btw, stevan 17:43
stevan autrijus: yes...
gaal brb, tea+snack dinner
ingy hola
ninereasons don't stop bouncing, shapr :)
autrijus nvm 17:44
I was confused about ref().
stevan shapr: you got a smokey on yur tail, back atcha!
shapr yay, citizens band!
ingy shapr: what's your 10-20 good buddy? 17:45
shapr I used to live in "circle city"
stevan I just used to watch "Smokey and the Bandit" when I was a kid :)
castaway oooh.. I love that film (those films)
stevan classics of american cinema :P 17:46
shapr How the heck do I describe my 10-20? Is there a CB name for Sweden?
castaway advises nothingmuch's SEE to brace for impact ,) 17:47
stevan shapr: "up in the great white north where the girls are pretty, but your freeze your,... over and out"
shapr haha
gcom hey all...i wrote an example assuming that $string.chars will return a list in list context, but i can't find anything that verifies this 17:49
seems like a nicer alternative to unpack or split, if it does indeed work 17:50
ninereasons I'm not sure that it should work that way, gcom 17:51
gcom the only documentation for char seems to be that it is one of the replacements for a length concept 17:52
*chars
autrijus gcom: it's not specced like that, although I can see it's interesting.
gcom: p6l! p6l!
gcom sorry, wrong forum :) 17:53
autrijus $ ./pugs -MTest -e "isa_ok(%ENV, 'Hash')"
ok 1 - The object is-a 'Hash'
$ ./pugs -MTest -e "isa_ok(%ENV, 'Scalar')"
not ok 1 - The object is-a 'Scalar'
# Failed test (-e at line 3, column 1)
# Expected: Scalar
# Got: Hash::Env
yay!
jabbot pugs - 1797 - TODOed all failing IO tests.
autrijus honest-to-god .isa
jabbot pugs - 1796 - * refactor lc/uc etc into op1Str and op1
pugs - 1795 - Added comment to t/force_todo.
iblech I TODOed all failing IO tests (r1797). Because of uncatchable dies in io_in_[while|for]_loops.t I had to manually skip() those tests.
stevan autrijus++
autrijus landed as r1798 17:54
enjoy
stevan iblech: thats good
autrijus: thanks :)
kungfuftr ah... home and fed 17:56
autrijus stevan: you may wish to update ChangeLog some :)
stevan autrijus: I will
autrijus danke
and I think I won't stay up and push 6.2.0 out when I'm dizzy
I think do it right after I wake up is better idea
crysflame oo, 6.2.0
:) 17:57
autrijus although I think the codebase is finally in okay shape now
I need to fix t/data_types/ now
gaal t/statements/gather....ok 3/5pugs.exe: Prelude.(!!): negative index 18:00
autrijus yup, saw that
kungfuftr pid still broken too? 18:01
autrijus worksforme, but I'm not on win32 18:02
kungfuftr i'm on freebsd 4.11
autrijus weird, worksforme.
$ ./pugs -e '$PID.say'
67834
iblech works for me, too (Linux, GHC 6.4)
stevan me too 18:03
kungfuftr oh... works now
must be the churn
jabbot pugs - 1801 - changed fail_is() usage to isnt()
pugs - 1800 - added todo_isnt/unlike(); added tests fo
pugs - 1799 - changed fail_is() to isnt() and added un
pugs - 1798 - * .isa() primitive.
stevan ok,.. so how are we doing here,.. what tests should I work on? 18:04
ninereasons my @a = 'a'..'d';
@a[1,2] = @a[2,1] 18:05
@a[1..2] = @a[2,1]
gaal stevan, what's fail_is? (is it obsoleted?)
ninereasons ('a', 'c', ('b', 'c'), 'd')
?
fail_is is now "isnt" 18:06
gaal ahh, thanks. 18:07
autrijus ninereasons: fixing that
fixed. you have a test for that? 18:08
ninereasons I was trying to think of a sensible to write it
my example is silly
autrijus sigh, Exp should be GADTs, that way we can avoid that sort of error. 18:09
but not 6.2.0.
Corion_ castaway: Back
ninereasons maybe it's something noone would ever do anyway
autrijus I'm sure someone will. 18:10
I will ;)
ninereasons so would I, perhaps, but for exactly opposite reasons that you would 18:11
(igorance ! eq foo)
autrijus anyway, the fix is in. 18:12
ninereasons you rock, autrijus 18:13
autrijus :) 18:14
ninereasons kungfuftr, have you succeeded in getting hs-plugins to install? 18:16
on freebsd 4.11 ?
jabbot pugs - 1804 - * ninereasons spotted that `@a[1..2]` im 18:23
pugs - 1803 - TODOed statements/* and rules/rules_refs
pugs - 1802 - * fix some golf syntax
ninereasons hmm. the tests now completely exhaust my knowledge of perl6 (and some).
back to the docs for me, I think. 18:24
autrijus nice 18:26
Corion_ pugs smoke chart at datenzoo.de/pugs/win2k.html (of r 1803)
18:27 Corion_ is now known as Corion
autrijus infinite slices now work again 18:27
ninereasons $infinite_slices++ ! 18:28
18:28 Corion is now known as Corion_
autrijus that should make the golfs happy. 18:28
gaal forum2.org/gaal/pugs/tests_winxp.html - r1794 - winxp/msys 18:31
had to inline the css though :(
jabbot pugs - 1808 - * add enterRValue 18:33
pugs - 1807 - * constant arrays no longer out-of-bound
pugs - 1806 - * =<> works again
pugs - 1805 - * infinite slices now work.
autrijus it's interesting comparing 1794 to 1803 :)
gaal we run on cygwin? 18:35
the tests should s/cygwin/msys/ afaict
stevan so are we not just TODOing test? (in the force_todo file) 18:37
autrijus: array refs seem to show up as Scalar with .isa(),.. is that correct? 18:41
autrijus no. fixing that
stevan autrijus: ok
autrijus as we speak, even
stevan t/builtins/list/reverse.t has that error a number of times 18:42
(in case your looking for a good test case)
gaal t/subroutines/subroutine.t infinitlooping again?
affter test 13: use @_ in sub 18:43
jabbot pugs - 1811 - force-TODO-ind t/builtins/list/sort.t &
pugs - 1810 - * for(){} should not be parsed as hash d
pugs - 1809 - forcing TODO on t/builtins/list/join.t
ninereasons this rate of revision is dizzying. it's like a wikpedia edit war without the acrimony 18:45
PerlJam Is there some reason jabbot reports them in reverse order?
stevan PerlJam: something to do with the the rotation of the earth and how fast development is 18:47
PerlJam stevan++ heh
stevan kinda like in the first Superman when Lois' car fell in the earthquake
ninereasons shouldn't they be in the right order in australia, then?
stevan ninereasons: (shhh dont spoil the illusion) 18:48
ninereasons stevan, i was hoping an australian would confirm !
autrijus stevan: that's fixed. I fixed ref.t too 18:49
as of r1812. 18:50
pugs> my $a = 123; $a.isa('Int') 18:51
bool::true
pugs> my $a = 123; $a.isa('StrStr')
bool::false
err, s/StrStr/Str/
ok. I need to 1)shower 2)journal 3)sleep. :)
ninereasons *applause* 18:52
autrijus meanwhile, please keep the tests that are regressions from before-itype time (i.e. those that really shouldn't fail) as fails
and force_todo nonregressions
and update ChangeLog
so I can fix the regressions and push out 6.2.0 when I wake up
much thanks =) 18:53
stevan autrijus: will do
autrijus #perl6++
castaway nm Corion, already got it, ta
autrijus thanks folks :D
18:53 Corion_ is now known as Corion
jabbot pugs - 1812 - * isa_ok works... ref() now works again 18:53
castaway night autrijus!
autrijus what, 70 commits today. 18:54
ninereasons 103 occupants (counting bots and zombies) build it and they will come
stevan autrijus: that depends on when you consider "today" to end :)
elmex i had a splendid idea.. why not let pugs parse java instead of perl6? ;) 18:56
Corion elmex: Feel free to write a Java parser+interpreter in Perl6 or Haskell
(plus runtime :) ) 18:57
Java itself shouldn't be too hard, at least much easier than Perl
elmex java sucks (sorry for the harsh word) 18:59
Corion brb (rebooting firefox) 19:00
ninereasons what is that, Corion? [Firefox 0.10.1/20041001] ? 19:02
Corion ninereasons: Chat!Zilla - JavaScript irc client written in/for FireFox 19:03
ninereasons [Firefox 1.0.2/20050317]
Corion ... has nice rendering, works with UTF8, and I could actually modify the UI, if I wanted to
See www.hacksrus.com/~ginda/chatzilla/
ninereasons mine is Chatzilla 0.9.67 [Firefox 1.0.2/20050317] (not on it now, though) 19:04
it was the firefox version # that caught my eye
Corion ninereasons: Eh - dunno - I didn't play with any of the FF version settings. I think I have 1.0.1 or whatever. 19:05
... maybe that's a bug or something.
ninereasons hm
autrijus hrm. {} is a closure.
Corion Yah. I'm using Mozilla Firefox 1.0pre
autrijus I want it to be a hash. I see little use for it as a closure. 19:06
Corion autrijus: I think "{}" is/could be an empty hash, while "{ }" could be a closure.
autrijus awwww
Corion Like &noop = { };
(so there is a use for empty closures)
welcome to Python land, where Whitespace matters ! :)
(but the time when whitespace didn't matter in Perl is long gone anyway) 19:07
gaal whitespace already does
Corion autrijus: Or disambiguate by +{}
gaal yeah :/
ninereasons autrijus, should we tell your mommy that you're not in bed?
Corion (like map does in P5)
ninereasons: He's in the shower.
ninereasons hah, Corion :)
autrijus yup.
stevan autrijus: BTW - $handle.isa(Handle) fails but ref($handle) returns 'Handle'
Corion We-ird. Now, t/data_types/array_ref.t seems to hang and eat 100% CPU 19:08
at r1812
(win2k)
autrijus I see. that. skip it for now?
mark it as a fail
stevan (do you mean me or Corion) 19:09
or both
Corion autrijus: Will do so (marking it as "skip" ? or into the untodo list ?)
"untodo" won't help much with an endless loop :)
autrijus skip_fail or something 19:10
gaal the harness should handle that kind of thing, by rights
Corion autrijus: Will do so, once my make test finishes
gaal my mind stopped working about thirty minutes ago or i'd do it
autrijus stevan: fixed
* 'Handle' is spelled 'IO' in S06.
Corion gaal: That requires alarm() or something like that. Which is hard to do.
autrijus r1813
gaal Corion, not reallt, becasue the harness is still p5 19:11
stevan svn up; make; :)
Corion gaal: Still requires alarm()
gaal so?
Corion gaal: Win32 native build doesn't have alarm() :) 19:12
gaal ah.
ultimately the test is run with open |. how does that work in win32? 19:13
jabbot pugs - 1813 - * 'Handle' is spelled 'IO' in S06.
Corion gaal: Win32 has open2 and open3
gaal they CreateProcess? 19:14
Corion gaal: Hmmm. alarm() is half-working on Win32 native Perl, BTW
gaal: Yes - they don't rely on the shell, which doesn't really do |foo, and CreateProcess with the correct handles. 19:15
gaal t/subroutines/subroutine.t still seems to get stuck on msys, perl 1812
well, threads work ok on win32, don't they? it means branching code and ickiness, but it can be done 19:16
Corion gaal: Sure ;)
Where/how do I install additional Haskell libs ? I'd like to install the fabled System.Win32.DLL, but everything I find are the "fptools" ...
gaal quesion is whether we can assume perl itself has threading enabled
on that platform anyway 19:17
Corion I get three unexpected successes - should I skip those or promote those?
stevan autrijus: how many hours till you wake up? (So I know when to get the changelog done by)
autrijus 10 hours 19:18
or 9
stevan Corion: where were the successes? 19:19
Corion stevan: Dunno - I think datenzoo.de/pugs/win2k.html showed them in undef.t somewhere 19:20
stevan what color are they? 19:21
Corion dark red 19:22
#15 in undef.t for example
stevan I expect the "non-existent hash subscript" one does work now
so that can be promoted
the FIXME parsefail right beloe that (line 63) might work now too 19:23
Corion go go go ! :)
jabbot pugs - 1816 - Fixed accidential overeager commit
pugs - 1815 - skipping 2 tests in array_ref.t for rele
pugs - 1814 - Handle is now spelt IO in the tests too;
Corion I'm making a new smoke chart
gaal yay - can't wait for when the last block in that file gets to be uncommented :)
Corion ... and I should get a second checkout for development, so my bogus modified versions don't get committed accidentially :( 19:24
stevan Corion: I am grabbing some lunch while I 'make test'
gaal: AUTOLOAD,.. YUK!!!! :P
gaal it's *refined* AUTOLOAD :)
stevan although this is better than perl5 AUTOLOAD
gaal: OT - is what I read in the backlog true,.. your going to jail soon? 19:25
gaal likely, yes.
Corion Hah. I just had a brilliant idea - I can wipe my whole source tree except the root .svn directory, so I get a clean checkout without having to re-setup the password/login/location
stevan because your refuse to do the military service in the occupied territories right?
Corion gaal: That sucks indeed - but last I knew, you didn't even have public/civil service in .il ... 19:26
gaal yes, that's true. but it's not like i'm dead set on jail :) if they offer me another task, i'll take it
autrijus oh wow. I did it. someone write a test for it?
Corion gaal: But props to you for not working in the occupied areas!
autrijus pugs> my $a = { 1 => 2 }; $a.ref
Hash
gaal autrijus, what, autoload? yes, bottom of undef.t, ages ago 19:27
stevan gaal++ # for standing up to The Man!
autrijus gaal: no, hash composition
aka anonhash.
stevan autrijus: I think we might have a test for that somewhere
if not we will write it :)
autrijus good. then watch out for unexpected successes :)
gaal Corion, there's compulsory military service in Israel, 18-36 months 19:28
(36 for 18-yr-o males)
stevan has to grab lunch for real before this coffee rots a hole in his gut
Corion gaal: In Germany too (but shorter), but I didn't know that you had an option out of the "military" part of the service. 19:29
ninereasons autrijus, that worked before ITypes - nice to have it back!
gaal oh, no, in the normal case you don't.
Corion All fails will be fixed by modifying the basic force_todo file, right?
autrijus ninereasons: really? somehow I don't remember that
Corion gaal: Oh ?
ninereasons I didn't write a test for it, but you can see it in the irc log from a couple of days ago 19:30
autrijus { 1 => 2 }.{1} # I don't thinks there's code for it
gaal i'm way after my regular service. like most israelis i get called to do reserve duty every now and then
autrijus really. interesting
gaal is old
Corion gaal: Ewww. That's ugly !
gaal autrijus, is that supposed to eval to 2? 19:31
autrijus yes.
Corion gaal: "luckily", the only thing They can draft me for is sweeping minefields ...
autrijus it now does
gaal they can draft me for sweeping the floor if they want :)
ninereasons my %hash = map -> @a { @a[0] => @a[1..Inf] } @z; 19:32
gaal wow, cool
ninereasons worked (but maybe it shouldn' t have) a few days ago
gaal ...and when i said "..offer me another task", i meant military duty that does not pertain to the OT 19:33
ninereasons '%hash.perl' in that case was '((\'factor\' => (\'a\', \'b\', \'c\')), (\'timers\' => (1, 2, 3)))'
from my @z = (['timers',1..3],['factor','a'.. 'c']) ;
jabbot pugs - 1817 - * anonymous hash literals. 19:34
autrijus ninereasons: that's a different matter
{} wasn't acting as anonsub 19:35
I mean anonhash
it's acting as a simple closure
in a pre-IType pugs:
pugs> {1 => 2}.{1}
(pugs: cannot cast into [VPair]: VSub (Sub {isMulti
that's what would have happened
ninereasons i'm not sure that what I posted is correct perl6, anyway. 19:36
Corion Ah - BTW - I've long been wanting the "-g" and "-m" options in Perl. "-g" is mostly -e "while ( $_ = =<> ) { say if (" ... -e "); };" and -m is -e "while ( $_ = =<> ) { say (" ... -e "); };". Does anybody use the -m switch at all?
autrijus I think it is.
Corion (-g for "grep" and -m for "map", obviously)
Khisanth g and m would make things very confusing 19:37
Corion and -g '$ARGV:' should become 'say "$ARGV:$_"'
Khisanth: In what sense confusing?
autrijus Corion: I have threatened for a couple years now to re-upload pmake to CPAN as "ake.pm".
perl -make test 19:38
perl -make install
Corion autrijus: :-)) Just like Tye McQueens ouse.pm , which you invoke as "perl -Mouse"
autrijus: :))
autrijus but in this case, I use small -m :)
Khisanth Corion: nvm, misunderstood what you were refering to
Corion autrijus: I see... You could live with "-Make", but "-make" is nicer... I could settle for -G , but I'd prefer -m for "map"...
autrijus in any case, that's p6l domain :) 19:39
Corion ... and having (Unicode!) grep and map is nice IMO.
autrijus (you can of course implement it now if you'd like)
Corion autrijus: I had originally planned that as an (semi serious) April Fools joke post to p5p :)
Khisanth autrijus: perl -Make would still work in that case wouldn't it?
gaal autrijus, it doesn't work for me? ./pugs.exe -e 'say {1,2}.{1}'
<SubBlock(<anon>)>
otoh, this does work: ./pugs.exe -e 'say [1,2].[1]' ==> 2
autrijus Khisanth: yes, but wouldn't be as nice.
Corion autrijus: Yeah - I think I'll implement it, and see if it'll be taken out or not. And patch it in again! SVN commit fight! :) 19:40
autrijus gaal: r1817?
heh. I'm glad we don't have revert wars
Corion autrijus: Well - the community is still fresh, and quick yelling on irc is better than prolonged flamewars on p6* 19:41
gaal yes
Corion ... and people here get into contact easier, and maybe even have met. 19:42
(mostly due to Autrijus World Tour 2004/2005)
shapr buys tickets for the 2006 tour
gaal Corion: you could write a test that, when it fails, patches the tree and svn ups. 19:43
incidentally {1,2}.{1}.say prints nothing
Corion gaal: I was considering writing a bot that watches the openfoundry RSS :)
Corion does automation for a living.
gaal well, prints a newline
Corion Hmmm. Maybe we should make the setEnv less fatal, so it can be caught with a try{} block (as before). Is it too late to make Pugs / Haskell patches? 19:44
autrijus it is never too late. 19:45
there is no code freeze in pugs development. :)
elmex not?
Corion autrijus: Yeah, but I don't want to make an unclean patch when everybody is in preflight :)
gaal yay! hanzi finally started appearing in my terminal!
autrijus Corion: go ahead and commit it in.
elmex i thought we could freeze now and go on implementing perl6 in perl6 ;) *G*
autrijus we have a VCS.
Corion autrijus: :)
autrijus elmex: we can't. maybe you can. pugs is certainly turing complete ;) 19:46
so go ahead if you'd like.
elmex autrijus: ;)
autrijus: i'm not a compiler specialist
even thought i wrote a topdown parser generator some weeks ago 19:47
shapr me neither!
Corion Oh. And in the future, I'd like an introspective list of compilers, if that's possible/sensible to want, so I can reduce the code in my compiler test :)
elmex (in perl)
autrijus journal up. my fingers are tired. I should probably sleep. :)
autrijus waves &
shapr g'night autrijus 19:48
gaal zzz also. bye! 19:55
ninereasons night gaal (<- note the correct spelling) 19:56
gaal thanks, ninereasons :) 19:57
mj41 pugs -e "say @ARGS[0]" args_no_argv 20:24
pugs -e "return 1;"
cannot return() outside a subroutine
Val (VList [VInt 1])
Is this ok?
ninereasons pugs -e "my &a = sub {return 1}; say a" 20:26
mj41 my $conf = require 'conf.p6'; 20:27
# conf.p6
return {
path => '.',
};
{ path => '.', }; # is ok 20:28
ninereasons r1817; oops - build now broken unless hs-plugins are installed 20:34
Corion back from phone :) 20:37
ninereasons does anyone else have freebsd 4.x ? I've failed to figure out this hs-plugins install problem.
Corion ninereasons: worksforme
theorbtwo Hm, and cognominal broke the AUTHORS convention.
Corion (Win2k, no plugins that I know of)
ninereasons on freebsd?
Corion ninereasons: No, no freebsd 20:38
ninereasons oh i see what you mean.
Corion Don't you know - *BSD is dying
:)))
ninereasons I'm covering all my bets, Corion
shapr covers all his bets with one finger.
ninereasons I'm on Win32, using freebsd over ssh, via linux via cygwin xwindows
stevan wonders if Corion really has three mouths or not
Corion Bah. I'm ditching the setEnv patches for Win32 and will just disable the test. 20:39
(as it's too late/I'm too tired to figure out how to make the Haskell failWith() in Compat.hs work like die() in Perl)
stevan: No no - you're thinking of the Beast Of A Thousand Young. I'm more like Cthulhu 20:40
ninereasons r1817 builds for you, Corion; without plugins installed?
Corion ninereasons: I think I don't have any plugins. I'm using the plain GHC 6.4 Win32 install 20:41
But I wouldn't know how to check if I have hs-plugins
ninereasons ?? (maybe BSD really is dying :(
stevan ninereasons: doubt it is dying (it is the base of Darwin and some ISPs love it) 20:42
ninereasons look for the make option -DPUGS_HAVE_HSPLUGINS=1
theorbtwo Corion: ghc-pkg describe plugins. 20:43
Or what ninereasions said.
Corion "cannot find package plugins"
theorbtwo In what way is it killing the build?
ninereasons Chasing modules from: src/Main.hs
Could not find module `Plugins':
jabbot pugs - 1819 - More sanity in External.Haskell. Things
pugs - 1818 - fixed t/builtins/string/lc.t; and more f
ninereasons ah .. 20:44
interesting.
-DPUGS_HAVE_HSPLUGINS=1 is in the build, without plugins installed
Aankhen`` goes to sleep.
Nite. 20:45
Corion theorbtwo: You mentioned System.Win32 - how would I install that module/package in Haskell ? (And where should I find it?) 20:46
theorbtwo ninereasons: ghc-pkg describe plugins ?
Corion: autrijus mentioned that, not me. 20:47
ninereasons ghc-pkg: cannot find package plugins
Corion theorbtwo: Ah. It's too late and I'm confusing people already :(
ninereasons I'll manually alter the Makefile, and see if that works
Corion Good night then ;) My idea of making setEnv work will have to wait... I'll just skip() the %*ENV tests on Win32
stevan anybody want to take a look at t/builtins/strings/length.t 20:48
it is only showing 6 test for me
but there should be 15
oh wait 20:49
I see it now
Khisanth the for?
stevan yeah
it was assuming the older flattening behavior
ok much better now :) 20:50
theorbtwo ninereasons: Try ghc-pkg describe plugins && echo "Yes"
ninereasons ghc-pkg: cannot find package plugins
theorbtwo Now try ghc-pkg describe base 20:51
Er, && echo "Yes"
ninereasons name: base
version: 1.0
.. YES
theorbtwo Hm. 20:52
No clue, sorry.
ninereasons it's alright - it was educational :)
shapr what's the problem?
ninereasons my build is showing that I have plugins, when I don't 20:53
-DPUGS_HAVE_HSPLUGINS=1
shapr are you using ghc6.4 ?
ninereasons ghc-pkg: cannot find package plugins
Glasgow Haskell Compiler, Version 6.4, for Haskell 98, compiled by GHC version 6.4
shapr what about "ghc-pkg -l|grep plugins" ? 20:54
ninereasons null
ghc-pkg -l|grep plugins || echo "NOPE"
NOPE
shapr well, you definitely don't have hs-plugins then.
theorbtwo shapr: Read inc/Pugs/MakeMaker.pm, have_ghc_package, near the bottom. 20:57
kungfuftr still waits for his compiles to finish
ninereasons if kungfufter's build completes, it's probably not OS relate (freebsd 4.11) 20:59
my problems, that is
kungfuftr ninereasons: ah... my smoke is still working... getting my laptop to a usable state 21:01
stevan kungfuftr: my tests were stalling on the t/examples/golf tests ,.. just FYI 21:02
kungfuftr stevan: kungfuftr.com/pugs-smoke.html 21:03
stevan kungfuftr: nevermind then :) 21:04
wow what happened with the builtins?
kungfuftr no idea... that only builds every hour 21:05
stevan strange
kungfuftr checks stuff
stevan theorbtwo: Profit? 21:06
theorbtwo It's a South Park / Slashdot reference.
stevan theorbtwo: You Bastard! 21:07
cognominal theorbtwo: oops. what should U do to correct my mistake
kungfuftr stevan: 2 runs seem to have run into each other 21:08
stevan kungfuftr: are you doing 'make clean'?
that was not working right for a little while
(leaving Makefile around and such) 21:09
cognominal hum, I guess that must be ASCII only for the rist field
theorbtwo Right. And utf8, not latin1, for the last field. 21:10
shapr bonjour Ovid
stevan Ovid!!
Ovid Salut!
stevan Ovid: the ITypes refactor is (mostly) complete
so give your HTML::Entities a try 21:11
Ovid So, if Pugs completely borked right now? I can't even compile. Something about "Could not find module 'Plugins'"
kungfuftr stevan: running it manually from scratch now
stevan Ovid: blame theorbtwo :)
kungfuftr: excellent
Ovid OK, I blame him. Still doesn't get me a compile, though.
stevan Ovid: those are hs-plugins, they are needed for using Haskell libraries in perl6 21:12
theorbtwo: isnt there a flag Ovid can switch off?
stevan looks for the hs-plugins URL too
ninereasons we used to have a check builtinswitched off the flag automatically
it doesn't seem to be working, right now 21:13
stevan Ovid: grab this ftp://ftp.cse.unsw.edu.au/pub/users/dons/hs-plugins/snapshots/
theorbtwo What's the third char of your name, cognominal?
Ovid OK, I see the plugins. What do I do with 'em? 21:15
stevan install 'em
Ovid ? All?
stevan wait lemme look 21:16
ninereasons autoconfig && ./configure && make
cognominal theorbtwo: e acute
Ovid ninereasons: was that directed at me or someone else?
stevan Ovid: grab the latest one
ninereasons piped to you, Ovid
it was given to me as instructions
cognominal ...stōæ½xE9phane
stevan 2050412 21:17
ninereasons (but I haven't been able to build plugins on freebsd, so far)
stevan (damn australians livin in the future)
Ovid: then follow ninereasons instructions re: autoconfig, etc. 21:18
Ovid ok
stevan then make sure to 'make install' and then 'make register'
ninereasons Ovid, then make install and make register
theorbtwo OK, authors file fixed.
ninereasons oops
stevan as for what it is these things are actully doing,.. I will refer you too theorbtwo
hey metaperl 21:19
theorbtwo sighs.
Something you wrote breaks the build, and people are all over you like it's your fault.
Ovid No. The instructions say I need to start with ./configure, but there's no such file.
kungfuftr bah, no building... missing pugins (hehe)
ninereasons autoconf
stevan autoconf makes one
ninereasons are you using linux, Ovid ? 21:20
theorbtwo ./autogen.sh
kungfuftr huh?
Ovid Mac OS X
stevan Ovid: I had the same issue,.. (damn Australian Haskell programmers cant write instructions worth a damn.,.. *grumble* *grumble*)
ninereasons autogen.sh makes me smile :)
Ovid :(
Oh well. I gotta get back to work. I'll deal with this later. 21:21
shapr autoreconf && ./configure && make
then as root, make install && make register
Ovid OK, trying now. 21:22
Ovid appreciates the help.
shapr if it's Haskelly stuff, feel free to poke me
kungfuftr how comes plugins is not optional?
theorbtwo shapr: It is.
stevan shapr++ # the lone voice of sanity ...
theorbtwo kungfuftr: It's supposed to be.
Ovid Thanks. How long have you been doing Haskell?
shapr as long as the #haskell channel has existed 21:23
Registered: 3 years 49 weeks 6 days (23h 0m 11s) ago
Ovid Damn!
castaway It is optional, just the checker seems to not always work.. go fix it ,)
shapr My fiancee got me two books and I started learning.
kungfuftr Could not find module `Plugins':
use -v to see a list of the files searched for
(imported from src/External/Haskell.hs)
*** Error code 1
jabbot pugs - 1820 - Fix AUTHORS file.
kungfuftr (soz)
ninereasons doesn't like the sound of the "Damn!"
Ovid My only experience with a Haskell programmer was someone who sneered at me because I use Perl. I should email him a link to pugs! 21:24
shapr I'm self-employed, so I didn't exactly learn quickly, but I did have fun.
castaway works for me:
/usr/bin/perl5.8.6 -Iinc Makefile.PL
*** Inline Haskell support disabled. If you want dynamic loading
of haskell modules, please install the hs-plugins library,
ftp://ftp.cse.unsw.edu.au/pub/users/dons/hs-plugins/snapshots/ ,
dated 2005-04-10 or later -- remember to "make register" too!
Ovid Actually, the "damn!" was a good thing.
shapr Ovid: would that be someone on extremeperl? I surely hope not =)
Ovid No. An ex-girlfriend's brother.
kungfuftr *blink* works now
ninereasons oh a "Damn" of appreciation
not of frustration :)
kungfuftr no it don't... bellecks
metaperl oh Ovid, you know galois.com is in Oregon don't you?
shapr The more I learn about programming, the more I realize it's all the same, there's no reason to fight.
Ovid I was referring to Gauloise cigarettes. 21:25
metaperl they are a haskell shop
castaway plenty of reason, just none of it useful ,)
stevan shapr: thems fightin words boy!
Ovid Curious.
shapr stevan: haha
it's true though, I've realized there's not a huge difference between monads and OOP
PerlJam shapr: there are bunches of Perl/PHP/Python/etc. fanatics that need to learn that lesson.
metaperl i didnt see your coment regarding cigarettes. I was just telling you about a haskell shop in oregon
shapr Though the difference is enough that I'd rather use monads.
PerlJam (about no fighting that is)
shapr Ovid: tried Gitanes? 21:26
kungfuftr theorbtwo: it may be optional, but it may still require it to be there to check stuff... etc.
=0)
ah... 2 secs
Ovid Oh! I thought you were referring to perlmonks.org/?node_id=446761. It's a post I made today and I referred to Gauloise cigarettes. Curiously, it's about the disagreements on the extremeperl list.
shapr oh my 21:27
Ovid shapr: no, I haven't. However, I quite smoking about half a year ago.
Hey, it looks like Pugs is compiling now :)
s/quite/quit/
kungfuftr nope
shapr after my private email exchanges with Rob, I have discovered he's defending his position rather than looking for answers.
kungfuftr =0/
shapr kungfuftr: you have plugins problems? 21:28
kungfuftr yup
Ovid You know, I really can't blame Rob. I've been there (I'm still there at times. It's tough to let go of pride)
ninereasons is so glad not to be alone in his problems anymore
kungfuftr freebsd 4.11
shapr I'm still there at times too, but I have friends who tell me to "get the fuck over it" and I'm smart enough to listen.
ninereasons same here
Ovid :) 21:29
shapr kungfuftr: dons developers on bsd, so I'd be surprised if plugins don't work on one of the flavors.
theorbtwo So: Is the no-plugins case still broken for anybody, or did it mysteriously fix itself?
Ovid Me, I just learn to have friends who are smarter than I am. It makes me listed more.
shapr But also dons should be awake in an hour or so, and you can ask him then.
Ovid: yeah, that's one thing I like about #haskell
kungfuftr shapr: i have no plugins... it just won't build, though is supposed to be optional 21:30
ninereasons still broken for me, theorbtwo and for kungfuftr
Ovid has a problem too, theorbtwo , but know the details
don't know ..
shapr I don't even try to be prideful on #haskell, there's nothing like getting squished by a 15 year old who knows twice as much as I do about type theory.
Ovid Mine is fix.
ninereasons one down, then. theorbtwo 21:31
mj41 use v6; 21:33
my $conf;
$conf = { 'a' => 1 };
say $conf.keys.[0];
my $conf2 = { 'a' => 1 };
say $conf2.keys.[0];
# will print 0 a
jabbot pugs - 1821 - More debugging for plugins problem.
kungfuftr ninereasons: try uncommenting line 193 of Module::Install::Pugs 21:34
theorbtwo ninereasons, kungfuftr, anybody else, try svn up and re-running.
castaway goes to bed
kungfuftr pugs is building directly after removing that comment
shapr The smartest people seem to have very little pride, I guess they realize how much there is left to learn. 21:35
theorbtwo I'd prefer to leave the waitpid call commented out; it reportedly causes infinite runs...
Ah, damn.
PerlJam mj41: surely that's an old pugs you've got?
theorbtwo It reportedly causes infinite runs for some win32s.
mj41 r1819 winXP
stevan shapr: that may be because prideful smart people tend to say dumb things when their pride gets in the way (hence lowering their status) 21:36
shapr interesting thought
that would make sense
kungfuftr theorbtwo: do it based on OS
stevan shapr: I meet many a really really smart person in my dot.com days who's ego made them seem like a complete idiot 21:37
kungfuftr ninereasons: any luck?
ninereasons i tried svn up first, kungfuftr - that failed 21:38
trying your suggestion
stevan finds un-TODO-ing passing test highly satisfying
mj41 my $conf;
$conf = { 'a' => 1 };
say ~$conf.kv;
# prints: 0 a 1
theorbtwo ninereasons: What did it print when it failed?
PerlJam mj41: Hrm.
mj41: oh! *my* pugs is an older one.
ninereasons do you really mean the line with '1;' on it, kungfuftr ? (no comment there) 21:39
PerlJam anyway, the whole variable initialization thing has been problematic for a while and no one with the right tuits + knowledge has been able to fix it yet.
ninereasons stille "Could not find module `Plugins':" theorbtwo 21:40
theorbtwo What?
What prints when you do a Makefile.PL run?
ninereasons first line: 'ghc-pkg', 'describe', plugins at inc/Module/Install/Pugs.pm line 176. 21:41
kungfuftr ninereasons: inc/Module/Install/Pugs.pm - 193
theorbtwo ninereasons: Yes, keep going, please.
kungfuftr ninereasons: is a call to waitpid
ninereasons 63638 at inc/Module/Install/Pugs.pm line 178. 21:42
63638 finished: 0x0 at inc/Module/Install/Pugs.pm line 180.
then, writing Makefile for ...
weird, kungfuftr ; my line 193 is simply "1;" 21:43
kungfuftr ninereasons: up to dat copy?
ninereasons svn stat shows no alterations to working copy 21:44
and up to 1821
kungfuftr ninereasons: look for "#waitpid"
ninereasons maybe I out to wipe out the working copy and check out again ?
Ovid Anyone else experiencing the tests hanging on t/examples/golf.............................ok 1/10?
stevan Ovid: I had that happen too 21:45
ninereasons found it on line 179, kungfuftr ; it looks like ...
Ovid Hmm ... I still don't feel comfortable installing this yet :(
ninereasons yes .. it looks like ...
it's working. thanks kungfuftr 21:46
kungfuftr mwuahahah
ninereasons strange about the line number difference.
kungfuftr *shrug* 21:47
ninereasons *shrug*
stevan Ovid: r1822 - I skipped the golf test 21:48
Ovid Thanks. Trying again. 21:49
kungfuftr theorbtwo: looks like the waitpid needs to be there for bsd
ninereasons a different problem in Ovid's case though. I think his problem went away when he got plugins to go in successfully. 21:50
(don't know his OS) 21:51
Ovid That's correct.
Mac OS X
ninereasons interesting
21:51 metaperl_ is now known as metaperl
kungfuftr memory leak in golf? 21:52
t/examples/golf.t
pugs: out of memory (requested 1048576 bytes)
stevan kungfuftr: when I interupted it,.. I found the process was still going in the background 21:53
svn up though,.. i skipped it
kungfuftr ah... k 21:54
theorbtwo tries one more checkin, sighs, declares it bed. 21:56
the the finished: foo line does not give 0x100 if you don't have it, or 0x0 if you do, then something odd is going on. 21:59
s:1st/the/if/
jabbot pugs - 1823 - One more try. 22:03
pugs - 1822 - more test cleaning and force-TODO-ing
ninereasons that is interesting, mj41. " $conf = { 'a' => 1 }; 22:15
$conf[0].kv;
('a', 1)
$conf.perl;
'((\'a\' => 1))'
$conf.kv 22:16
(0, ('a' => 1))
jabbot pugs - 1824 - More progress filling out S28draft.pod 22:23
dvergin__ always loves hearing from jabbot. 22:24
Ovid I still think there are serious problems here. I am now hanging on "t/subroutines/subroutine....................NOK 13"
Don't make me learn Haskell and fix these damned things! 22:27
(just kidding)
shapr tokes on a lambda 22:28
the first one comes free man
kungfuftr shapr: that's right... get him hooked!
Ovid muses trading Prolog lessons for Haskell lessons.
shapr grins 22:29
I give Haskell lessons for free.
Deceptively free... I think open source is resource competitive, and the resources are the minds of the developers. 22:30
Juerd Hey, Ovid! 22:37
We're online at the same time. That's special.
Ovid Hi Juerd. 22:38
Juerd Do you grok my explanation of the := swap thing?
Ovid Yes, but I still think it will confuse programmers. Several folks here were confused by ($x,$y) := ($y,$x) and there was disagreement over what it meant. 22:39
Juerd (Though it only makes sense in the sane world where references are references, and names, containers and values are separated from eachother, and from where Larry left a month or so ago) 22:40
Ovid: It's no more confusing than with s/=/:=/
Ovid: It does the same thing, but swapping the variable names while keeping the values in place, instead of exactly the other way around: swapping the values while keeping the variables (and their names) in place
The larger the value (possibly megabytes in size), the greater the benefit. 22:41
Ovid Well, aliasing and binding is good in the general sense, but trying to swap with aliases suggests to me that variables were incorrectly used (or named) in the first place. 22:42
Juerd swapping with = is like working with cp on the command line
:= is mv.
Ovid: They were either incorrectly used/named or equivalent in meaning in either case
The end result, in the language, as seen by the programmer, is exactly the same. 22:43
This is why this example is generally done with $x and $y, or $foo and $bar and never with meaningful names, because this thing just doesn't happen much in practice.
Ovid Hopefully this won't cause a problem. I'll worry later as "idiomatic Perl 6" starts to evolve. 22:44
:)
Juerd I think := is a big and important thing.
It's a key element to performance - something really missing in Perl 5. It's only available implicitly. 22:45
Re idioms: many will go away, because all the often used idioms are going to be part of the language itself
We lose the ST because sort itself now caches 22:46
The selectselect idiom is gone because, well, $/ is gone :)
(And $\, $|, etc)
slurp does away with @ARGV = $fn; <>
Ovid Oh, don't get me wrong. I agree that := is really important. I just am concerned about an odd little corner case with it where reasonable programmers have had different expectations. 22:47
Unreasonable programmers may be even worse.
Juerd I think explaining them both is a good way to demonstrate how := differs from =
They both swap, only the thing that's swapped is different. 22:48
I think the swap case can more clearly explain the difference than normal assignment with many print statements to show what's happening.
When explained right, at which I certainly failed :)
kungfuftr Ovid: t/subroutines/subroutine.t is hanging for me too 22:57
Ovid Bummer. 23:02
Yeah, I've run it twice. This is probably a similar problem to the examples/japh that stevan skipped. There's something serious there :(
Juerd Note that many drivers have no way of knowing what data sources might be available for it. These drivers return 23:10
an empty or incomplete list or may require driver-specific attributes.
Sorry, wrong window.
There is also a data_sources() method defined for database handles.
shapr grabs the handle of a nearby database 23:11
jabbot pugs - 1825 - skipping tests in subroutine.t 23:14
stevan Ovid: subroutine hang is fixed
this was a parsefail which was recently uncommented 23:15
kungfuftr stevan: beat me to it... dammit
Ovid OK. Updating now.
stevan so not something terribly serious I think
kungfuftr: beware my svn fu !!!!!!
stevan is dreading the Changelog duty given the number of commits today 23:16
shapr consiers making stevan's life harder. 23:17
stevan shapr: go ahead, make my day ;| # <<< clint eastwood squint 23:18
shapr grins
kungfuftr makes lots of one character changes to files
Juerd Oh, the changelog loves that. 23:24
stevan hmmm,... bug or correct 23:25
my @array3 = (@array1, @array2); say +@array3 # prints 2
./pugs -e 'my @array1; my @array2; my @array3 = (@array1, @array2); say +@array3' 23:26
shouldn't that flatten @array1 and @array2?
Juerd It should, as the RHS of ARRAY = should be in list context 23:33
Not list-of-scalar-contexts :) 23:34
stevan Juerd: I assume that this: my @array5 = ( @array2[2, 1, 0], @array1[2, 1, 0] );
is the same thing then
Juerd That's only for sub arguments and binding, except when slurpy...
Same thing as what?
stevan it should flatten as well since the slices would return arrays 23:35
Juerd +@array5 should be 6, and not grep defined, @array5
Slices return lists 23:36
Not arrays
Juerd zzz
Good night
stevan thanks Juerd
jabbot pugs - 1827 - force-TODO-ing and test cleanup 23:53
pugs - 1826 - force-TODO-ing and test cleanup
Ovid Only 190/4236 subtests failed :) 23:59

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!