crysflame a period after was not would be nea 02:12
i don't think the comma is required, unless you're saying it like "That part of my car exploded, though, was not"
as the commas are used to interjecct. was not isn't an interjection. yey elements of style.
Steve_p heh, comma-talk continues :) 02:41
Odin-LAP I thought comma had already been implemented? :D 02:42
crysflame heh 02:47
autrijus hey. 02:56
hm. I should make it crystal clear that Pugs wants ActivePerl/MinGW's Perl 03:02
and not CygPerl. 03:03
Odin-LAP How strong of a REPL loop thingy is Pugs going to have? :> 03:05
autrijus I'd like eventually to implement most of hugs commands :) 03:06
see Shell.hs
in particular, :l and :e will be helpful 03:07
as would :b
patches welcome :)
Odin-LAP Hm.
Odin-LAP ponders going off on a (or yet another) tangent and learning Haskell... 03:08
autrijus you can read a few pages of YAHT per day :)
Odin-LAP Hm? 03:09
autrijus see "Where can I learn more about Haskell?" 03:10
svn.perl.org/perl6/pugs/trunk/docs/...rview.html
crysflame autrijus: hmm, yep.
autrijus yaht = yet another haskell tutorial
crysflame crystal clear.
autrijus yeah. it now dies if $^O eq 'cygwin'. 03:11
crysflame heh, neat
why cygwin bad?
^is
autrijus not bad, just lack of a porter for ghc
hence, bitrotted
clkao morning autrijus
autrijus ghc is more MinGW minded. 03:12
crysflame ah.
autrijus hi clkao.
crysflame does ghc compile under cygwin?
autrijus not sure. never tried.
going to airport now. 03:20
hopefully resume network in a hour
autrijus waves &
scw Hello, I've just maintioned that thet word 'Synopses' in home page and overview should be a miss-spelling of 'synopsis'? 05:37
obra that's a plural, no? 05:38
mdiep yes, synopses is plural of synopsis
scw hmm... but, however, the url uses synopsis as well.
obra and? the url points to a singular. 05:41
scw: as a native speaker, it looks sane to me.
scw obra: Oh! Then that's fine :)
obra heh. (you'll see that perl.org uses synopses in the text as well ;) 05:42
scw (It seems that I should be more concentrated in English class...) 05:44
obra what's your native language, scw? Do note that 'synopses' isn't very common, even for native speakers 05:46
scw obra: Chinese, I'm from Taiwan.
obra an. XD will you be at yapc.tw? 05:47
mdiep also, hypotheses is the plural of hypothesis
scw obra: Sure! I've registered and payed!
obra cool. I'll meet you there
scw Are you one of the speakers? 05:48
obra <- Jesse Vincent (yes) 05:49
scw obra: Oh! You talked (and will talk) about RT! I participated YAPT.tw last year, too! 05:51
obra ah :) 05:53
Yes, RT is my baby
mugwump young child more like ... it'l be a teenager before you know it 06:06
so speaking of things in double digits, has anyone tackled pugs tests for Synopses 10+? 06:14
autrijus rehi from HKG! 07:31
clkao happy? 07:32
autrijus the whitespace problem introduced by luqui's qwinv patch has been fixed.
was very silly
$x < $y > $z
# parsed as $x{'$y'} $z
chained comparison now works again. 07:33
looking thru the test failures now
sig, ($x-- + $x--) is calculated thrice. 08:05
autrijus moans the lack of referential transparency
fixing.
also, hi abstraction, edgewalk1r :)
abstraction hey 08:11
autrijus how did you find here? 08:12
<- preparing 6.0.9 release now
abstraction #perl
autrijus ahh.
autrijus goes there 08:13
abstraction so how much of the known perl6 syntax has been implemented in pugs? 08:14
autrijus class and objects is missing; rules is missing 08:15
the rest is pretty much all there
but there's lots of corner cases.
so we're porting perl5's unit tests here
to ensure that we have a solid foundation 08:16
and then build class/rules on them.
abstraction good stuff
autrijus oh, and "unpacking" in subroutine and := binding is not yet there.
but will be soon :)
yeah. it's fun.
so, got time to help out? 08:18
currently it's mostly porting $your_favourite_perl5_code to perl6, and/or write tests. 08:19
abstraction I assume I'm going to know haskell?
autrijus you don't have to if you don't want to :)
you can just treat pugs as a black box that happens to run some perl6 code...
...and when it does not, yell it here so we can fix
(see "Overview" in pugscode.org) 08:20
abstraction sounds like a good way to get started with perl6 syntax
autrijus I certainly hope so :)
abstraction now we just need a haskell to pbc util :) 08:21
autrijus see "Compiler backend" part in the overview :)
but it's more important to be Correct in treatment of synopses at this moment. 08:22
once the basic stuff is solid, we can then have a pretty good idea of how to generate parrot AST.
and once that's done, pugs can be ported to perl6 and finish the bootstrap.
(of course, if PGE/P6C is ready by then, we won't need to do the bootstrap here.) 08:23
hi sorje.
theorbtwo (Or we could bootstrap the other way, and make a haskell to parrot compiler. More work, but could be interesting... at least to somebody that actually knows haskell.) 08:24
autrijus yup. if people want to do that, look at GHC.Net
Core->pasm is certainly interesting. 08:25
(Core is GHC's internal language)
it's far larger in scope though.
it will also help if someone interested in C can look at parrot/ast/. 08:26
and report back what can be done with it to make Pugs AST align with parrot AST 08:27
sorje hi autrijus
autrijus hey.
because it looks like parrot ast has a builtin ast->pasm compiler 08:28
if that is the case, we may no longer have to write any compiler at all.
just emit parrot ast and let parrot take over.
scw autrijus: Hello, when I was writing Advanced calculus homework, some part of the overview was translated to Chinese XD 08:38
autrijus: w.csie.org/~b92103/01Overview.zh.html
autrijus ooh! 08:39
viirya: coordinate! :)
scw++
scw autrijus: The translation is not very smooth, however...
autrijus please translate "Apocryphon" as "ōæ½xA5~ōæ½xA8ōæ½xE5", i.e. "Pugs ōæ½xA5~ōæ½xA8大ōæ½xA4@: Pugs ōæ½xADpōæ½xB9ōæ½xBAōæ½xB7ōæ½xA7ōæ½xBBōæ½xA1" 08:40
scw ok.
gugod is this in only html format ? we could put it as a aprt of p610n 08:42
s/aprt/a part/
autrijus gugod: POD can't handle named links
gugod: if you will translate it to Kwid, I'd be very very grateful 08:43
gugod right
autrijus # svn.openfoundry.org/pugs/lib/Perl6/...lkwid.kwid
gugod that's actually my idea
autrijus and once it's in kwid, then sure, it can be part of p6l10n
*nod*
gugod: HTML2Pod may be useful 08:44
on cpan
i.e. hack it a bit and let it emit kwid instead
gugod right 08:46
autrijus hm. 08:50
can any Val be used as hash key?
in p5 we stringify all keys
can p6 use arbitary values?
currently it is a problem because Int 2 and Num 2.0 and String '2' can point to different values
which is clearly broken
jiing scw: s/ōæ½xA8ōæ½xD1ōæ½xC4m/ōæ½xB0^ōæ½xC4m/ 08:51
autrijus I'll use Hash as Str keys only for now. 08:52
viirya autrijus: hum? 08:53
autrijus viirya: see scw's url above -- the text needs some work but would be a good start
just don't want you two to duplicate work, that's all :) 08:54
asked the hashkey question on p6l.
viirya autrijus: I have done part of it. :)
abstraction can p6 hash keys be anything? string, int, obj, etc? are are they supposed to be stringified?
autrijus abstraction: I can't find it in the synopses.
abstraction: pugs is currently Anything
but it creates a problem because Str "2" and Int 2
can point to different keys
which is I think not what most people will accept 08:55
abstraction but Str 2 and Int 2 are different
autrijus they are.
abstraction makes sense to me then
autrijus my %hash = (1..4); %hash<1> # undef, wtf?"
or, even worse
my %hash = (1..4); %hash{1.0} # undef? 08:56
abstraction p5 people aren't used to types
autrijus so very true.
abstraction but how often do you have a hash with various types? 08:57
er mixed types
autrijus well, in typed languages
you can restrict the key type
abstraction probably not very often
autrijus so it's usually not a problem.
but by default, keys are just Scalars
abstraction ah
autrijus maybe we can restrict hash keys to be unified type. 08:58
it's slightly bizzare.
well, very bizzare and radical.
abstraction 1 type to rule them all.
autrijus and that is the Scalar type? ;)
CDybedahl Synopsis 9 says: 09:14
To declare a hash that can take any object as a key rather than just a string, say something like:
my %hash is shape(Any); 09:15
Seems to me that means that hash keys are strings by default.
I guess you could argue that if you ask for any sort of keys, you're also asking for "2", 2 and 2.0 to be treated as different. 09:17
autrijus oh. 09:19
is shape().
okay, thanks!
CDybedahl++
string interpolation bug finally tracked down.
luqui was using prefix:<~> to concat strings. fancy that :-)
mm my passport was left on the plane. 09:22
mm but they found it.
autrijus about to board to Beijing
see you in ~3hrs, hopefully with 6.0.9 tarball :)
autrijus waves &
abstraction later autrijus 09:23
CDybedahl This command: pugs -e 'for (1..4) -> $i { if (($i%2)!=0) {say $i} }' 09:25
...prints 1, 3 and 4. Surely that's not right?
If you change != to == it doesn't print anything at all.
abstraction that't not right 09:26
luqui poke autrijus 09:28
compare difference between: my @a = (1,2,3); "@a" 09:29
and my @a = (1,2,3); " @a"
re: changes in r325. we need to give the whole string string context somehow... 09:30
viirya autrijus: www.viirya.org/01Overview.zh.txt 10:06
autrijus: it's early translated version. I just do half of it. You could check first. 10:07
autrijus: If not urgent, I will finish the rest in next two days. 10:08
rgs makes an rpm of pugs 13:44
Juerd At the DPW, pugs was mentioned a lot of times :) 13:46
rgs it's a hot topic at fosdem too
Juerd It's great to see that everyone's so up to date 13:47
scw jiing: fixed, thanks :) 14:03
viirya: WOW! Your translation is wonderful! 14:05
viirya scw: thanks, just made half of it. 14:07
scw viirya: In 'What is the Perl 6 bootstrapping problem?', you translated 'Perl 6 Rules' into 'Perl 6 ōæ½xB3Wōæ½xABh' but leave the same word untranslated in 'Why did you choose Haskell?' 14:15
viirya scw: I will corrent it, it should not be translated. 14:18
scw viirya: It seems leave it untranslated is the best solution ^^" 14:19
viirya scw: yes. 14:22
the error is updated, and a bit of translation added. 14:46
&
Steve_p Hmmm...tests for pairs seem broken 14:53
autrijus there's no wireless in the hotel :-/ 14:59
the latency is >2snd on dialup :-/
>2seconds even 15:00
scw autrijus: So svk rules? XD
autrijus I'm afraid so
clkao great open source conference venue 15:03
wolverian autrijus: hey, I haven't yet asked that pattern question on p6l, I'll read the pattern guards paper before doing that. (just in case you were wondering) 15:11
15:24 edgewalk1r is now known as edgewalker
autrijus ok, cool 15:28
clkao: conference begins tomorrow morning; then we'll have wifi
hopefully will extend into the night... this is not even 14.4kbps
about 9600 baud
shapr Hey, did you send me the pdf already?
autrijus has a deja vu feeling of telix and trumpet wind^Hsock
shapr haha, winsock :-) 15:29
Been a few years since I've used trumpet winsock.
Sure was a cool feeling the first time I got it working...
autrijus shapr: aut.dyndns.org/tmp/01Overview.pdf 15:31
shapr yay thanks!
autrijus but you can also just convert from tml...
I wonder if there is a way to translate html to pdf while preserving the hyperlinks.
distiller maybe, but I don't have a copy of that.
OOo certain ^Hly doesn't do that.
shapr I would very like that. I don't want to do latex again, I want wiki publishing with optional pdf/ps output.
autrijus riiight. I have a PDF::FromHTML module on CPAN. 15:32
but it doesn't grok hyperlinks yet.
wolverian: sure, take your time. :) thanks!
viirya++ # wonderful translation. i'm in no rush, so take your time :) 15:33
17:00 metaperl_ is now known as metaperl
autrijus ew^W^Wwhew. 17:09
All tests successful, 4 subtests skipped.
Files=40, Tests=616, 26 wallclock secs ( 0.00 cusr + 0.00 csys = 0.00 CPU)
commit takes forever...
sigh. can't release anything under this bandwidth 17:10
will have to wait till tomorrow morning :-)
wolverian dialup? people still use those? :/ 17:14
autrijus and this is supposed to be one of the better beijing hotels :-/ 17:15
(conference venue)
wolverian vandalize the minibar in retaliation. 17:17
obra how censored is your network? 17:18
autrijus I don't know because I tunnel everything thru ssh
obra I suppose you could do everything over ssh -C -D
autrijus nor do I want to know, really :)
obra heh
statico finds a censored internet kind of creepy 17:26
autrijus ^H^H^Hjournal up. 17:27
Khisanth hmm
autrijus see you tomorrow when I got more bandwidth :)
Khisanth autrijus: why are there some funny letters comming up from some of your lines?
autrijus Khisanth: extremely high latency renders control chars as literals 17:28
Khisanth ah
autrijus or, if you are talking about Big5 code, then they are big5 code.
you can usually tell one from another by checking for double quotes around them :) 17:29
Khisanth well it's erm 0x52 and 0x7F
autrijus oh. so definitely control code
'nite! & 17:30
wolverian good night! 17:31
integral I can't seem to find how to read a single line from a while, S02 just mentions using * with for to read all the lines 18:00
anyone know where I can be enlightened?
s/while/file/ 18:01
nnunley integral: That would be how -- for *$handle { block is called for each line in file handle with $_ containing the current line } 18:05
So... I would guess that $line = *$handle would only fetch a single line. 18:06
Just as $line = <FILE> does, unless you've turn read mode slurpy. 18:07
integral hmm, ok, but...
also in S02, it mentions that the * unary op forces list context on its arg 18:08
nnunley reads. 18:10
Might require reading A02
integral ah, it says FHs are just iterators now 18:12
Steve_p Ah! Better. 18:59
Steve_p hopefully provides an extinguisher for flames in ingy's journal
nnunley steve_p: You're steve_p on use.perl, too? 19:00
clkao speters i guess? 19:01
Steve_p yep 19:02
ingy hola 19:12
I got flamed?
pjcj 4/las ingy 19:13
bah ...
statico hi ingy 19:17
ingy Steve_p: nice comment 19:19
hi statico
sooo.... I finished and checked in perlkwid.kwid 19:20
anyone care to review it for me?
Steve_p ingy, Ovid took offense 19:21
ingy Steve_p: yeah, I think he was too wired into a particular mode 19:22
that's ok
clkao because he manages grants?
ingy Allison grokked me I think. So I'm not worried 19:23
statico on kwid?
Steve_p clkao, on money and being an open source developer with a family
ingy statico: use.perl.org/~ingy/journal/23384 19:24
clkao *nod*
ingy statico: can you review my perlkwid spec?
clkao added pugs to ProjectsUsingSVK
Steve_p I prefer the "Steal Money" method of grants 19:25
statico ingy: sure
statico agrees with the ten points 19:29
Steve_p ingy, did you kill .over or do I need to read more
ingy Steve_p: .over is now .list 19:37
Kwid ends up being significantly different from Kwiki 19:38
In many ways it improves on Kwiki
In some ways it has to deal with Pod legacy
but Pod legacy isn't necessarily bad 19:39
Pod is a rather nice language, and Kwid makes it much nicer to read and work with 19:40
I really like the {*bold*} {/italic/} {`code`} constructs 19:41
as bullet proof versions of the common *bold* /italic/ `code` 19:42
integral can you put unbalanced {s and }s in those? 19:43
ingy sure
integral :-)
ingy the ending marker for '{`' is '`}' 19:44
integral so can you put `} in there?
ingy so anything else goes
{{``}`}} is one way
integral ah, that's a good way
ingy {`\``} is the other 19:45
a single backtick as code is {```} 19:46
oops 19:47
{`\`}`} is the other
:) 19:48
backticks for code is sooooo nice
also there is the '{ ', ' }' markers 19:49
{ don't *format* anything in /here/ } 19:50
aka the 'asis' format
(it also makes whitespace be non-breaking) 19:51
integral sounds good :-)
ingy Backslash at the end of a line overrides folding
which is logical
it is an escaped newline 19:52
Kwid comments are handy too 19:56
20:44 edgewalk1r is now known as edgewalker
ingy autrijus: I got your copy of Perl 6 Now 20:44
autrijus: hopefully it will be coffee stained and mangled by the time you get it
obra oh, ingy, my flights changed. I'll be arrivign earlier by a couple days 20:50
ingy obra: they changed, or you changed them? 20:59
wolverian hm, okay, I think I finally understand pattern guards. 21:07
new terminology makes simple things hard. 21:08
integral hmm, seems someone's found the == problem 22:07
buu Hrm. I need ghc
integral buu: www.haskell.org/ghc/download_ghc_62...ml#windows 22:08
buu The pugs source doesn't need a C compiler does it?
integral: It doesn't run under cygwin properly =/
integral buu: on Intel GHC directly generates code
buu: no, that's why you run it natively
buu Yeah, but then it doesn't work with cygwin's perl
Whic is just a pain
integral well use your other perl 22:09
buu I don't have two perl's at the moment
integral oh
buu great 22:14
I installed ActivePerl, ran the makefile.pl, got a makefile, ran nmake and get 'invalid argument' 22:15
Which version of nmake am I supposed to have?
That's annoying =[ 22:20
Juerd How do people drive automatic cars and not fall asleep? 22:39
And don't you terribly want something to do for your left foot?
mugwump at least that means they both get stiff on a long drive not just the right foot 22:41
s/foot/leg/
need pedals down there to get some exercise while driving 22:42
wolverian get a bike? :) 22:43
Juerd mugwump: I don't care what the pedal's for, but let my left foot do something
If even pump up the airbag
Give it something to do
rgs listen to some jazz and beat the rythm.
Juerd For example
This temporary replacement car doesn't have audio
Which sucks
mugwump I go mad without lots of breaks 22:44
rgs ingy: your journal entry made me laugh. It's so you.
mugwump tho if I don't have any good breaks a bit of trance or even D'n'B will suffice 22:49