|
pugs.blogs.com/dist/Perl6-Pugs-6.2.13.tar.gz released! | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com Set by audreyt on 17 October 2006. |
|||
|
00:11
glasser joined
00:37
systems`` joined
00:47
vel joined,
systems`` left
|
|||
| audreyt | xinming: your .close change somehow broke 200+ tests... | 01:01 | |
| lambdabot | audreyt: You have 1 new message. '/msg lambdabot @messages' to read it. | ||
| audreyt | I'll take a look after nap | ||
| @messages | |||
| lambdabot | lanny said 2h 54m 50s ago: You'll want to check the smokes for Win32. r14463 (just before GHC threading turned on) has 2 failures and after 231 (including having to kill pugs by hand on 2 test to | ||
| continue). | |||
| audreyt | @tell lanny it's not me, it's xinming... | 01:02 | |
| lambdabot | Consider it noted. | ||
|
01:05
ferreira left
|
|||
| Limbic_Region | audreyt - what's your next timezone anyway? | 01:09 | |
| audreyt | brazil | ||
| Limbic_Region | perlbot .bz | 01:10 | |
| perlbot | .bz is Belize | ||
| eviltwin_b | .br :) | ||
| Limbic_Region | ah | ||
|
01:23
justatheory joined
|
|||
| audreyt | repaired | 01:27 | |
| svnbot6 | r14480 | audreyt++ | * Pugs.Types: IO::Dir is not a subtype of IO, really... | ||
| r14480 | audreyt++ | * Pugs.Prim: Socket::close and IO::close should be two different things. | |||
| audreyt | @tell xinming never mind, fixed | ||
| lambdabot | Consider it noted. | ||
| eviltwin_b | don't think you @tell-ed him to start with; only a snark to lanny | 01:29 | |
| audreyt | that's correct. | 01:32 | |
| @tell lanny never mind; fixed :) | |||
| lambdabot | Consider it noted. | ||
| audreyt | sorry for the snarky tone; intense sleep deprivation does that for a person... | 01:33 | |
| eviltwin_b | enh. actually I think I was being snarky there (but same comment; was up until 3am yesterday then got up earlyish today. mmm, weather futzing with my head) | 01:34 | |
| xinming | audreyt: sorry for my rudeness about introducing that bug... >_< | 01:36 | |
| lambdabot | xinming: You have 1 new message. '/msg lambdabot @messages' to read it. | ||
| xinming | @messages | ||
| lambdabot | audreyt said 9m ago: never mind, fixed | ||
| audreyt | xinming: no, it's actually not your bug; it's that IO::Dir was incorrectly considered a subclass of IO | ||
| but that's wrong because $dir.say doesn't do anything meaningful | 01:37 | ||
| your change is fine :) | |||
| xinming | en, thanks. | ||
| audreyt | (it just revealed an existing bug) | ||
| xinming | BTW | ||
| while we pass the variable to perl 5, is $a sames as \$a while we pass it to perl 5 please? | 01:38 | ||
| hmm, I mean from Pugs | |||
| The bugfix you fixed yesterday for me seems treat $a and \$a are the same | 01:39 | ||
| my $a; $t.process("hello", $v, $a); which would act as $t.process("hello", $v, \$a); | 01:40 | ||
| audreyt | well it depends on what the other side wants | ||
| if it wants something it can write back (as is the case for .process) | |||
| then it's rw | |||
| in which case $a will work too | 01:41 | ||
| I think, anyway | |||
|
01:43
prism joined
|
|||
| xinming is confused about this small difference... | 01:44 | ||
| so, you mean, $a is just an alias passing to the perl 5 program, right? | 01:45 | ||
| audreyt | right | ||
| xinming | and by default, it's ro, But we need to write to it, | ||
| so, It's an rw alias. | |||
| hmm, not sure about this. >_< | 01:46 | ||
| audreyt | basically in perl5 land, when we have a pugs value | ||
| if you teat it as a scalar ref | |||
| then it behaves as a lvalue, that is you can assign to it | |||
| and because in Pugs, $x=1 and $$x=1 and $$$x=1 and $$$$x=1 means the same thing | 01:47 | ||
| (if $x is a simple uninitialized scalar variable) | |||
| the extra \ got canceled away | |||
| but my mind isn't working really right now; I should probably sleep instead of offering potentially misleading/bogus explanation :) | 01:48 | ||
|
01:51
lanny joined
|
|||
| xinming | audreyt: Just go and sleep. :-) | 01:51 | |
| lanny | Hiya. Sorry to point at your stuff first, audreyt. I had reviewed xinming's changes and they seemd trivial so... | ||
| lambdabot | lanny: You have 2 new messages. '/msg lambdabot @messages' to read them. | ||
| lanny | Thanks for the every-so-quick-fix. You are gonna burn yourself out at this pace. :) | 01:52 | |
|
01:52
SubStack joined
|
|||
| lanny | @messages | 01:52 | |
| lambdabot | audreyt said 50m 5s ago: it's not me, it's xinming... | ||
| audreyt said 19m 31s ago: never mind; fixed :) | |||
| audreyt | lanny: not likely :) been like this for 1.5+ years now :) | ||
| lanny | Let me just say then that I'm glad you are using your powers for good instead of ebil. | 01:53 | |
|
01:54
LimbicRegion joined
|
|||
| audreyt | I'm using it for fun, actually, but thanks :) | 01:54 | |
| audreyt fades into sleepland & | |||
|
01:54
LimbicRegion is now known as Limbic_Region
02:15
thepler joined
02:25
thepler joined
02:57
weinig is now known as weinig|sleep
03:07
avarab joined
03:09
avarab is now known as avar
03:19
avarab joined,
avarab is now known as avar
03:30
avarab joined
03:42
avarab_ joined
03:51
avar joined
04:02
Eidolos joined,
avarab joined
|
|||
| svnbot6 | r14481 | lanny++ | * t/regex/p6regex/p6regex.t -- stab at un-todo of unexpected passed tests | 04:04 | |
|
04:58
root4o joined
|
|||
| root4o | hi, { ... } is not implemented yet , right ? | 05:00 | |
| eviltwin_b | L<S06|The &?BLOCK object> contains a code snippet using ?? :: (instead of ?? ||) | 05:01 | |
| ?eval sub x {...}; x() | |||
| evalbot_r14481 | *** ... - not yet implementedā¤ at <eval> line 1, column 8-11 | ||
| eviltwin_b | looks that way | ||
|
05:02
MacVince joined
|
|||
| eviltwin_b | hm, actually that looks like it DTRT: {...} means "this sub is not yet implemented" | 05:02 | |
| ?eval sub x {!!!}; x() | 05:03 | ||
| evalbot_r14481 | Error: !!! - not yet implemented | ||
| eviltwin_b | ?eval sub x {???}; x() | ||
| evalbot_r14481 | Error: Cannot cast from VUndef to Handle (VHandle) | ||
| eviltwin_b | whoops, that wan't what I meant | ||
| MacVince | Hi all! | 05:04 | |
| root4o | 10x | ||
| eviltwin_b | huh, yes it was, but {???} apparently isn't implemented | ||
| ?eval ... | 05:05 | ||
| evalbot_r14481 | *** ... - not yet implementedā¤ at <eval> line 1, column 1 - line 2, column 1 | ||
| root4o | this also doesnt seems to work ? right... | 05:06 | |
| my $args1 = \(1,2,3); | |||
| eviltwin_b | huh? | ||
| root4o | foo(|$args1) | ||
| eviltwin_b | the point is that {...} *means* "not yet implemented" | ||
| root4o | i'm getting | 05:07 | |
| eviltwin_b | oh, you're on to something else | ||
| root4o | pugs: don't know how to toVV': VList [VInt 1,VInt 2,VInt 3] | ||
| but | |||
| $args1 = \{ x => 5 }; | |||
| works | 05:08 | ||
| eviltwin_b | ?eval my $x = \(1,2,3); sub x ($a, $b, $c) {say "Hi!"}; x(|$x) | 05:15 | |
| evalbot_r14481 | Error: No compatible subroutine found: "&x" | ||
|
05:15
BooK joined
|
|||
| eviltwin_b wonders if he's too tired to be trying to figure this out | 05:16 | ||
| root4o | ?eval my $x = \(1,2,3); sub x ($a, $b, $c) {say "Hi!"}; x(|$x) | 05:17 | |
| evalbot_r14481 | Error: No compatible subroutine found: "&x" | ||
| root4o | eval my $x = \(1,2,3); our sub x ($a, $b, $c) {say "Hi!"}; x(|$x) | ||
|
05:17
nothingmuch joined
|
|||
| root4o | ?eval my $x = \(1,2,3); sub x ($a, $b, $c) {say "Hi!"}; x(|$x) | 05:17 | |
| evalbot_r14481 | Error: No compatible subroutine found: "&x" | ||
| eviltwin_b | you need the leading ? | ||
| root4o | yes i got that :) | ||
| eviltwin_b | yeh, I'm slow :) | 05:18 | |
| root4o | ?eval my $x = \(1,2,3); our sub x ($a, $b, $c) {say "Hi!"}; x(|$x) | ||
| evalbot_r14481 | Error: No compatible subroutine found: "&x" | ||
| root4o | ?eval my $x = \{a => 5}; sub x ($a) {say "Hi!"}; x(|$x) | ||
| evalbot_r14481 | OUTPUT[Hi!ā¤] Bool::True | ||
| root4o | eval my $x = \{a => 5}; sub x ($a) {say "Hi!"}; x(|$x) | ||
| this works | |||
| pairs also doesnt seems to work | 05:19 | ||
| eviltwin_b | ?eval my @x = \(1,2,3); sub x ($a, $b, $c) {say "Hi!"}; x(|@x) | ||
| evalbot_r14481 | Error: No compatible subroutine found: "&x" | ||
| root4o | when I type '?' does ppl on the irc see it ? | 05:20 | |
| eviltwin_b | yes | ||
| root4o | good | ||
| :) | |||
| i'm tring to find a thing to write test, for... but it is more harder to find something than to write the test ;) | 05:21 | ||
| eviltwin_b | ?eval my $x = \(1,2,3); say $x.perl | ||
| evalbot_r14481 | OUTPUT[\CaptSub {c_feeds = [MkFeed {f_positionals = [VPure (IFinite 1),VPure (IFinite 2),VPure (IFinite 3)], f_nameds = fromList []}]}ā¤] Bool::True | ||
| eviltwin_b | ?eval my $x = \(1,2,3); say $x | 05:22 | |
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| root4o | and then when something dont work ;) | ||
| eviltwin_b | ?eval my @x = \(1,2,3); sub x {say "Hi!"}; x(|@x) | ||
| root4o | yes it has to be Capture ,right ? | ||
| evalbot_r14481 | OUTPUT[Hi!ā¤] Bool::True | ||
| root4o | hmm :) | ||
| eviltwin_b | yes, it needs to be a Capture | ||
| root4o | so @ instead of $ | 05:23 | |
| eviltwin_b | ?eval my $x = \(1,2,3); say $x | ||
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| root4o | ?eval my @x = \(1,2,3); sub x (@a){say "Hi!"}; x(|@x) | 05:24 | |
| evalbot_r14481 | OUTPUT[Hi!ā¤] Bool::True | ||
| eviltwin_b | ?eval my $x = \(1,2,3); sub x (Int $a, Int $b, Int $c) {say "Hi!"}; x(|$x) | ||
| evalbot_r14481 | Error: No compatible subroutine found: "&x" | ||
| root4o | ?eval my @x = \(1,2,3); sub x (@a){say @a}; x(|@x) | ||
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| eviltwin_b | that still seems wrong, the point of unary | is that it's supposed to unwrap the array | ||
| root4o | yes | 05:25 | |
| eviltwin_b | ?eval my $x = \(1,2,3); sub foo (Int $a, Int $b, Int $c) {say "Hi!"}; foo(|$x) | ||
| evalbot_r14481 | Error: No compatible subroutine found: "&foo" | ||
| root4o | it gets capture | ||
|
05:25
avarab is now known as avar
|
|||
| root4o | but seems it can not unwrap it ! | 05:25 | |
| ?eval |(1,2,3) | 05:26 | ||
| evalbot_r14481 | (1, 2, 3) | ||
| eviltwin_b | ?eval my $x = \(1,2,3); sub foo (Int $a, Int $b, Int $c) {say "Hi!"}; foo(*$x) | ||
| evalbot_r14481 | Error: ā¤Unexpected "x"ā¤expecting term postfix, operator, ":" or ")" | ||
| eviltwin_b | ?eval my $x = \(1,2,3); sub foo (Int $a, Int $b, Int $c) {say "Hi!"}; foo(|($x)) | 05:27 | |
| evalbot_r14481 | Error: No compatible subroutine found: "&foo" | ||
| root4o | ?eval $x = \(1,2,3); say *$x | ||
| evalbot_r14481 | Error: ā¤Unexpected "x"ā¤expecting term postfix, operator, ":" or "," | ||
| eviltwin_b | no, that was me thinking wrong | ||
| root4o | ?eval $x = \(1,2,3); say |$x | 05:28 | |
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| root4o | ?eval $x = \(1,2,3); say *|$x | ||
| evalbot_r14481 | OUTPUT[any(VNum Infinity,VV (VPure (CaptSub {c_feeds = [MkFeed {f_positionals = [VPure (IFinite 1),VPure (IFinite 2),VPure (IFinite 3)], f_nameds = fromList []}]})))ā¤] Bool::True | ||
| root4o | ?eval $x = \(1,2,3); say $x.yaml | 05:30 | |
| evalbot_r14481 | OUTPUT[--- !hs/IScalar ā¤- !hs/VV ā¤ - !hs/Pugs.Val.Val ~ā¤ā¤] Bool::True | ||
| eviltwin_b | ?eval my $x = \(1,2,3); say |$x | ||
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| eviltwin_b | ?eval my $x = \(1,2,3); say |($x) | 05:31 | |
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| eviltwin_b | ?eval my $x = \(1,2,3); say(|($x)) | ||
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| root4o | ?eval $x = \([1,2,3]); say |$x | ||
| eviltwin_b | hrm, that looks not good | ||
| evalbot_r14481 | pugs: don't know how to toVV': VRef <Array:0xb7b805f0> | ||
| root4o | what is 'toVV' mean ? | 05:32 | |
| eviltwin_b | dunno but it sounds internal | ||
| root4o | ?eval $x = \(a => 1); say |$x | ||
| eviltwin_b is still working his way throuhg the Synopses | |||
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| eviltwin_b | see, I thought unary | "unwrapped" the cpature but that looks like not | 05:33 | |
| root4o | ?eval $x = \(a => 1); sub x($a) { $a }; say x($x) | ||
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| root4o | ?eval $x = \(a => 1); sub x($a) { $a }; say x(|$x) | 05:34 | |
| evalbot_r14481 | OUTPUT[<capt>ā¤] Bool::True | ||
| root4o | ?eval $x = \(a => 1); sub x($a) { $a }; say x(*|$x) | ||
| (22:35:03) evalbot_r14481: OUTPUT[<capt>ā¤] Bool::True | |||
| evalbot_r14481 | OUTPUT[any(VNum Infinity,VV (VPure (CaptSub {c_feeds = [MkFeed {f_positionals = [], f_nameds = fromList [("a",[VPure (IFinite 1)])]}]})))ā¤] Bool::True | ||
| eviltwin_b | I don't think we need the "say" there | 05:36 | |
|
05:37
shobadobs joined
|
|||
| root4o | ha | 05:37 | |
| aha | |||
| eviltwin_b | ?eval my $x = \(1,2,3); |$x | 05:41 | |
| evalbot_r14481 | \CaptSub {c_feeds = [MkFeed {f_positionals = [VPure (IFinite 1),VPure (IFinite 2),VPure (IFinite 3)], f_nameds = fromList []}]} | ||
| eviltwin_b | ?eval my $x = \(1,2,3); |(|$x) | ||
| evalbot_r14481 | \CaptSub {c_feeds = [MkFeed {f_positionals = [VPure (IFinite 1),VPure (IFinite 2),VPure (IFinite 3)], f_nameds = fromList []}]} | ||
| root4o | ?eval $x =\{a => 5}; |$x | 05:44 | |
| evalbot_r14481 | {("a" => 5),} | ||
| root4o | ?eval $x =\(1,2); |{$x} | 05:45 | |
| evalbot_r14481 | {Syn "block" {Var "$x"}} | ||
| root4o | ?eval $x =\(1,2); {|$x} | ||
| evalbot_r14481 | \CaptSub {c_feeds = [MkFeed {f_positionals = [VPure (IFinite 1),VPure (IFinite 2)], f_nameds = fromList []}]} | ||
| root4o | ?eval $x =\{(1,2)}; |$x | 05:46 | |
| evalbot_r14481 | \{Syn "block" {Syn "," {1;ā¤ 2}}} | ||
|
06:11
root4o left
06:28
marmic joined
06:39
seano joined
06:46
DHGE joined
06:53
nothingmuch joined
07:54
lumi joined
08:03
AzureStone joined
08:06
foo\ joined
08:31
SubStack joined
08:40
SCalimlim joined
08:52
iblechbot joined
09:00
pjcj joined
09:05
kanru joined
09:10
ditto joined
09:13
DHGE left
09:22
larsen_ joined
09:28
dduncan left
09:29
nothingmuch joined
09:33
qmole joined
09:40
Tzoa joined
09:41
larsen__ joined
09:45
Tzoa left
10:31
yodo joined
10:34
frankg joined
10:39
stennie joined
10:56
lumi joined,
marmic joined,
andres joined,
Steve_p joined,
f0rth joined,
TimToady joined,
ingy joined,
fridim joined,
ajs joined,
awwaiid joined,
holoway joined,
chip joined,
sri_ joined,
dvorak joined,
orafu joined,
revdiablo joined,
perlbot joined,
rafl_ joined,
hcchien joined,
cj joined,
kolibrie joined,
jferrero joined,
frankg joined,
qmole joined,
nothingmuch joined,
kanru joined,
iblechbot joined,
SCalimlim joined,
SubStack joined,
foo\ joined,
shobadobs joined,
MacVince joined,
weinig|sleep joined,
buubot joined,
Odin- joined,
bennymack joined,
lollan joined,
jabbot joined,
araujo joined,
cnrocks joined,
wilx joined,
plural joined,
bcorn joined,
audreyt joined,
Juerd joined,
webmind joined,
PerlJam joined,
leo joined,
pmichaud joined,
wolverian joined,
yuval joined,
statico2 joined,
mdiep_ joined,
spoop joined,
Narcisse joined,
stevan joined,
meppl joined,
takesako joined,
takanori_ joined,
autark joined,
buu joined,
nine joined,
jdv79 joined,
rafl joined,
Ikarus joined,
svnbot6 joined,
nipotaway joined,
gnuvince joined,
jojo joined,
spinclad joined,
Shabble joined,
jiing joined,
masak joined,
osfameron joined,
beppu joined,
cognominal joined,
Yappo joined,
eviltwin_b joined,
baest joined,
ajs_home joined,
kcwu joined,
norageek joined,
kalven joined,
woremacx joined,
mako132 joined,
kane-xs_ joined,
ruz joined,
xinming joined
11:01
weinig|sleep is now known as weinig
11:03
buetow joined
11:23
chris2 joined
11:33
AzureStone joined
11:38
Psyche^ joined,
pjcj joined
11:50
Psyche^ is now known as Patterner,
elmex joined
12:13
ditto joined
12:37
andres_ joined
|
|||
| svnbot6 | r14482 | kudra++ | Summaries for this week | 12:40 | |
| meppl | "require" doesnt work inside a class? | 13:48 | |
| audreyt | nopaste? or commit a test? | 13:49 | |
|
13:56
stevan_ joined
|
|||
| pasteling | "meppl" at 89.54.90.94 pasted "require does not work here?" (10 lines, 114B) at sial.org/pbot/20597 | 14:10 | |
| meppl | k | ||
| audreyt, so, in this paste require does not work | |||
|
14:10
weinig is now known as weinig|bbl
|
|||
| meppl | the error-message is: | 14:11 | |
| *** No compatible subroutine found: "&Fest::Foo" | |||
| at ./Fest.pm line 7, column 9-33 | |||
| it does not work, when i use "require" in a class, which is in a extra file | |||
| xinming | BTW, may I ask what does "nopaste" mean please? Every time I saw nopaste, I know I would paste somewhere in trained reflex... :-P | ||
| meppl | "require" works in a "class{}" | 14:12 | |
| xinming, paste code into here ;) sial.org/pbot/paste | 14:13 | ||
| audreyt | meppl: well try ::Fest::Foo | ||
| xinming: nopaste is the older paste service nopaste.snit.ch/ | |||
| meppl | xinming, "nopaste" ---> not paste it direct in the channel | ||
| audreyt | meppl: problem is that require does not yet introduce a class literal | ||
| not exactly sure if it should | |||
| maybe yes | |||
| meppl | hm | ||
| no, that doesnt work, too | 14:14 | ||
| whatever | |||
| i can also use "use" | 14:15 | ||
| momentaneous | |||
| ;) | |||
| thx | |||
| audreyt | ::Fest::Foo.new ? | ||
| that doesn't work? | |||
| meppl | oh | ||
| audreyt, that works | |||
| yes | |||
| thank you very much audrey | 14:16 | ||
| audreyt | ok. the problem is that | ||
| require Foo::Bar | |||
| can't tell if F::B is a class (in which case Foo::Bar should be registered as literal) | |||
| or if it's a package (in which case it should not. Or should it?) | |||
| so in writing ::Fest::Foo you delay the resolution to runtime | |||
| meppl | ah, okay | 14:17 | |
| audreyt | principle of least surprise tells us that maybe registering a literal regardless of class/moduleness is sane | 14:18 | |
| audreyt ponders about it some more | 14:19 | ||
| meppl | what about that? require "/home/non/Sense.pm"; | 14:36 | |
| require $foo; | 14:37 | ||
| my $bar = ::$foo.new(); | |||
| does not work ;) | |||
| if $foo were edited to home::non::Sense | 14:38 | ||
| audreyt | yup | ||
| ::($foo) will work | 14:39 | ||
| meppl | thx | 14:40 | |
| nothingmuch | audreyt: got a few spare minutes for a brain dump? | 14:49 | |
| wrt MO::Emit | |||
| eviltwin_b | dumb question about capture objects? | 14:52 | |
| audreyt | eviltwin_b: there's only dumb answers... | 14:53 | |
| eviltwin_b | ?eval my $x = \(1,2,3); sub foo (Int $a, Int $b, Int $c) {say "Hi!"}; foo(|$x) | ||
| audreyt | nothingmuch: wait me for ~20min? | ||
| evalbot_r14482 | Error: No compatible subroutine found: "&foo" | ||
| audreyt | need to finish breakfast | ||
| eviltwin_b | shouldn't that be interpolating? | ||
| audreyt | nothingmuch: of course you can braindump ahead | ||
| eviltwin_b | a couple of us were tossing that around last night (~8 hours ago) | ||
| nothingmuch | audreyt: i'll wait | 14:54 | |
| thanks =) | |||
| }:) | |||
| eviltwin_b | 9 hours ago :) | ||
| audreyt | eviltwin_b: weird... seems interpolation is like completely broken :) | ||
| I'll check. a sec... | 14:55 | ||
| eviltwin_b | we were wondering if it was implemented yet. further exploration made it look like the Capture object was passed through unchanged | ||
| audreyt | that is the case though it shouldn't | ||
| nothingmuch | audreyt: i'll hop to the post office | ||
| eviltwin_b | (i.e.. declare with a single untyped argument, that argument is rvealed to be a Capture) | ||
| nothingmuch | ~30 mins | ||
| sounds good? | |||
| audreyt | compare $x = [1,2,3] | ||
| nothingmuch: sure | |||
| which does interpolate correctly | |||
|
15:28
vel joined
15:33
[particle] joined
|
|||
| audreyt | eviltwin_b: there you go | 15:46 | |
| thanks for prodding | |||
| svnbot6 | r14483 | audreyt++ | * Interpolation of Capture objects via prefix:<|> implemented, | ||
| r14483 | audreyt++ | suggested by eviltwin_b++: | |||
| r14483 | audreyt++ | sub f ($x, $y) { ... } | |||
| r14483 | audreyt++ | my $capt = \(1,2); | |||
| r14483 | audreyt++ | f(|$capt); # now works | |||
| r14483 | audreyt++ | Previously only Array, Hash, Scalar and Pair were intepolated. | |||
| r14482 | kudra++ | Summaries for this week | |||
| audreyt | nothingmuch: hmm it's ben nearly an hour... | 15:47 | |
| nothingmuch: I had to nap a bit. please braindump ahead when you're back but obra will abduct me to some coffee tasting soon | 15:48 | ||
| gaal | audreyt: in a few weeks i'll prolly have some more time - and can start moosing over more stuff to newland in the parser | ||
| audreyt | gaal: good, mostly it's just the Sub signature producer | ||
| once it's there we can implement sub Main | |||
| which is a nice standalone task | |||
| gaal | well, and then everything else! | ||
| audreyt really looks forward to sub Main | 15:49 | ||
| gaal | eg oldcode, pads(!), lexical pragmas | ||
| audreyt | I mean sub MAIN. | ||
| nodnod | |||
|
15:49
Khisanth joined
|
|||
| nothingmuch | audreyt: back | 15:49 | |
| audreyt | oy, there goes my napping | ||
| gaal | I'd like to be / under the sea / in an octopuses garden / in sub MAIN | 15:50 | |
| nothingmuch | heh =) | ||
| gaal | er, wrong song | ||
| nothingmuch | you can nap if you prefer | ||
| audreyt | I'll nap _and_ IRC at the same time | ||
| so dump ahead | |||
| nothingmuch | heh | ||
| it's supposed to be bidi ;-) | |||
| but okay | 15:51 | ||
| i'm trying to figure out MO::Emit | |||
| been doing so for the past 2-3 days | |||
| i am trying to design it so that it's portable | |||
| to some extent | |||
| the mandatory incest we require: | |||
| the emitter has an undefined output, but a well defined input (responder interfaces) | 15:52 | ||
| sometimes the output can be responder interface, if it's a dynamic runtime | |||
| (or partly) | |||
| audreyt | (all runtimes are dynamic) | ||
| nothingmuch | e.g. in the case of the p5 emitter the output is packages | ||
| all our runtimes are dynamic | |||
| audreyt | correct. | ||
| nothingmuch | however, MO, by design, can also do static systems | ||
| anymoose | 15:53 | ||
| audreyt | yes, but a compelling use case does not yet exist to generalize that so bduf yagni et | ||
| cetera | |||
| nothingmuch | yes, i'm ignoring the cases i don't care about in terms of implementation | ||
| i'm just being accurate in my descriptions ;-) | |||
| however | 15:54 | ||
| for my purposes the p5 emitter is not dynamic per se | |||
| in that the target runtime doesn't twiddle MO objects all over the place | |||
| so dynamic -> foreign dynamic is the same as dynamic -> static as far as the emitter is concerened | |||
| (set aside the fact that the second is a lot more work ;-) | |||
| i need two runtimes for p5: | 15:55 | ||
| one is the interpreted one, in which there are actually calls to $ri->dispatch( $invocation, $responder ); | |||
| and the "compiled" one, in which $ri's are flattenned into packages | |||
| and $invocation must be ::Method | |||
| and $responder must be invocant | |||
| the flattenning into packages is incest #1 that the emitter must do | |||
| audreyt | define incest | 15:56 | |
| nothingmuch | (e.g. take apart a private interface, layer it over, and codegen methods that are actually dynamic with if ( caller == ... ) { $body_one->() }) | ||
| incest == related objects that know a lot about eachother | |||
|
15:56
stevan_ joined
|
|||
| audreyt | oh. tight coupling | 15:56 | |
| ok | |||
| nothingmuch | that is, you can't just toss any old object into the emitter | ||
| the polymorphism only takes you so far | |||
| tight coupling sounds like a good thing sometimes ;-) | 15:57 | ||
| anymoose | |||
| the second type of incest we need to take care of is method body <-> emitter | |||
| the problem is that to write generic accessors | |||
| that don't care what the emitter's output is | 15:58 | ||
| audreyt | ...is impossible | ||
| nothingmuch | i need some sort of abstracting out of calling conventions, etc | ||
| well, it si possible, for very simple accessors | |||
| audreyt | you need another layer | ||
| nothingmuch | what I need is macros | ||
| audreyt | yes. | ||
| nothingmuch | e.g. method_call( $inv_ast, $method ) | ||
| but perl 5 doesn't have these | |||
| audreyt | write one for it then | 15:59 | |
| nothingmuch | i can kludge them as simply runtime calls | ||
| (fugly..., sub method_call { ... }) | |||
| or i can codegen some perl using Code::Perl (not very transformable) | |||
| or I can finish Perl::Generate (too long) | |||
| audreyt | what transormatino you going to do? | ||
| I think Code::Perl makes a lot of sense | 16:00 | ||
| nothingmuch | e.g. the method call macro | ||
| it does | |||
| audreyt | as a proof of concept | ||
| nothingmuch | but it's holder objects aren't powerful enough | ||
| it's nice as an emitter AST | |||
| but not as a highlevel AST | |||
| audreyt | my pt is that HL AST is not needed for the specific case of ->pkg p5 emit | ||
| nothingmuch | indeed | 16:01 | |
| but Code::Perl is not enough for fake method call vs. real method call going into the same AST node | |||
| which means i have to rewrite subtrees | |||
| which means pain | |||
| also, it doesn't have a walking API | |||
| audreyt | it's localized limited pain though | ||
| nothingmuch | my plan B is to go 100% macro | ||
| audreyt | walking api is trivial | ||
| just reify the calls | |||
| wrap them with all same-named functions | 16:02 | ||
| nothingmuch | since i'm only emitting very specific code, i write macros for all these "things" (a method call, a return, etc)a | ||
| nd the emitter has a class for macros | |||
| which does the relevant Code::Perl calls | |||
| audreyt | that reifies them into data structure that represents suspented call | ||
| presto, walkability | |||
| yes, it's similar to the macros you just described | |||
| nothingmuch | yep | ||
| any other ideas? | |||
| err,nevermoose, it won't get any better than this | 16:03 | ||
| okay, the second thing is how should the input params to the emitter be designed? | 16:04 | ||
| i'm thinking there should be low level emitter that deals with responder interfaces | |||
| and an "assistant" that futzes the meta classes | |||
| audreyt | define futz | 16:05 | |
| nothingmuch | well, the problem is that responder interfaces <-> perl 5 packages are not a 1:1 fit | ||
| to support SUPER:: etc natively we decompile the responder interfaces into actual classes by first emitting the superclasses | |||
| if the MRO and perl's MRO disagree | 16:06 | ||
| we kludge by masking the superclass | |||
| hence the unit being emitted is a collection of classes, typically | |||
| but that is problematic, because classes are not the one true way of creating responder interfaces | 16:07 | ||
| audreyt | why is that prblmtc? | 16:09 | |
| you only mro-kldg on cls clstr | |||
| nothingmuch | that's for the " | ||
| compiling" perl 5 emitter | |||
| the non compiling emitter will simply fmap the input objects | |||
| audreyt | yeah so say you have a nonclass rspd if | ||
| nothingmuch | recursively replacing method defs with code refs | ||
| audreyt | yes yes | ||
| a non class rspdif will not participate in c3 | 16:10 | ||
| it will handle mro some other way | |||
| afaics | |||
| nothingmuch | yes | ||
| audreyt | so why is that problmtc? | ||
| nothingmuch | mm... i suppose the input to the emitter is just responder interfaces... =P | ||
| audreyt | I thougt you said that yourself. | 16:11 | |
| 17:52 < nothingmuch> the emitter has an undefined output, but a well defined input (responder interfaces) | |||
| nothingmuch | i guess i'm searching for something a bit more like a compilable unit | ||
| heh | |||
| oops =) | |||
| well, that was a slip then | |||
| audreyt | I think it's orthogonal | ||
| compunit means your analysis has a bound | |||
| but for MO whenever you hit "emit" that's a compunit | 16:12 | ||
| by definition | |||
| so don't worry about that | |||
| nothingmuch | hmm | ||
| fairynuff | |||
| audreyt | sepcomp will just trigger MO once for each compu | ||
| (sorry for abbrs I'm napping so suspending laptop with my palms ) | |||
| nothingmuch | ... and it'll be the emitter's problem i guess, to also glean data from the already compiled units | ||
| audreyt | or it calls into a nonlocal analyzer | 16:13 | |
| nothingmuch | okay, i'll let you nap for real and stop talking nonesense ;-) | ||
| audreyt | without storing state itself | ||
| nothingmuch | polymorphic ;-) | ||
| audreyt | what, I talk nonsesne? | ||
| :) | |||
| nothingmuch | i'll let you $unit, and stop talking nonesense | ||
| not i'll let you $unit | |||
| where 4unit == ( $subunit, and stop talking nonesnese ) | |||
| ;-) | |||
| if that was the case, i'd i wouldn't use "and stop", because that's bad english | 16:14 | ||
| audreyt | well it makes a lot of non nonsense | ||
| nothingmuch is having coders block | |||
| audreyt | which represents a thunk of sense | ||
| nothingmuch | hence undecisiveness | ||
| audreyt | (cf. Haskell is Not Not ML) | ||
|
16:15
stef_ joined
|
|||
| audreyt | oh. then write some more component tests | 16:15 | |
| I hear that's good for you | |||
| nothingmuch | i've been doing Crypt::Util when my brain would freeze | ||
| audreyt | gravedigging module? | ||
| nothingmuch | =) | ||
|
16:15
stef_ joined
|
|||
| audreyt | "This module provides an easy, intuitive and forgiving API for wielding crypto-fu." | 16:16 | |
| nothingmuch | use Shovel; open my $fh, $crypt; while( my $goodies = <$fh> ) { } =) | ||
| audreyt | :D | ||
| cool, that part of PlRPC I liked | |||
| so abstracting it out into a ::Util makes sense | |||
| nothingmuch | oi! i didn't know | ||
| i'll go through PlRPC and extract goodies | 16:17 | ||
| eviltwin_b | "suggested"? | ||
| S06 | |||
| nothingmuch | huh? | ||
| audreyt | Obviously you still haven't quite memorized all the synopses (of all CPAN modules)... :-) | ||
| nothingmuch | that was perl 5 | ||
| audreyt | eviltwin_b: "suggested to implement" | ||
| nothingmuch | wait wait wait | ||
| oh, i thought that was you | |||
| eviltwin_b | ah, was not clear that it wasn't implemented yet | ||
| nothingmuch | my nickcoloring script died, sorry =( | ||
| audreyt | eviltwin_b: I forgot to implement it during the releng frenzy is all | 16:18 | |
| eviltwin_b | ah | ||
| audreyt | because it's not tested! | 16:19 | |
| (or was it) | |||
|
16:19
andres__ joined
|
|||
| audreyt | eviltwin_b: can you add the relevant tests to t/data_types/capture.t ? | 16:20 | |
| in additon to &foo.call($c) it should also test foo(|$c) | |||
| eviltwin_b | not at the moment | ||
| audreyt | someone else then? :) | ||
| I'll nap for real... | |||
|
16:21
ferreira joined
|
|||
| eviltwin_b | I have a nonfunctional setup at the moment, found out last night my dports is thoroughly horked so I'm doing a full reinstall | 16:21 | |
| hence most stuff (including svn and dependencies) is currently nonfnctioning | |||
| audreyt | eviltwin_b: mail juerd at convolution.nl with your real name and preferred login name | ||
| to get a feather.perl6.nl / perlcabal.org account | 16:22 | ||
| which is handy for things like webirc and pugshacking | |||
| afk... & | 16:23 | ||
|
16:24
Limbic_Region joined
16:26
lisppaste3 joined
16:40
DHGE joined
16:47
prism joined,
TimToady joined
16:54
weinig|bbl is now known as weinig
16:56
bernhard joined,
justatheory joined
|
|||
| nothingmuch | @seen cmarcelo | 17:07 | |
| lambdabot | I saw cmarcelo leaving #perl6 3d 12h 20m 38s ago, and . | ||
|
17:08
thepler joined,
chris2 joined
17:18
penk joined
17:20
offby1 joined
17:21
aufrank joined
|
|||
| offby1 | here's a dumb question: what's the perl6 equivalent of the Camel Book? | 17:21 | |
| I.e., the one place I can go to find out how to actually write perl6 code? | |||
| nothingmuch | the synopses are the closest thing we have | ||
| offby1 | *sigh* | ||
| well, maybe that's not as bad as I'd thought. | |||
| nothingmuch | there's also a plethora of tutorials | 17:22 | |
| offby1 | they are kept up to date with pugs, yes? | ||
| nothingmuch | it's more the other way around | ||
| offby1 | see, I don't want a plethora, because then I have to decide which to pick. | ||
| nothingmuch | pugs tries to keep up to date with them | ||
| offby1 | I want just one canonical authoritative site. | ||
| nothingmuch | right, that's why i assumed you'd prefer the synopses =) | ||
| offby1 | OK, I'll see if my pugs checkout includes them | ||
| nothingmuch | it doesn't | 17:23 | |
| dev.perl.org/perl6/doc/synopsis.html | |||
| lambdabot | Title: Synopses - perl6 | ||
| offby1 | oh :-| | ||
| thanks | |||
| nothingmuch | also in svn.perl.org | ||
| and Perl6::Bible | |||
| offby1 | to be honest, I can never remember if it's the Synopses, Apolocyptamundo, or the Ex-Weegeeans that I want to consult. | 17:24 | |
| aufrank | I thought pugs/docs/Perl6/Spec/ was equivalent to the syns these days | ||
| nothingmuch | oh | ||
| offby1 isn't the only one confused, apparently | |||
| nothingmuch didn't know | |||
| perhaps /me is out of date | |||
| spec.pugscode.org is a good way to see how pugs <-> syns interact | |||
| that said, synopsis describe and specify perl 6 | 17:25 | ||
| apocalypses are earlier (so now out of date) | |||
| and explain more of the why behind some things, or what some things should be | |||
| and plural(exegesis) were supposed to help explain apocalypses using a practical code sample | |||
| aufrank | exegeses | 17:26 | |
| nothingmuch | ta =) | ||
|
17:27
nelhage joined
|
|||
| offby1 | thanks | 17:28 | |
| Not much; you? :-) | |||
| nothingmuch | ta =) | 17:29 | |
| bah | |||
| this is not a terminal | |||
| offby1 | Ici n'est pas un pipe. | ||
| Are the synopses kept under revision control somewhere? I prefer reading things locall via my editor, than on the web ... | 17:30 | ||
| *whine whine* | |||
| nothingmuch | svn.perl.org/perl6/doc | 17:32 | |
| lambdabot | Title: Revision 13276: /doc | ||
| offby1 | thanks!! | 17:38 | |
| offby1 smacks forehead | 17:39 | ||
| I already had a checkout of that. Yeesh. | |||
| nothingmuch | heh | ||
| offby1 | I bet I asked for the URL here back then, too :-| | ||
| I need to be _forced_ to use Perl6 for a few months, rather than dabbling in it. That way I'll remember this stuff. | |||
|
17:40
bonsai joined
|
|||
| offby1 happily reads S02 | 17:43 | ||
|
17:49
mako132_ joined
17:58
Aankhen`` joined,
TimToady joined
17:59
zdeqb joined
18:01
zdeqb left
|
|||
| offby1 | OK, my first real question. I want a hash whose values are anonymous list references. In Perl5, I'd add a new element to one of those lists by saying push ``@{%dict_hash{$key}}, $value'', but with pugs, I get an error: ``Unexpected "{%" expecting "@" or "::"''. What's the syntax that I'm looking for? | 18:20 | |
| actually I guess in Perl5 I'd really say ``push @{$dict_hash{$key}}, $value'' | |||
| but you get the idea | |||
| wolverian | you don't need to deref explicitly in perl6. | 18:21 | |
| offby1 | go on. | 18:22 | |
| wolverian | %hash{$key}.push($value); | ||
| offby1 | oho! I'll try that. | ||
| hmm, now I must show you what I'm _actually_ doing, rather than the simplified version :-) | 18:23 | ||
| %dict_hash{&bag($chopped)}.push ($chopped); | |||
| that yields | |||
| Unexpected "($" expecting comment, operator, statement modifier, ";" or "}" | |||
| so -- I want to call the "bag" function on $chopped, and use that as my hash key ... | 18:24 | ||
| wolverian | you can't use whitespace so freely. | ||
| offby1 | ah | ||
| wolverian | otherwise it looks fine, except the & is not required. | ||
| offby1 | superstition :-) | ||
| wolverian | it shouldn't be used in perl5 either. | ||
| (except when you need it. which is probably never. :) | 18:25 | ||
| offby1 | fabulous! | ||
| thanks, mr wolverian ! | |||
| wolverian | you're welcome :) | 18:26 | |
|
18:29
cjeris joined
|
|||
| offby1 | OK, next: how do I open a file, and tell Perl (or Pugs) "This file is encoded in Latin-1, so please don't whine about "decodeUTF8: bad data '\243'" ?? | 18:30 | |
| I bet it's just an extra arg to "open" | |||
| alas, I can't remember where "open" is documented | |||
| eviltwin_b | S16? | 18:32 | |
| offby1 | lemme look | ||
| dev.perl.org/perl6/doc/synopsis.html doesn't mention an s16 | 18:33 | ||
| lambdabot | Title: Synopses - perl6 | ||
| eviltwin_b | hm, it's a draft | ||
| offby1 | nor do I find one in /usr/local/src/langs/perl6-doc/ | ||
| eviltwin_b | I'm at perlcabal.org/syn/? (the ? is part of the url) | ||
| lambdabot | Title: Official Perl 6 Documentation | ||
| eviltwin_b | redirected from spec.pugscode.org | 18:34 | |
| offby1 | yikes, there really _is_ no one place to find the perl6 equivalent of the Camel Book ... :-( | ||
|
18:34
Limbic_Region joined
|
|||
| eviltwin_b | I think perl6 is still a bit too young for that | 18:35 | |
| offby1 | guess so | ||
| eviltwin_b | not to mention pugs vs. parrot vs. other implementations | ||
| offby1 | well, I'd assume all those implementations would hew to the same spec. | ||
| s/assume/hope/ | |||
| eviltwin_b | heh | ||
| perl evolves by mutation, as always, the spec just documents the successful mutations :) | 18:36 | ||
| offby1 | hmm | ||
| wolverian | the idea is, actually, that the implementation adhere to the spec. | ||
| offby1 | well, S16 didn't answer the question anyway. | ||
| wolverian | in any case, I don't think pugs has IO layers yet. | 18:37 | |
| offby1 | Any idea how to tell perl that my file is Latin-1 ? | ||
| oh | |||
| wolverian | you could always look at the source :) | ||
| ingy | seen TimToady | 18:39 | |
| jabbot | ingy: TimToady was seen 1 days 13 hours 7 minutes 49 seconds ago | ||
|
18:49
aufrank joined
18:52
glasser joined
|
|||
| offby1 | so ... Ints and Strings are different types, in that ``3.perl'' and ``"3".perl'' yield different strings ... but it appears that if I use an Int as a hash key, it gets coerced to a string. Am I understanding correctly? | 18:57 | |
| eviltwin_b | as I understand it, that's the default. you can declare hash using general objects (Any) or other types as keys | ||
| offby1 | aha | 18:58 | |
| that coercion indeed mimics perl5's behavior. | |||
| (except Perl6 is better because I can use _big_ ints as keys. yay) | |||
| hm. How _would_ I write that declaration? ``my Int %h'' doesn't sound right -- I'd expect that to declare that the _values_ are Ints, not the keys. | 19:00 | ||
| wolverian | that's defined in the data structures synopsis | 19:01 | |
| offby1 | aha, I'm looking in the wrong place, then. | ||
| wolverian | perlcabal.org/syn/S09.html#Hashes | 19:02 | |
| lambdabot | Title: S09 | ||
| offby1 | hmm, isn't doing what I'd expect: | 19:03 | |
| my Any %h{Int} | |||
| %h{3} = "Three" | |||
| %h{"4"} = "Four" | |||
| %h | 19:04 | ||
| (("3", "Three"), ("4", "Four")) | |||
| wolverian | pugs doesn't necessarily implement that yet. | ||
| offby1 | ah | ||
| but it looks like I'm doing the right thing then? If so I'll just write it that way and ignore that pugs doesn't yet do it. | 19:05 | ||
| wolverian | yup | 19:06 | |
| offby1 | thanks. | 19:07 | |
| nelhage | "You should write a test" :) | 19:22 | |
| wolverian | that, too. :) | ||
| nelhage | OK. I've built and installed pugs from svn; How do I run stuff in the examples/ dir? neither perl examples/foo.pl nor pugs examples/foo.pl is working | 19:23 | |
| offby1 | I dunno, but I'll be curious to hear the answer too. | 19:24 | |
| in the meantime, my next q: | |||
| my $cache_file_name = "dict.cache" | |||
| my $cache = open ($cache_file_name, :w) | |||
| --- That yields | |||
| *** dict.cache w1: openFile: does not exist (No such file or directory) | |||
| that's puzzling, since I think I'm telling it to create a new file | 19:25 | ||
| pugs/t/builtins/io/open.t seems to imply I'm doing it right... | |||
| wolverian | you don't want the whitespace there, anyway. | 19:26 | |
| offby1 | *smack* | ||
| gotta figure out how whitespace works | |||
| eviltwin_b | GNU style conventions are not valid Pugs (and suck anyway IMO :) | 19:27 | |
| offby1 | just habit | ||
| wolverian | perlcabal.org/syn/S02.html#Whitespa...d_Comments | ||
| lambdabot | Title: S02, tinyurl.com/y89dt4 | ||
| offby1 | yeah, I read it, but it hasn't sunk in yet | ||
| wolverian | the relevant bit here is that whitespace before a postfix operator is invalid (unless escaped) | 19:28 | |
| offby1 stares blankly | |||
| eviltwin_b | look for "long dot" | ||
| wolverian | yeah. | 19:29 | |
| nelhage looks for the postfix operator in offby1's code fragment, but fails to find it. | |||
| eviltwin_b | <( )> | ||
| wolverian | nelhage, () | ||
| eviltwin_b | (that is, the parens are a postfix_circumfix operator | ||
| nelhage | hmm, ok. I guess that makes sense | 19:30 | |
| eviltwin_b | a postfix operator which takes an argument in circumfix | ||
| nelhage | (*very* new to p6 here) | ||
|
19:30
MacVince joined
|
|||
| eviltwin_b is still quite new to it as well but is working through the Synopses and seems to be retaining *something* at least | 19:30 | ||
| offby1 | I think my rule will be: never use any whitespace anywhere :-| | 19:31 | |
| eviltwin_b | that won't help, there are cases where it's mandatory :> | ||
| offby1 | *smack* | ||
| I think my rule will be: add and delete whitespace at random until it compiles :-| | |||
| eviltwin_b | hey, it could be Python... | ||
| offby1 | anyway. | ||
| so: what's the inverse of "say $foo.perl" -- in other words, what's the perl6 equivalent of "do $filename" ? | 19:32 | ||
| wolverian | use a module :) | 19:33 | |
| offby1 | huh? | 19:34 | |
| I don't understand how that answers my question. | |||
| maybe I wasn't clear. | |||
| wolverian | generally you want to use Foo; instead of do "foo.pl" | 19:35 | |
| so. anyway. good night! | |||
| offby1 | "say $foo.perl" is a delightfully easy way to get an unparsed string that, when re-parsed by perl, creates a value that is just like $foo. So it's a handy way to write perl data to files. Now, if I do this, I am eventually going to want to read that data back, and turn it back into a perl value. How do I do this second step? | ||
| wolverian | eval() | 19:36 | |
| offby1 | eval what? | ||
| the string? | |||
| oh, of course. | |||
| offby1 tries | |||
| eviltwin_b | you want that anyway, really, IMO --- forces you to at least think about validation etc. | 19:37 | |
| offby1 | wolverian: seems to work; thanks. | ||
| eviltwin_b: say again? | |||
| eviltwin_b | if you just slurp data in with "do" (even in perl5) you're being lazy and potentially very insecure | 19:38 | |
| offby1 | because the "data" might contain Evil Code. Sure. | ||
| eviltwin_b | making it harder to do that is therefore a Good Thing IMO | ||
| offby1 | uh | ||
| not if I already know beforehand that the stuff is safe, it isn't. | |||
| anyway, calling "eval" doesn't seem noticeably safer. | 19:39 | ||
| eviltwin_b is a professional paranoid and thinks that "already know beforehand" is a road paved with good intentions | |||
| offby1 | good point, I'll grant you. | 19:40 | |
| eviltwin_b | it's not, but by breaking it up it should aty least suggest putting checks between the read and the eval | ||
| offby1 | aaahhh | ||
| gotcha | |||
| although ... | |||
| ugh. | |||
| any such checks would require me to do my _own_ parsing, wouldn't they? | |||
| surely there's a safe, standard way to serialize and unserialize data? YAML, perhaps? | 19:41 | ||
| eviltwin_b | I think there is (or is intended to be) but haven't seen it yet | ||
| oh, if you're intent on being lazy I get the impression open($file).slurp.eval is supposed to work | 19:42 | ||
| offby1 | sounds good. | ||
| works indeed. | 19:43 | ||
| eviltwin_b | as to parsing, that's not too horrible an idea --- go take a look at grammrs | 19:45 | |
| er. Grammars | |||
| eviltwin_b is impressed --- has been noodling that concept for a while | 19:46 | ||
| offby1 | I refuse to do that sort of parsing. It's unthinkable that perl6 won't eventually provide a safe way to do this. | 19:47 | |
| eviltwin_b | worse come to worst you could fall back to perl5 YAML :> | 19:48 | |
| oh look, S29 mentions an evalfile(). (now if there were an indication of any of these things are actually implemented...) | 19:50 | ||
| offby1 | I notice that there's already a .yaml method for generating the output, so I wouldn't be surprised if there were a corresponding method for reading it back in | 19:51 | |
| ooh, look how a file handle stringifies: | 19:58 | ||
| <VHandle ({handle: dict.cache}> | |||
| that's odd | |||
| unbalanced left paren. | |||
| I know that isn't meant to be parsed, but it still looks funny ... | |||
| SamB | indeed it does... | 20:03 | |
| I bet it was a mistake | |||
|
20:05
MacVince joined
|
|||
| gaal | offby1: also check out .yaml | 20:17 | |
|
20:17
ferreira joined
|
|||
| gaal | ?eval { moose => 'elk', 42 => "lala" }.yaml | 20:18 | |
| evalbot_r14483 | "--- !hs/IScalar \n- !hs/VUndef VUndef\n" | ||
| gaal | ugh. | ||
| eviltwin_b | ...or then again,maybe not:> | ||
| gaal | ?eval my %x = ( moose => 'elk', 42 => "lala" ); %x.yaml | ||
| evalbot_r14483 | "--- \n? 42\n: lala\n\n? moose\n: elk\n\n" | ||
|
20:19
trym joined
|
|||
| gaal | the other direction is eval $str :lang<yaml> | 20:19 | |
| offby1 | let's try that. | 20:23 | |
| gaal | ?eval { why => 'did that happen earlier' }.perl | 20:25 | |
| evalbot_r14483 | "\\undef" | ||
| offby1 | cool | ||
| gaal | interesting. | ||
| very not cool :( | |||
| offby1 | I meant "cool" because one can read yaml natively | ||
| gaal | weirdly, it's newVal undef. | ||
| yes. it roundtrips objects too :) | |||
| nelhage | ?eval my $x = {whats => "up"}; $x.perl | ||
| evalbot_r14483 | "\{(\"whats\" => \"up\"),}" | ||
| gaal | nelhage: yes, I suspect it's parsed as a thunk, which may be correct, but then showing it as an undef is incorrect. | 20:26 | |
| ?eval { "this is an anonyblock" }.perl | 20:27 | ||
| evalbot_r14483 | "\\undef" | ||
| gaal | yeah. | ||
| should stringify as CODE0xdeadbeef or something. | 20:28 | ||
| offby1 | indeed | ||
| although a disassembler would be cool too :-) | |||
| gaal | "something" being perhaps 'CODE:{ "this is an anonyblock }' | ||
| yeah :) | |||
| newVal will make this not-so-difficult | 20:29 | ||
| though I need for audreyt to jiggle my memory about how we capture original layout/comments etc. now that newAST doesn't have annotations | 20:30 | ||
| offby1 | *gasp* | 20:32 | |
| there's actually a plan for saving comments and so on in parsed code? | |||
| gaal | of course. how else would we do refactoring IDEs? :) | ||
| offby1 | ga-roovy | ||
| nelhage | gaal: by abusing horrible hacks, of course. | 20:33 | |
| offby1 | sure | ||
| that technique was good enough for my granddaddy | |||
| gaal | nelhage: you mean writing a Z80A backend to pugs, writing the optimizing refactoring engine + AI that infers the comments that went with it on the output, and prettyprinting the result back to Perl 6? | 20:34 | |
| ok, I'd like to see that | |||
| nelhage | Huh. That would be pretty awesome. | 20:35 | |
| I was thinking more along the lines of jesse's fsck.com/~jesse/extract | |||
| (Warning: this code has been known to traumatize small children) | 20:36 | ||
| gaal | that's... intereting | 20:40 | |
|
20:41
jferrero joined
21:02
scsibug joined
21:07
putter joined
|
|||
| putter | openfoundry down? | 21:08 | |
| ah well. bbl & | 21:11 | ||
|
21:19
mauke joined
|
|||
| eviltwin_b | hm, I'm having sporadic network problems in general | 21:31 | |
| nothingmuch | night | 21:34 | |
|
21:43
ferreira left
21:45
shobadobs joined
|
|||
| obra | gaal: it _works_ | 21:52 | |
| it's scary, but works | 21:53 | ||
|
21:56
larsen_ joined
|
|||
| audreyt | openfoundry worksforme | 21:58 | |
| clkao | audreyt: what was the moving to commitbit plan? | 21:59 | |
|
22:03
[particle] joined
22:12
putter joined
|
|||
| putter | re openfoundry, / svn: REPORT request failed on '/pugs/!svn/vcc/default' / svn: REPORT of '/pugs/!svn/vcc/default': Could not read status line: connection was closed by server. (svn.openfoundry.org) | 22:12 | |
| eviltwin_b | 502 Proxy error | 22:13 | |
| cute | |||
|
22:15
weinig is now known as weinig|bbl
|
|||
| Patterner | for interesting values of cute | 22:20 | |
| offby1 | I got that too, and hour or two ago | ||
| s/and/an/ | |||
| audreyt | just for commits? | ||
| pulling seems to be fine | 22:21 | ||
| working now to get us a new svnhost. | |||
| Patterner | "svn up" fails | ||
| eviltwin_b | s/interesting/ironic/ :) | 22:22 | |
| offby1 | works OK nwo. | 22:23 | |
| now. | |||
| audreyt | read worksforme and write is borked. | 22:24 | |
| hopefully will get a new .perl.org host for pugs soon | |||
| (I don't have openfoundry shell anymore so can't fix it at the moment :/) | |||
| offby1 idly wonders if Google could be talked into hosting it | |||
| audreyt | there's already a google pugs host. | 22:25 | |
| offby1 | there you go, then! problem solved :-) | ||
| audreyt | they lifted quota limit for us | ||
| well, problem is all committers need to get gmail accounts... | |||
| and @perl.org with commitbit support doesn't have that restriction | |||
| but yes, googlecode will be our rw fallback | 22:26 | ||
| offby1 | audreyt: I must say, your work on pugs leaves me in awe. It's the only free software project that I am genuinely excited about, _and_ believe has a prayer of actually being completed. | ||
| audreyt | but @perl.org should be main | ||
| Patterner | I think i have 100 invites to gmail left :) | ||
| audreyt | Patterner: yeah but the good nicknames have run out | ||
| clkao | audreyt: luckily there's openid support planned | ||
| audreyt | e.g. I had to be audreyt.org instead of audreyt | ||
| and I'd like to maintain some sort of continuity in nicknames | |||
| offby1 nods | |||
| offby1 hates it when he finds he's not the first "offby1" in some namespace :-| | 22:27 | ||
| Patterner | you don't like [email@hidden.address] | ||
| eviltwin_b | suid audrey? :) | ||
| clkao | 22:28 | ||
| SamB has "naesten", should have grabbed SamB if it was still there though... | 22:32 | ||
| offby1 bellows belligerently: there's only room enough in this community for _one_ "offby1"! | |||
| audreyt | offby1: re genuine excitement, I'm very very happy to hear that :) | 22:34 | |
| offby1 | I imagine I'm one of many, but | 22:35 | |
| audreyt | re completed, sure it will :) | ||
| but we first need to fix the most nonfun part, which is a nonfunctional svn master... working on it now | |||
| (fortunately clkao has a completely nointerrupted-by-skip svk mirror of the repo) | |||
| clkao: can you do the svndumpfilter into huge file now? | |||
| clkao | sure | 22:36 | |
| offby1 | audreyt: if I donate to the Perl Foundation, does that support y'all? | ||
| clkao | i think buying her shiny cloth | 22:37 | |
| cloths would help ;) | |||
| audreyt | offby1: donating money for keeping the new pugsmachine running would be nice | 22:38 | |
| offby1 investigates | |||
| I'm far too lazy to contribute code, but money I can handle. | 22:39 | ||
| audreyt | clkao: I've got a large enough deposit on Issey retailers that will buy me another year's clothes or so... | 22:40 | |
| offby1 | hmm, the dropdown only lists "Perl Development Fund", and three other less-likely-looking things | ||
| audreyt | obra: which fund would this be? | ||
| obra | i don't know for sure. offby1: for large donations, all kinds of magic can happen. | 22:41 | |
| but I know that's hard to pull off sometimes. | |||
| offby1 | obra: ain't gonna be large, unless your idea of "large" is much smaller than mine :-) | ||
| obra | understood | ||
| I've been trying to get pugs-donation stuff streamined. but i'm not tied in enough. | |||
| offby1 | ka-ching | 22:43 | |
| well, the perl foundation donation stuff was easy enough to do that I've just done it. | |||
| hope they throw it at y'all. | |||
| audreyt | eventually, indirectly, some of them will... | 22:44 | |
| offby1 | sounds like getting your paycheck after tax and social security has been deducted ... :-( | 22:46 | |
| SamB | maybe canned food? | 22:48 | |
|
22:48
SubStack joined
|
|||
| audreyt | mmm spam | 22:55 | |
| Limbic_Region raises his hands above his head in triumph | 22:56 | ||
| NP complete -> O(1) in < 8 hours | |||
| offby1 | ? | ||
| audreyt | Limbic_Region: yay! | 22:57 | |
| Limbic_Region | of course, uploading the db so I can show it off it going to take > 10 hours | 22:58 | |
| (slow upload)-- | |||
|
23:00
trym joined
23:03
Aankhen`` joined
23:05
jferrero is now known as JoaquinFerrero
23:13
bsb joined
23:19
root4o joined
23:20
jferrero joined,
jferrero is now known as JoaquinFerrero
|
|||
| putter | svn up, svn ci still down. | 23:29 | |
|
23:30
fridim left
|
|||
| putter | are there "using svk with pugs" instructions somewhere? | 23:30 | |
| (switching to svk is nominally on putter's todo list for this next week) | 23:31 | ||
|
23:31
offby1 left
|
|||
| putter | but for now, have fun all. & | 23:31 | |
|
23:37
SubStack joined
23:39
diakopter joined
23:44
[particle] joined
23:51
diakopter left
|
|||