pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, elf: etc.) (or perl6: for all) | irclog: irc.pugscode.org/
Set by Tene on 29 July 2008.
meppl good night 00:36
pugs_svn r22853 | lwall++ | [Cursor] some prep work for matching against NFG integer arrays 00:37
pugs_svn r22854 | pmurias++ | [smop] native capture returns less nulls 10:02
riffraff hi 13:32
pmurias riffraff: hello 13:36
riffraff: what brings you here on a quiet misty weekend? ;)
riffraff actually, a sunny weekend here in rome :) 13:37
so instead of going out for a walk I slack on irc ;)
pmurias riffraff: why do people on irc always have better weather? ;) 14:06
riffraff ahah 14:09
masak pmurias: statistical bias due to people not tending to brag about their bad weather? :) 14:17
masak @tell rindolf Don't know if you ever got help with the "No handler to delete" error, but at least half of the cause is that you're using the Rakudo REPL. See rt.perl.org/rt3/Ticket/Display.html?id=58258 14:27
hello? lambdabot? 14:28
:/
@tell rindolf The other half of your problem is a bug in Rakudo, one which I've now reported as [perl #60304]. Thank you for spotting it. 14:37
rindolf So who should I talk to about adding a test to Rakudo? 18:05
masak rindolf: have you backlogged? did you see my messages to you? 18:08
rindolf: adding a test to Rakudo consists of adding a test to the Perl test suite in the Pugs repo
rindolf masak: your messages to me?
masak rindolf: I'll get the for you from the past. 18:09
rindolf masak: what?
masak: do you have a different nick?
masak rindolf: no.
masak but you weren't online at that time 18:09
and lambdabot was absent too
rindolf masak: ah. 18:10
masak: ah, then how am I supposed to see your messages?
masak I'll post them here for you
(there's also an archive for this channel)
here's the first message:
Don't know if you ever got help with the "No handler to delete" error, but at least half of the cause is that you're using the Rakudo REPL. See rt.perl.org/rt3/Ticket/Display.html?id=58258
here's the second:
The other half of your problem is a bug in Rakudo, one which I've now reported as [perl #60304]. Thank you for spotting it. 18:11
for future reference, the IRC archives for this channel are at irc.pugscode.org/
...for easy backlogging :)
moritz_ re 18:13
masak moritz_: OH HAI
moritz_ @tell DarkAkorn I'm back now... do you still need help with the IRC logs? 18:28
masak lambdabot's down... we miss her :/ 18:29
moritz_ I noticed, yes ;)
masak & # food 18:33
rindolf masak: here? 19:46
masak rindolf: more or less, yes. 19:50
rindolf masak: I'd like to implement the @$array_ref feature in Rakudo - how do I go on?
I wrote a test file.
But where do I put it? 19:51
moritz_ where is it specced?
rindolf moritz_: I don't know.
masak rindolf: meet moritz_, our test suite pumpking
you're in good hands.
rindolf masak: :-) 19:52
moritz_ rindolf: try to grep in the spces where @() is mentioned...
for example S03 mentions it sever times
so it should go in t/spec/S03-operators/ somewhere 19:53
either t/spec/S03-operators/context-forcers.t or a new file
masak context-forcers.t sounds right to me. 19:54
rindolf Where are the spec/synopses repos? perlcabal.org/syn/ 19:55
moritz_ svn.perl.org/perl6/doc/trunk/design/syn/
rindolf What is the difference between context-forcers.t and context-forcers.rakudo ? 19:57
literal moritz_: that doesn't look up-to-date
e.g. S17 there is just a placeholder while the one on perlcabal.org has lots of stuff in it 19:58
mncharity It would be nice of Perl 6 had an active planet. It would provide greater visibility to the work which is happening.
moritz_ rindolf: the .rakudo file is generated from .t, and some tests might be TODOed or skipped
rindolf moritz_: ah.
moritz_ literal: some specs also live in the pugs repo, in docs/Perl6/Spec/ - that's a bit confusing :/ 19:59
mncharity planetsix.perl.org/ exists, but it (1) currently is only showing a single recent post, with apparently no way to get previous ones. So it's either broken, or badly misconfigured. (2) it's missing blogs. Eg perlgeek.de/blog-en/perl-6/ . 20:00
I'd create an Elf blog as well, if there were a place to put it.
Perhaps a smop blog as well? 20:01
masak mncharit++ # ideas
mncharity So, any thoughts on who maintains/has access to planetsix.perl.org/ ? With objectives of get moritz's blog added, and have some history accessible? 20:02
moritz_ mncharity: since it's perl.org, ask [email@hidden.address] 20:03
mncharity: I think I suggested a blog once, and it got accepted immediately
(but I won't suggest my own blog, it doesn't feel right ;)
mncharity masak: :)
masak TimToady: S02 says: "If you wish to supply a C<Pair> as a key, you must compose an outer C<Pair> in which the key is the inner C<Pair>: %hash = (($keykey => $keyval) => $value);". But Rakudo doesn't require the parens to produce such a hash. Are the parens superfluous and just for clarification, or does Rakudo take liberties with the precedence? 20:04
mncharity moritz_: will do. I've your permission to suggest your blog? Suggested name? 20:04
moritz_ mncharity: sure you have. Just use my name (Moritz Lenz), since all others seem to use their name as well ;) 20:05
mncharity ok. in progress. 20:06
moritz_ mncharity++ # taking action 20:07
masak rindolf++ # writing tests 20:09
mncharity moritz_: perlgeek.de/blog-en/perl-5-to-6/ seems to have p6 content as well. and which doesnt appear in perl-6/. what's the right thing? 20:10
TimToady masak: => is supposed to be right associative, not left
masak TimToady: i.e. Rakudo has it wrong.
TimToady yes
masak files rakudobug mail
moritz_ mncharity: both ;) perl-5-to-6 is more tutorial and feature show-off, perl-6/ is news.
mncharity: perl-5-to-6 needs dupe filtering, because the rss feed sometimes has dupes 20:11
rindolf masak: www.shlomifish.org/Files/files/code...-test.diff 20:12
masak: thanks. :-) 20:13
masak rindolf: the tests themselves look good, but since Rakudo doesn't yet implement @$x...
...they will not only not pass, but fail to parse.
rindolf masak: I'm trying to implement it.
moritz_ uhm, I don't know if @$x[1] does what you mean, because [] has tighter precedence
masak moritz_: good point. 20:14
moritz_ and in general that's not wrong, because you can just say $x[1]
so you'd neet to write that as @($x)[1] instead
s/neet/need/
masak rindolf: do you think you could add a fudge directive? if you do, I can commit your patch. (unless you already have commit rights to the Pugs repo)
rindolf masak: actually, I used to have a commit bit on the Pugs repos. 20:15
masak rindolf: moritz_ is right. there's no need in p6 to convert to array before indexing
rindolf: did you misplace it?
rindolf masak: OK.
masak: no, I don't think I did.
masak rindolf: why then "used to"? 20:16
rindolf masak: well, maybe it was removed or something.
masak I've never heard of anyone being deprived of his commit bit.
rindolf It should be still active.
masak aye.
rindolf masak: fine, I'll remove the first two tests that I added. 20:17
masak ...and add a fudge directive. kthxbai.
rindolf masak: OK. 20:18
masak: but I'd like to try to implement it in Rakudo now
masak rindolf: :)
rindolf: I'd advice you to hang out on #parrot on irc.perl.org in that case. 20:19
moritz_ rindolf: I don't want to stop you, but pmichaud is going to re-work the list context stuff anyway
rindolf moritz_: ah.
masak was going to say "aw, pmichaud gets all the good stuff!", but realized that this is actually a good thing. pmichaud++ 20:20
masak heh, I wrote a whole bug report about Rakudo's => being left-associative, before I realized that it's right-associative after all... 20:23
mncharity moritz_: err, where is the feed for perlgeek.de/blog-en/perl-6/ ? 20:25
moritz_ mncharity: perlgeek.de/blog-en/perl-6/index.rss 20:26
it's advertised wrongly on the page itself :/
mncharity sent. yay progress. 20:31
masak progress++ 20:42
mncharity I hope to get back to elf sometime this week. Perhaps first with a spot check of how STD is coming along. When gimme5 is doing useably correct parses of elf and STD.pm (perhaps already is?), then I clearly become the elf bootstrap blocker, and I expect I'll allocate some solid days to get it done. Else perhaps back to low intensity hacking on the rx_on_re->elf merge, in preparation for later STD. 20:46
May also attempt a Boston.PM hacking evening. If boston.rb can consistently manage it, why not boston.pm? Could do an elf night. 20:47
@tell ruoso Any interest in doing a smop blog? For planetsix.perl.org/ ?
lambdabot taking Sunday off? 20:48
moritz_ ENOƛBOT
masak lambdabot down :(
mncharity snif
masak it's a foretaste of how helpless we'll be ten years down the road, when our bots desert us... 20:49
mncharity nifty character between ENO and BOT.
lol
masak mncharity: yes, λ :)
moritz_ it's lambda + stroke
mncharity I'm looking forward to getting my first email from someone I kown is already deceased.
ahhh 20:50
moritz_ I didn't find the normal lambda in my char map :/
masak moritz_: why the stroke?
ah.
moritz_ because somebody stroke down lambdabot? ;-)
mncharity so much unicode fun to look forward to
masak I just did 'C-u C-\ greek RET l' :)
emacs++ 20:51
mncharity I'm off. happy hacking. & 20:52
rindolf Hmmm.... @($x) works. 20:55
moritz_ rakudo: my $x = [1, 2, 3]; say @($x).elems 20:56
p6eval rakudo 31963: OUTPUT[3␤]
moritz_ rakudo: my $x = [1, 2, 3, 0]; .say for @($x);
p6eval rakudo 31963: OUTPUT[1 2 3 0␤]
masak yes, but that's wrong :)
it loops over it as if it were one element 20:57
don't worrt, I've reported that to rakudobug :P
s/worrt/worry/
masak perl6: my $x = 'a'; die "Something's wrong" if !($x eq 'a' | 'b') 21:31
p6eval pugs: OUTPUT[*** Something's wrong␤ at /tmp/wWuLw0Rc99 line 1, column 14-38␤]
..elf 22847: OUTPUT[Something's wrong at (eval 123) line 4␤ at ./elf_f line 3861␤]
..rakudo 31963: RESULT[Bool::False]
masak my Rakudo r32280 says "Something's wrong". can anyone confirm? 21:32
moritz_ masak: mine too (+ a backtrace) 21:33
masak yes, sure.
good, then it's not just me.
this is a regression of some sort.
don't we have a test for this? 21:34
masak runs spectest_regression
moritz_ masak: that's the "boolean context doesn't collapse junctions" TODO 21:35
masak oh.
I'm pretty sure it used to work.
we had such code in November.
...which now fails :/
moritz_ rakudo: my $x = 'a'; say perl(!($x eq 'a' | 'b')) 21:36
p6eval rakudo 31963: OUTPUT[Could not find non-existent sub perl␤current instr.: '_block11' pc 86 (EVAL_12:28)␤]
masak moritz_: try the method.
moritz_ rakudo: my $x = 'a'; say (!($x eq 'a' | 'b')).perl
p6eval rakudo 31963: OUTPUT[Bool::False␤]
moritz_ uhm.
masak Something's wrong.
moritz_ rakudo: my $x = 'a'; say "yes" if (!($x eq 'a' | 'b'))
p6eval rakudo 31963: RESULT[Bool::False]
moritz_ rakudo: my $x = 'a'; say "yes" if (!($x eq 'a' | 'b')); say "done" 21:37
p6eval rakudo 31963: OUTPUT[done␤]
masak why the outer pair of parens?
moritz_ historical artifact 21:38
masak rakudo: my $x = 'a'; die "Something's wrong" if !($x eq 'a' | 'b')
p6eval rakudo 31963: RESULT[Bool::False]
masak moritz_: perl 5 doesn't need that either.
dang, how come the bug is suddenly gone?!
oh, right. it never showed up here on p6eval... 21:39
that must mean that we were fine by r31963.
how come p6eval runs such an old version of Rakudo, btw? 21:40
moritz_ svn.perl.org firewall :/
masak ach so. 21:41
moritz_ I'm now re-checking the cronjobs
:q
wrong window
masak you can quit IRC, but you can never leave >:) 21:42
moritz_ "you can checkout any time you like, but you can never leave" - was it that you were thinking of? ;-) 21:45
masak moritz_: Hotel California, aye. 21:46
moritz_ very cool song ;) 21:51
masak one of the best. 21:55
masak moritz_: are you sure this regression is part of the TODO you mentioned? even considering that it appeared after r31963? 22:00
moritz_ masak: no, not sure. 22:01
moritz_ there are now some passing skipped tests in t/spec/S03-operators/precedence.t 22:02
but I'm not quite sure that the tests are right :(
masak moritz_: the symptoms do seem to indicate that junctions aren't coverted to a boolean 22:03
pugs_svn r22855 | moritz++ | [t/spec] some unfudgings for rakudo
masak $ perl6 -e 'my $x = "a"; say (!($x eq "a" | "b")).perl' gives 'any(Bool::False, Bool::True)' here
moritz_ that should not happen[tm]
mncharity planetsix.perl.org/ Robert++ 22:05
masak moritz_: but you say it's reported already? what's the ticket number? 22:09
moritz_ masak: dunno, but it's definitively a TODO to implement proper autothreading in the dispatcher. 22:10
masak: I don't think there's a TODO for this particular item.
masak writes something about it to rakudobug
moritz_ rakduo: my $x = "a"; say (!($x eq "a" | "b")).perl
rakudo: my $x = "a"; say (!($x eq "a" | "b")).perl 22:11
p6eval rakudo 31963: OUTPUT[any(Bool::False, Bool::True)␤]
moritz_ I CAN HAZ NEW RAKUDO ON SRVR
it's just reported as old :/ 22:12
should say 32288 in a few minutes, though
masak moritz_++ 22:13
moritz_ rakudo: say "hi"
p6eval rakudo 31963: OUTPUT[hi␤]
masak now we'll get all the latest regressions, right here in the channel!
moritz_ how write the perl6 RT queue bot? 22:14
masak sorry, what?
(that didn't parse)
moritz_ s/how write/who writes/ 22:15
moritz_ should go to bed now

Powered by Cro and the Raku® Programming Language.

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

Thank you!