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.
00:22 avarab joined 00:24 nekokak joined 00:25 hexmode joined 00:30 avarab_ joined 00:33 seano left 00:51 Pominx joined 01:01 thepler joined 01:08 AzureStone joined 01:35 diotalevi__ joined 01:48 frankg joined, nine joined, jdv79 joined, rafl joined, Ikarus joined, svnbot6 joined, nipotaway joined, gnuvince joined, jojo joined, lisppaste3 joined, mdiep joined, hexmode joined, spinclad joined, scw joined, Shabble joined, jiing joined, Narcisse joined, meppl joined, masak joined, Odin- joined, lambdabot joined, osfameron joined, weinig joined, araujo joined, dduncan joined, nothingmuch_ joined, thepler joined, demq joined, diotalevi__ joined, Pominx joined, avarab_ joined, nekokak joined, dmq joined, Alias_ joined, justatheory joined, cmarcelo joined, nperez joined, explorer joined, ruoso joined, Daveman joined, Eimi joined, takanori joined, prism joined, theorbtwo joined, beppu joined, cjeris joined, idiotmax joined, Patterner joined, Lorn joined, xinming_ joined, penk joined, cognominal joined, BooK joined, shachaf_ joined, Yappo joined, Eidolos joined, eviltwin_b joined, buubot joined, TimToady joined, baest joined, [particle] joined, gugod joined, Gothmog_ joined, stevan joined, buu joined, ingy joined, ajs_home joined, Odin-LAP joined, evalbot_r14413 joined, mugwump joined, ofer0 joined, fridim joined, kcwu joined, ajs joined, norageek joined, stennie joined, spoop joined, awwaiid joined, Grrrr joined, takesako joined, kalven joined, knewt joined, Khisanth joined, holoway joined, kane-xs joined, ashleyb joined, jabbot joined, woremacx joined, mako132 joined, chip joined, miyagawa_ joined, drbean joined, autark joined, pasteling joined, sri_ joined, f0rth joined, ruz joined, kane-xs_ joined, obra joined, clkao joined, broquaint joined, Maddingue joined, cookys joined, mtve joined, GeJ joined, SamB joined, integral joined, tewk joined, rgs joined, arguile joined, lidden joined, TreyHarris joined, j0sephi joined, silug joined, dvorak joined, statico joined, orafu joined, revdiablo joined, perlbot joined, rafl_ joined, hcchien joined, cj joined, kolibrie joined 01:54 shachaf_ joined 01:55 lambdabot joined 02:01 cjeris left 02:06 nipotaway is now known as nipotan
svnbot6 r14422 | audreyt++ | * Pugs.Lexer: Avoid the annoying and not helpful "expecting comment" 02:17
r14422 | audreyt++ | in parse errors.
cmarcelo audreyt: ping 02:19
audreyt sorry can't stay
lambdabot audreyt: You have 1 new message. '/msg lambdabot @messages' to read it.
audreyt at baggage claim
will be back in 3hr 02:20
cmarcelo ok
svnbot6 r14423 | audreyt++ | * Pugs.Parser: Implement the "Closing hash curly may not terminate a line"
r14423 | audreyt++ | rule as mandated by the new spec.
cmarcelo forgot "offline commit in a plane" svk use case :o) 02:21
eviltwin_b knows a number of people who're prone to that 02:22
(hm, most of them are here, come to think of it :> )
02:40 weinig is now known as weinig|zZz 02:45 buu joined 02:50 buubot joined 02:55 agentzh joined, mako132_ joined 02:57 SCalimlim joined
agentzh class diagram for Moose: perlcabal.org/agent/moose.png 02:58
generated from a single command: $ perl rt2png.pl moose Moose "Moose|MOP" 02:59
audreyt, stevan, nothingmuch: please check the sanity of perlcabal.org/agent/moose.png :) 03:02
TreyHarris what's the Perl 5 equivalent of / (\d+) { $0 < 256 or fail } / ? is there one? 03:03
audreyt yes it involves (??{...}) 03:04
qr/(\d+) (??{$1 < 256 ? '(?!)' : ''})/x 03:05
or some such.
TreyHarris audreyt: welcome to the united states :)
audreyt boarding plane to portland in 5min now... bbl 03:06
agentzh fglock: class diagram for Pugs::Compiler::Perl6: perlcabal.org/agent/pcr.png it's generated automatically by my plotter based on PPI and Class::Inspector. not sure if it's sane. :) 03:09
these pictures are a bit huge, so they may look fuzzy in the web browser. ;) 03:17
svnbot6 r14424 | raptor++ | action 'if' cond; - added
Khisanth agentzh: not with a big enough screen! :P
agentzh Khisanth: nodnod 03:18
:)
Khisanth although it definitely does NOT work when you only have 256MB of memory
agentzh indeed 03:19
i'll upload the diagram generator to cpan (possibly with the name UML::Class) once it's mature enough.
SamB Khisanth: oh well! I never even build pugs anymore anyway. 03:20
TreyHarris I think the actual equivalent in perl 5 is qr/((?<!\d)\d+(?!\d)) (??{$1 < 256 ? '' : '(?!)'})/x
Perl 6 regexes are soo much nicer :) 03:21
SamB that does look pretty damn nice, as regexes go!
TreyHarris SamB: "/ (\d+) { $0 < 256 or fail } /" looks nice? or you were sarcastically saying that that perl 5 monstrosity looks nice? or you really think it does look nice? :) 03:22
SamB the perl6 version!
TreyHarris ah. k, yes :) 03:23
SamB what kind of fool do you take me for?
TreyHarris though.... hmmmm.... maybe that's not the equivalent perl 5. do any of the current engines compile the above RE?
SamB Even a (former?) Python user can recognize a nice looking regular expression when he sees one... 03:24
TreyHarris Was thinking about the "2360" case, which I don't know would match 236 or not.
Khisanth TreyHarris: strange RE you got there
TreyHarris Khisanth: it's from S05.
Khisanth I mean the perl 5 version 03:25
TreyHarris oh, how so, is there a better way?
i'm feverish with flu, so i'm willing to bet i missed something obvious
Khisanth I mean the \d+ is already greedy, the look arounds appear to be completely unnecessary
eviltwin_b I was wondering about that myself 03:26
TreyHarris Khisanth: oh? try it, leave them out, it matches 236 when you have 2360
Khisanth hmm 03:27
I guess the (??{}) is forcing it to backtrack :)
TreyHarris yes, of course :)
eviltwin_b oh, I see, /x is on by default in perl6 I take it
Khisanth yes 03:28
TreyHarris being able to finely control backtracking is a Perl 6 neologism
eviltwin_b and yes, it would backtrack and try again; that's the nature of regex matching
TreyHarris eviltwin_b: you need to read S05... backtracking is no longer "the nature" of matching; as i just mentioned, you can control it pretty finely in Perl 6
eviltwin_b well, it's certainly the nature of perl5 (and other regexp engines) backtracking, which was the actual subject there 03:29
TreyHarris *nod* yes
you're right of course. 03:30
must catch my bus, i will see scrollback when i get home & 03:32
eviltwin_b I admit to being well behind the curve on p6 stuff, need to do a lot of reading up :> 03:33
TreyHarris eviltwin_b: smartlinks make it MUCH faster to get up to speed
eviltwin_b (for those wondering wtf the crazy is: see getcwd.pl in perl5 :) 03:34
TreyHarris i'd actually recommend you read the specs that way: go to smoke.pugscode.org, find some recent smoketest with a SYN link, and use that to read the synopses. then you'll know as you're reading what's implemented yet and what's not
& really now
eviltwin_b I think that's the only contribution I made that had any staying power (and was about the last thing I expected to stick around...)
04:31 SubStack joined
cmarcelo @tell audreyt re moh: my next step was going forward and implement stuff related to classes (MO::Compile::Class::MI et al) but some design question on how to that come up, the main one being mapping (to Hs) the idea of subclassing and "rw" attributes ("just go for IORef"?). any ideas on this? or Pugs code that deals with related stuff? I'm looking now at OOHaskell paper (pointed by SEITC) to gather ideas. 04:44
lambdabot Consider it noted.
cmarcelo @tell nothingmuch (maybe you have some ideas on my questions to audreyt, see logs near this msg...) 04:46
lambdabot Consider it noted.
cmarcelo lambdabot: tks
04:48 mr_ank joined 04:56 seano joined 04:57 stennie left 05:15 BooK_ joined 05:25 buetow joined 05:31 justatheory joined 05:34 justatheory joined 05:39 frankg_ joined 05:44 frankg_ is now known as frankg 05:54 daz_buubot joined 05:57 b00t joined 05:58 glasser joined 05:59 daz_buubot joined
audreyt safely arrived to Schwern's. 06:10
going to sleep now :) be back in 10 hours or so
@tell cmarcelo yes IORef is the way to go. in RealPugs it'd be TVar but it's the same
lambdabot Consider it noted.
clkao audreyt!
audreyt hi clkao 06:12
clkao: sorry can't stay up any longer :) turns out the gel control is based on individual bottle's volume, not the total amuont, so I get to pass intact with crazy amount of stuff in many small bottles 06:13
06:13 ofer0 joined
clkao audreyt++ 06:14
eviltwin_b the terrorists have won...
06:15 daz_buubot joined 06:26 Gothmog_1 joined 06:31 nox_ghost joined 06:36 buubot joined 06:46 iblechbot joined 06:49 marmic joined 06:50 daz_buubot joined 06:53 daz_buubot joined 06:58 daz_buubot joined 07:03 DHGE joined 07:09 daz_buubot joined 07:21 zakharyas joined 07:37 explorer joined 07:41 DHGE joined 08:03 drrho joined 08:52 dakkar joined 08:53 DHGE joined 09:00 andara joined
nothingmuch_ @tell cmarcelo in the first revision of MO all the is rw were is ro. I think we should have an is ro version, where it's just records, and a tvar/IORef version that is modifyiable by the runtime. when moh steps into action the IORef-y version is simplified into the read only version, so that the other code can stay purely functional 09:09
lambdabot Consider it noted.
09:12 reZo joined 09:38 dduncan left 09:39 flw joined 09:44 nothingmuch joined 10:02 buetow joined 10:04 anatoly1 joined, anatoly1 left 10:26 ruoso joined 10:36 elmex joined 10:42 trym|work joined 10:44 avarab_ is now known as avar 11:07 ofer0 joined 11:14 xinming joined, chris2 joined 11:17 Psyche^ joined 11:19 agentzh joined
agentzh [particle]: your SmartLink.pm looks like a poetry. :) 11:20
11:20 thepler joined
agentzh i said to myself while reading your code, "i should have written smartlinks.pl that way..." 11:21
and i'm also delighted to see my UML class diagram generator can also process your SmartLink.pm, a Moose-based modern perl 5 OO module: 11:22
svn.berlios.de/svnroot/repos/unisim...rtlink.png
lambdabot tinyurl.com/yhye4r
agentzh goes back to slide-making. 11:25
too many classes, too many talks...
sigh.
avar Is there any $" in Perl 6? 11:33
11:33 Psyche^ is now known as Patterner
avar I.e. is "@arr"; { local $" = ":"; "@arr" } now "@arr[]" and @arr.join(":") ? 11:34
agentzh i hope not. :) 11:37
audreyt no. closest is overriding array stringification locally 11:38
which is cleaner
or even overriding infix:~ locally... 11:39
11:39 iblechbot joined
avar I'm not asking for it. But since perlvar isn't finished this sort of thing isn't very clear:) 11:39
What's the output of: 11:40
$ grep url .svn/entries
for the rw pugs repos, I'm using the googlecode one and I want to commit a tyop fix:)
audreyt I think $_ $! $/ are the only 3
googlecode one is not kept uptodate
svn.openfoundry.org/pugs/ is rw 11:41
lambdabot Title: Revision 14424: /
avar yay
avar nuked his svn:) 11:43
11:43 mdiep_ joined
avar nd -type f -name 'entries' -exec perl -pi -e 's[url="pugs.googlecode.com/svn/trunk"][url...org/pugs]' {} \; 11:43
lambdabot pugs.googlecode.com/svn/trunk"][url...org/pugs]'
avar missing @ :)
s/@/"/
audreyt not sure it's going to work 11:44
11:45 moeko joined 11:46 AzureStone joined
avar is my openfoundry svn passwd the same as the website passwd? 11:57
11:59 autark joined
audreyt yes 12:04
12:05 fglock joined 12:11 AzureStone joined 12:12 Limbic_Region joined
nothingmuch avar: svn switch does the same thing 12:13
lambdabot nothingmuch: You have 1 new message. '/msg lambdabot @messages' to read it.
nothingmuch @moosages
lambdabot cmarcelo said 7h 26m 37s ago: (maybe you have some ideas on my questions to audreyt, see logs near this msg...)
nothingmuch --relocate 12:14
avar avar@Arsia:/usr/local/src/pugs$ grep url .svn/entries url="svn.openfoundry.org/pugs" 12:22
lambdabot Title: Revision 14424: /
avar I can't commit with the same passwd that I use to log in to rt.openfoundry.org
avar@Arsia:/usr/local/src/pugs$ svn ci -m'* Tyop: "an one-element" => "a one-element"' 12:23
Authentication realm: <svn.openfoundry.org:80> RT-Foundry Login
Password for 'avar':
Sending docs/Perl6/Overview/Data.pod
Authentication realm: <svn.openfoundry.org:80> RT-Foundry Logi
audreyt that is right 12:25
just type in the same passwd again
svnbot6 r14425 | audreyt++ | * util/version_h.pl: Don't run `svn info` if there's no .svn.
avar audreyt: Doesn't help, sadly 12:26
I'll poke it later mmmaybe
avar.lir.dk/~avar/Data.pod.txt
Limbic_Region audreyt is safe and sound in Portland? 12:27
audreyt Limbic_Region: yup
Limbic_Region how is the jet lag?
audreyt avar: hmm weird. reet your passwd on web and try again?
oh and in any case the Data.pod is now ours
it's from upsteam svn.perl.org
which is why you can't commit
avar arr
audreyt I think
oh wait 12:28
no it's Overview not Spec
never mind me!
Limbic_Region oh, audreyt - I meant to @tell you. I added seano last night who already had a foundry acct (which I didn't realize initially)
audreyt it's just 5:28am in the morning
avar audreyt: hihi:)
Limbic_Region when I sent an invite to his email - it was never received
avar I reset my passwd earlier, been trying with the new one
Limbic_Region and searching for an existing openfoundry member wasn't very intuitive
but you probably have very little if any involvement anymore huh
audreyt avar: hm. try the old one ?
avar In any case I should get to $wrk, and I can just post patches anyway:) 12:29
audreyt Limbic_Region: indeed
avar audreyt: yeah:)
Limbic_Region well, it worked out in the end so I guess it doesn't matter
svnbot6 r14426 | audreyt++ | * Perl6::Overview::Data: s/an one-element/a one-element/ from avar++
avar wee! 12:30
12:33 mj41_ joined
fglock agentzh: wow - it looks much more complex in this format :) re colabti.de/irclogger/irclogger_log/...el=19#l165 12:41
lambdabot Title: #perl6 2006-10-19,Thu, tinyurl.com/yz8aj8
Limbic_Region um, my pugs appears to be fubar 12:56
perlbot nopaste
perlbot Paste your code here and #<channel> will be able to view it: sial.org/pbot/<channel>
pasteling "Limbic_Region" at 129.33.119.12 pasted "output from smoke" (56 lines, 2.5K) at sial.org/pbot/20556 12:57
13:06 vel joined 13:10 Psyche^ joined, Psyche^ is now known as Patterner 13:13 newbie joined
newbie 􏿽xA3􏿽xBF 13:14
some
13:14 newbie left 13:23 seano joined, ofer0 joined
svnbot6 r14427 | fglock++ | Pugs-Compiler-Rule - fixed '>>' token 13:36
r14428 | fglock++ | v6 - merged prefix::<$> from misc/pX 13:39
13:44 vel joined
xinming pasteling: url 13:46
hmm,
fglock sial.org/pbot/paste 13:48
xinming paste.lisp.org/display/28291 13:50
fglock: thanks
fglock: I went to #emacs for the url. ;-)
avar svnbot would be awesomer if it included info on what files were changed 14:01
[particle] isn't there a mailing list for commit logs?
(parrot has one)
svnbot6 r14429 | yiyihu++ | r14582@HomePc: xinming | 2006-10-19 22:00:53 +0000 14:03
r14429 | yiyihu++ | A ugly hack of fixing a compliation failure bug in pugs.
xinming hmm, I don't think the bug is fixed anyway. >_< 14:04
xinming goes to recompiling again. 14:05
14:05 hexmode joined
nothingmuch www.youtube.com/watch?v=NZNTgglPbUA&eurl 14:05
lambdabot tinyurl.com/z645u
14:08 elmex joined 14:14 elmex_ joined 14:15 cjeris joined 14:31 nipotan is now known as nipotaway 14:35 andara left 14:42 weinig|zZz is now known as weinig|bbl 14:57 andara joined
andara PUGS_SAFEMODE=1 pugs fails for r14429 ... 14:59
fglock stevan: ping 15:00
15:00 explorer joined
fglock @tell stevan you have studied Scheme, right? - does it make sense to have a Scheme backend? 15:05
lambdabot Consider it noted.
eviltwin_b next it'll be prolog :) 15:09
TimToady cool. 15:11
eviltwin_b: do you have a commit bit yet?
eviltwin_b hm? I'm not actually working on anything, nor am I in any position to do so. maybe when I've caught up on things (as if...) 15:12
TimToady all the rest of us are in the same state, but that doesn't stop us. :) 15:13
we all just change things on the fly while we're trying to catch up on things... 15:14
andara fglock: a core Scheme backend would be very nice
fglock I wonder how it compares to Haskell 15:15
andara fglock: in what sense?
fglock portability, speed, memory use ... 15:16
TimToady eviltwin_b: allbery (at) ece cmu etc.?
stevan fglock: scheme might be nice
lambdabot stevan: You have 1 new message. '/msg lambdabot @messages' to read it.
stevan @moosages
lambdabot fglock said 10m 55s ago: you have studied Scheme, right? - does it make sense to have a Scheme backend?
eviltwin_b yes 15:17
stevan fglock: scheme should have most of the features needed,.. but a full backend effort is quite a lot of work
TimToady eviltwin_b: commit bit is on its way. while you're "catching up", add your name to AUTHORS as first checkin is the tradition. :) 15:19
fglock stevan: can we reuse the object system?
TimToady and have fun 15:20
theorbtwo Tangentially related to perl6 and pugs: blog.moertel.com/articles/2006/10/1...gs-problem
lambdabot Title: A type-based solution to the "strings problem": a fitting end to XSS and SQL-inj ..., tinyurl.com/uv94m
theorbtwo (Haskell-based.)
TimToady 散歩します。＆ 15:24
theorbtwo wonders if we need a babelbot. 15:25
(JA): It walks. &
dakkar better: taking a stroll &
theorbtwo That was google translate. I'm pretty well monolingual. 15:26
fglock lunch & 15:27
dakkar theorbtwo: mine was with a dictionary ;-) 15:28
15:28 weinig|bbl is now known as weinig 15:35 nox_ghost joined
stevan fglock: Scheme does not have an object system 15:36
it has *many* object systems :P 15:37
LISP has an "official" one with CLOS
15:37 drrho joined
stevan but object systems in Scheme/LISP are like templating systems in Perl,.. everyone has to write one eventually 15:37
15:41 vel joined 15:44 buetow joined 15:57 rlb3 joined
fglock stevan: I mean, can we reuse tiny-clos (or other) ? 16:09
16:10 chris2 joined
stevan fglock: potentially 16:13
we would still need to add the Role stuff on top
and a few other bits which CLOS does not have
16:13 Al__ joined, Al__ left 16:16 Corion joined, Corion left 16:20 dakkar left
fglock stevan: people.cs.uchicago.edu/~robby/pubs/...06-fff.pdf - about mixins & traits in Scheme 16:22
lambdabot tinyurl.com/y3h7qa
16:29 Limbic_Region joined 16:38 Odin-LAP joined
stevan fglock: very nice :) 16:46
fglock: what is the motivation to do this? 16:48
fglock stevan: I'm just looking around if there are other alternative backends 16:49
16:50 weinig is now known as weinig|bbl 17:21 seano joined 17:32 ruoso joined 17:43 takanori_ joined
TimToady getting a very consistent error: user error (unhandled element: EStr "~") 17:48
xinming TimToady: :-P I don't know if It is introduced by my ";" :-P 17:49
TimToady: if I don't add ';' It wouldn't even compile.... >_< 17:50
TimToady I believe it's a yaml error 17:52
but the usual two culprits are up-to-date
17:57 kanru joined, justatheory joined
demq will perl6 have something like perl5s \C regex pattern? 18:00
can we remove it from perl5.10?
18:01 penk joined
TimToady [:bytes .] but will fail if string doesn't present a bytes interface. 18:02
demq so will it allow access to the byte representation of the internal string? 18:03
because \C can be used to make perl get very unhappy. 18:04
since it can absorb a byte of a multibyte sequences making any further processing of the string dubious.
i wonder why it was added actually. 18:06
eviltwin_b struck me as a quick hack to enable byte processing instead of unicode. not well thought out 18:07
then again, that's perl <6 for you :>
demq yeah, it looks to me like a "we can do this, so why not?" without bothering to do any research on the why not. 18:08
im talking about \C not whatever perl6 will do.
it looks like nobody has used \C either. at least judging by the google code search 18:10
im assuming that perl6 wont include a way to make a regex cause perl to segv. 18:15
18:15 c6rbon joined
TreyHarris demq: sure... "/ { dump; } /" ;) 18:24
18:25 elmex joined
demq perl6 will have the same dump keyword? 18:25
Juerd Possibly.
[particle] waves to TreyHarris
PerlJam I hope perl6's dump means something different from perl5's though :) 18:26
Juerd demq: Normal strings will probably not have a bytes interface. Buffers will, and they won't have codepoint interfaces. 18:27
demq thats too bad. its a wasted keyword mostly.
Juerd demq: It's possible to have a string that presents both interfaces, but this requires extra care, either by the type's implementation, or the user.
I personally think it makes no sense to have a string present itself both as bytes and as characters.
TreyHarris PerlJam: there will, I assume, be *some* way to call dump(), even if it isn't called that. 18:28
TreyHarris waves to [particle]
Juerd PerlJam: Why? What's wrong with dump?
(The function, not its name) 18:29
PerlJam Juerd: nothing is wrong with the functionality, just the name really 18:30
eviltwin_b mmm, namespace wastage
Juerd PerlJam: I agree. It should be somewhere in Internals:: 18:31
(In perl 5)
TreyHarris or POSIX maybe? does win32 support core dumps?
18:31 weinig|bbl is now known as weinig
TreyHarris POSIX.dump() would be fine 18:32
since I can't see one wanting to actually dump (in the .perl or .yaml sense) POSIX...
TreyHarris &'s to meetings 18:34
18:35 vel joined
TimToady Juerd: strings with multiple abstraction levels make sense only if string positions and lengths are not numbers but opaque abstractions. 18:37
and to the extent you treat them as numbers, they must be numbers that know which unit of abstraction is intended. 18:40
wolverian what do you say to a person who calls state an improper abstraction? 18:42
demq depends who they are 18:44
wolverian the opponent :) 18:45
TimToady you say, "But, but, all abstractions are improper to an aristotellian..."
demq i think you ask them to clearly define proper abstraction, and then wander off before they can finish.
wolverian my knowledge of the classics is lacking. 18:47
(er, thanks!)
PerlJam installs ghc 6.6 to witness the big speed-up 18:50
18:58 weinig is now known as weinig|bbl
Juerd TimToady: Opaque abstractions and byte level manipulation don't go well together. 19:01
19:02 spoop joined
PerlJam Juerd: That opaque CPU abstraction seems not to mind ;-) 19:02
eviltwin_b kinda wishes darwinports had 6.6. or even qa working ghc-devel (currently some 6.5 but gets me a patch error) 19:14
PerlJam ghc 6.6 certainly doesn't compile quickly does it? 19:16
19:16 buubot joined
PerlJam I think pugs compiles faster. 19:17
Limbic_Region pugs certainly runs faster 19:21
svnbot6 r14430 | andara++ | [runpugs] 19:38
r14430 | andara++ | -improved error handling (try not to die).
r14430 | andara++ | -Web::Terminal: new generic version with command history, supports GHCi and
r14430 | andara++ | many more. (not live yet)
19:44 larsen_ joined 19:45 BooK joined 19:47 andara joined
svnbot6 r14431 | fglock++ | v6/emitter(2) - added lightweight coro implementation example 19:54
19:57 andara left
svnbot6 r14432 | audreyt++ | * DrIFT.YAML: Emit better error messages than "unhandled EStr ~" 20:06
r14433 | audreyt++ | * Pugs.AST.Internals: Serialize Pragma correctly, and bump CompUnit version.
fglock re r14431: it's roughly based on www.bookshelf.jp/texi/onlisp/onlisp_21.html 20:13
lambdabot tinyurl.com/y6qxyu
fglock and on the implementation of <prior> in PCR
20:14 lisppaste3 joined 20:26 ruoso joined 20:35 ToniEisner joined, vel joined 20:36 ToniEisner left, multic joined 20:37 bernhard joined 20:42 weinig joined 20:57 fglock left 21:02 mauke joined
TimToady @tell audreyt it appears the "Closing hash curly may not terminate a line" error happens before isHashOrPair gets a chance to decide whether the block is in fact a hash block. 21:09
lambdabot Consider it noted.
cj is there anything fancy I need to do to commit code to the pugs repo?
shouldn't I be able to use the credentials I set up for the ossf page?
maybe it takes a while for the credentials to propagate 21:10
TimToady dunno offhand
audreyt maybe the auth failed somehow? 21:11
lambdabot audreyt: You have 1 new message. '/msg lambdabot @messages' to read it.
audreyt TimToady: can you try a commit and see if it works for you?
cj: try switch to svn.openfoundry.org/pugs/ and see if it works?
cj thanks audreyt 21:12
audreyt @messages 21:13
lambdabot TimToady said 4m 23s ago: it appears the "Closing hash curly may not terminate a line" error happens before isHashOrPair gets a chance to decide whether the block is in fact a hash block.
audreyt oops. fixed 21:14
svnbot6 r14434 | lwall++ | trivial change please ignore
21:14 ofer0 joined
cj wow. she is fast. 21:14
TimToady yes, commit works for me
src/Pugs/Parser/Util.hs:169:4: parse error on input `where' 21:16
svnbot6 r14435 | audreyt++ | * Pugs.Parser.Util: Repair the hash-compose-at-end-of-block
r14435 | audreyt++ | detection to not misfire on non-hash blocks. Reported by TimToady++.
r14436 | audreyt++ | * Oops, typo. :)
audreyt those pesky right parens... they always insist to balance
cj hurm. failed again. 21:17
$ grep url .svn/entries
url="cjcollier@svn.openfoundry.org/pugs"
I'll wait a few and try again. I just set up the user no more than an hour ago
audreyt weird. it shouldn't take any time to propagate 21:18
cj maybe it shouldn't be cjcollier@...
audreyt shouldn't have mattered
cj wouldn't think...
21:18 yuval joined
audreyt gugod: are you still in touch with ossf code/people? it appears new users cant commit suddenly 21:19
TimToady cj: the admin page says you're still unregistered
cj hopes it's not a PEBKAC
audreyt no you're the second one to report
demq move the chair and try again.
audreyt will have to wait another 5hr though
demq :-)
cj ah. maybe I need to register with the project and not just the site?
audreyt demq++
heh indeed!
there we go 21:20
gugod: false alarm, please discard
cj: commit again?
cj checking out again. give it a few... 21:21
TimToady audreyt: getting a lot of *** Cannot cast from VUndef to Pugs.AST.Internals.VCode (VCode) at blib6/lib/Test.pm line 278, column 5-17
cj Committed revision 14437.
thanks :)
svnbot6 r14437 | cjcollier++ | - adding myself (CJ)
audreyt looking 21:22
TimToady looks like Bool needs to be Maybefied maybe
yuval eh, from INSTALL: "PREFIX doesn't actually work right now" 21:24
any hackaround to make it work? installing to /lib /bin and /libexec is not FHS compliant...
TimToady blowing away Test.pm.yml doesn't help
TreyHarris cj: yay, i was afraid i had somehow screwed up (not like there's much to screw up in handing out a commitbit)
cj TreyHarris: looks like it was just my ignorance :) 21:25
TreyHarris what should i have told you that would have cleared it up, so i'll tell the next person? :)
cj audreyt: what did you do to add me to the project? Could I have added myself? 21:26
21:26 multic left
audreyt cj: no, requires prjadmin 21:27
commitbit.pugscode.org/
lambdabot Title: Pugs -- OSSF
cj TreyHarris: have an admin flip the bit for the user after they register, methinks
TreyHarris audreyt: i didn't know about that step... i've just been going to the URL above and inviting people, and it seemed to work in the past... what else do i need to do? 21:28
audreyt TimToady: someone broke &?ROUTINE. I suspect the strictness fix
TreyHarris: it did work usually
mabe cj instead of clicking on the link registerred anothe email instead 21:29
cj I went to this one: rt.openfoundry.org/Foundry/Home/Gue...l?id=24500 21:30
lambdabot Title: Member Login -- OSSF, tinyurl.com/ym6k5a
audreyt that should've worked...
TreyHarris audreyt: i was about to be cute and change one of my world clocks from Taipei to Portland. :) but then I realized knowing what time it is in openfoundry land is useful, and I already know what time it is in Portland.... 21:31
audreyt yeah... 21:32
TimToady: fixed 21:33
it's indeed the strict fix
svnbot6 r14438 | audreyt++ | * isStrict: $?STRICT shouldn't be a global; rename it to $*STRICT.
21:34 buubot joined
TimToady that's a funny way to say it, given that $*STRICT is global... 21:35
21:37 weinig is now known as weinig|food
audreyt what seano did was setting up so all $?FOO become global... 21:37
TimToady seems like compile-time's $*STRICT should get mapped to lexically scoped $?STRICT. 21:38
audreyt yes.
or rather, push to the envPragmas structure 21:39
not sure we need to expose all pragmas via $?FOO
maybe just one single %?PRAGMAS
TimToady sorry, unfortunate use of the "map" word...
audreyt you can say "fmap" which is safe in almost all circumstances...
TimToady :D
on the other hand, $? is sort of there for exposing such things in the first place. 21:41
audreyt multiple strict.pm with diff long names may all want $?STRICT though 21:42
21:42 ofer0 joined
TimToady strictness is not really a unitary concept in any case 21:43
audreyt right. 21:44
21:44 dduncan joined
TimToady though P6 really only does strict 'vars'... 21:45
having legislated the others out of existence.
21:45 rlb3 joined
TimToady okay, seems to be smokin' smoothly now. 21:47
audreyt woot
TimToady audreyt: under what circumstances did you get rid of "expecting" errors?
still does it for $foo $bar
audreyt expecting operator, statement modifier, ";" or end of input 21:48
I consider that useful
well, "end of input" is not
but the other 3 is
TimToady I think mere mortals would rather see: "You can't put two terms in a row without an operator between them."
audreyt so just "expecting operator ? 21:49
we can certainly special case each error state into a diagnostics
21:49 Limbic_Region joined
TimToady yeah, that would be an improvement 21:49
; and statment modifiers are just very loose operators... 21:50
statement even
21:53 nezcaz joined
nezcaz hello * 21:54
Limbic_Region salutations !
TimToady howdy do
21:55 djuls joined, djuls left
PerlJam "pregnant term: expecting operator" :-) 21:56
21:56 ruoso joined
nothingmuch audreyt: seen something like this before? rt.cpan.org/Ticket/Display.html?id=22289 21:57
jabbot nothingmuch: I havn't seen something, nothingmuch
lambdabot Title: #22289: Segfault after install, tinyurl.com/y2zlf7
nezcaz ah, need coffee i'am bloody tired 21:58
audreyt nothingmuch: no not at all
nothingmuch very mysterious =/
audreyt TimToady: done 21:59
audreyt goes back slideshacking
svnbot6 r14439 | audreyt++ | * Parse message fix: 22:00
r14439 | audreyt++ | - sigilled vars are taken as an unit in "unexpected" errors
r14439 | audreyt++ | - semicolon, eof and statement modifiers are not mentioned anymore
r14439 | audreyt++ | pugs> $x $y
r14439 | audreyt++ | Unexpected "$y"
r14439 | audreyt++ | expecting operator
r14439 | audreyt++ | at <interactive> line 1, column 4
nothingmuch audreyt: how does Module::AutoInstall get into it? the YAML::Syck makefile.pl doesn't use auto install 22:06
Juerd ?eval $x 3 22:07
22:07 evalbot_r14413 is now known as evalbot_r14438
evalbot_r14438 Error: ␤Unexpected "3"␤expecting operator, statement modifier, ";" or end of input 22:07
Juerd ?eval $x say blah
evalbot_r14438 Error: ␤Unexpected "say"␤expecting operator, statement modifier, ";" or end of input
Juerd Oh, it lags behind one revision :)
Limbic_Region not really 22:08
22:08 norageek is now known as norageek2
Limbic_Region it probably isn't finished compiling 22:08
Juerd Which makes it lag behind one revision...
I'm not saying it *always* lags behind. Just that it lags behind, implying now. 22:09
Limbic_Region well, given the rate of commits - always probably isn't too far from the truth ;-)
Juerd :)
TimToady well, the basic problem is that you should never use the English present tense for present tense. 22:10
we never say "I run" to mean "I am running"
nothingmuch urf
TimToady "I run" always means "I run generally" 22:11
Juerd I've always known that English is weird in many ways. But this I even find hard to believe :)
Limbic_Region as in "I run to work every morning"
Juerd Sorry....
I know that English is weird in many ways. But this I am finding hard to believe.
TimToady that's okay, you can find it hard to believe in general. :)
Juerd I don't find it hard to believe in general. Or, well, I'm not sure yet, as I'm already finding it a bit easier to believe :) 22:12
TimToady I'm sure there are also exceptions...
Juerd How on earth can a language get this weird? 22:13
Where you can't use present tense for present tense, or simple past for past tense...
TimToady The English look at the Americans, and the Americans look back at the English... 22:14
nezcaz latin :p
Juerd Latin made more sense in this respect.
Limbic_Region TimToady - have you read (m)?any of Steven Pinker's books?
TimToady nope. I think my wife has though.
but then she reads pretty much everything. 22:15
Juerd should perhaps unsubscribe from a few mailinglists
And just read the weekly summaries
TimToady: Then she must really love the Gutenberg project :) 22:16
Limbic_Region hasn't read much non-fiction lately which isn't the norm
Juerd doesn't read much at all.
I'm too impatient
TimToady I think she probably read most of the Gutenberg project while it was still on paper.
Juerd TimToady: Heh
Limbic_Region wow 22:17
Juerd Of technical books I often only get halfway through
s/Of t/T/
And I don't even begin reading fiction
Juerd loves Perl Hacks 22:18
I can read it in random order and it still makes sense
nezcaz use Math::TrulyRandom to be sure that u are random enough :p 22:22
22:23 ofer0 joined 22:24 mako132_ joined
nezcaz s/are/ 22:24
TimToady most of my failures appear to be a bare block at statement level being taken as a hash because the code inside was all commented out. I don't think statement block should turn into hash ever. 22:32
@tell audreyt most of my smoke failures appear to be a bare block at statement level being taken as a hash because the code inside was all commented out. I don't think statement block should turn into hash ever. 22:33
lambdabot Consider it noted.
nezcaz lol
22:38 thepler joined
svnbot6 r14440 | lwall++ | hash curlies can't end a statement 22:40
r14441 | audreyt++ | * <TimToady> I don't think statement block should turn into hash ever.
22:40 seano joined 22:46 glasser joined 22:51 explorer joined, elmex_ joined
svnbot6 r14442 | lwall++ | .perl/eval roundtripping needs to isolate from statement boundary. 22:54
TimToady also possible that .perl shouldn't put out something that in isolation would be misinterpreted at a statement boundary... 22:56
23:00 elmex joined 23:05 weinig|food is now known as weinig 23:15 elmex joined 23:25 polettix joined, SubStack joined 23:33 drbean_ joined
GeJ M􏿽xF8􏿽xF8se lambdacamels 23:40
23:41 dmq joined 23:47 buubot joined 23:56 glasser` joined
svnbot6 r14443 | lwall++ | more bare block cleanup 23:58
r14443 | lwall++ | s{...}{...} cleanup

Powered by Cro and the Raku® Programming Language.

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

Thank you!