Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2006-10-15

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 Limbic_Region let me confirm I can make the problem disapear by changing the .t file
00:01 Limbic_Region hrm - that's odd
00:01 Limbic_Region it is working as expected at home (but still failing)
00:01 Limbic_Region though at work it is definately fubar
00:01 Limbic_Region and also failing
00:02 Limbic_Region TimToady - I can certainly check both $actual and $expected - give me a sec
00:02 TimToady it seems to be set to native.  Not sure why notepad should screw it up unless your svn client is screwing up the native property
00:03 mako132_ joined perl6
00:04 TimToady but still, inside perl everything should canonicalize to \n if text mode is set on file inputs.
00:05 TimToady so my guess is still that slurp doesn't, since there probably aren't \r in the .t
00:06 TimToady (or notepad would be happier with it)
00:06 Limbic_Region say $new_config_text ~~ /\r/ ?? 'Y' !! 'N'; # says Y
00:06 Limbic_Region which is the first option to is()
00:06 Limbic_Region s/option/argument/
00:07 Southen_ joined perl6
00:07 TimToady so yes, slurp is defective.
00:07 TimToady maybe I can figure that out myself.
00:08 Limbic_Region will be standing by to test
00:11 TimToady There's no place like Haskell...There's no place like Haskell...
00:12 * Limbic_Region reminds TimToady not to forget the Ruby er Pugs slippers
00:21 Southen joined perl6
00:24 TimToady well, supposedly Haskell's builtin readFile is gonna default to text mode, but it seems like that isn't happening here...
00:24 TimToady I think we need an audreyt for this.
00:25 Limbic_Region smoke is finished btw
00:25 Limbic_Region few minutes faster still
00:25 TimToady you're just as pure as Ivory Soap!
00:26 Limbic_Region yep, 99 failures and I would say as many as not are due to the eqv thingy
00:26 Limbic_Region ?seen audreyt
00:26 lambdabot audreyt is in #perl6, #ghc and #haskell. I last heard audreyt speak 8h 7m 25s ago.
00:26 Limbic_Region well, depending on how long she stayed up after she stopped talking
00:27 Limbic_Region she might be up soon
00:32 TimToady this is all orthogonal to the notion that eqv should also be canonicalizing, of course.  but if we fix the slurp textmode maybe we can put that off...
00:33 TimToady actually, I believe audreyt is on record that the correct way to canonicalize things like eqv is to canonicalize everything coming into perl the same, so I guess it comes out to non-orthogonal under that view.
00:34 TimToady on the other hand I'm not sure Unicode cares about crlf vs lf canonicalization...
00:35 TimToady then there's the whole debate over choosing C or D...
00:35 Limbic_Region well, I have been living under a rock for the last few decades and I can't speel unicode
00:36 Limbic_Region I am interested in being able to treat text with ambivelance towards functionally equivalent control characters most of the time
00:36 Limbic_Region but when I want to look closer, I want to be able to do that too
00:45 explorer joined perl6
00:46 TimToady dinner &
01:05 nothingmuch INGY!!!!!!
01:05 nothingmuch Module::Compile breaks with the << operator
01:06 nothingmuch $bit <<= 1;
01:06 nothingmuch $bit <<= 1;
01:06 nothingmuch 86e1fb5a98b9c2c2c5314ff534059c87e035aa6e
01:06 nothingmuch VERY NAUGHTY!
01:06 nothingmuch why are you even parsing heredocs?
01:20 mauke joined perl6
02:03 svnbot6 joined perl6
02:05 evalbot_r14340 joined perl6
02:21 svnbot6 r14341 | audreyt++ | * Pugs.Prim: Instead of using ByteString.Char8's file IO primitives
02:21 svnbot6 r14341 | audreyt++ |   (which defaults to binary mode instead of text mode on win32),
02:21 svnbot6 r14341 | audreyt++ |   use Haskell's built-in ones, which handles CRLF transparently.
02:21 svnbot6 r14342 | audreyt++ | * Charnames.hs: Trim unused import.
02:31 audreyt resmoking on win32, expecting all green...
02:31 audreyt if so, then repair js is next for me, then changelog
02:31 Southen_ joined perl6
02:52 Southen joined perl6
02:58 svnbot6 r14343 | audreyt++ | * build_pugs.pl: Force relinking of Main.hs to ensure HsJudy
02:58 svnbot6 r14343 | audreyt++ |   wasn't accidentally picked up.
03:17 Southen_ joined perl6
03:32 TimToady r14342 smokes fine on Linux.
03:45 reZo joined perl6
03:49 audreyt win32 is fine except for all_parse.t
03:49 audreyt fixing
03:49 audreyt (CRLF is fixed)
03:49 audreyt (brb)
03:49 TimToady yay, I thought readFile was being the problem...
03:50 mako132_ joined perl6
04:10 gugod joined perl6
04:19 nothingmuch http://nothingmuch.woobling.org/Crypt-Util/
04:19 lambdabot Title: Index of /Crypt-Util
04:20 nothingmuch feedback welcome
04:20 nothingmuch xMAC is TODO
04:20 nothingmuch salted hashes might be added too
04:51 Salzh joined perl6
05:01 Salzh ?say "hello everyone"
05:01 lambdabot Maybe you meant: faq map slap
05:10 Southen joined perl6
05:12 mako132_ joined perl6
05:12 svnbot6 r14344 | audreyt++ | * be a bit more relaxed about 'zero time' in t\unspecced\times.t
05:15 BooK joined perl6
05:16 Salzh ?eval say "hello everyone"
05:16 evalbot_r14340 is now known as evalbot_r14344
05:16 evalbot_r14344 OUTPUT[hello everyone␤] Bool::True
05:21 Southen_ joined perl6
05:27 svnbot6 r14345 | audreyt++ | * Makefile.PL: Re-enable -threaded runtime for Win32.
05:33 svnbot6 r14346 | audreyt++ | * Make system() nonfatal always; instead it merely sets $!.
05:33 svnbot6 r14347 | audreyt++ | * system.t: unTODO.
05:36 svnbot6 r14348 | audreyt++ | * chmod.t: Fix skip on win32.
05:45 Salzh ?eval @a = (1, 2, 3, 4); @b = (5, 6, 7); @a>>+=<<@b; say join " ", @a;
05:45 evalbot_r14344 is now known as evalbot_r14348
05:45 evalbot_r14348 Error: ␤Unexpected ">>+=<<@"␤expecting "::", term postfix, comment, operator, statement modifier, ";" or end of input
05:48 svnbot6 r14349 | audreyt++ | * Pugs.Prim: `foo.exe` - make it nonfatal.
05:48 svnbot6 r14350 | audreyt++ | * all_parse.t: There's no need to turn \ into / -- it was triggering a modifying-constant error on Win32.
05:48 audreyt win32 shiuld smoke clean now.
05:49 audreyt resmoking...
05:50 Southen joined perl6
05:50 TimToady me 2
06:00 gaal hey! sorry for not helping out with change logging this weekend, got sidetracked.
06:00 gaal we need something to make that task easier! :)
06:18 Southen_ joined perl6
06:19 nothingmuch_ joined perl6
06:21 iblechbot joined perl6
06:38 TimToady r14350 still mighty fine on Linux.
06:48 Southen joined perl6
06:54 lanny joined perl6
06:55 lanny ?eval say "newline"
06:55 evalbot_r14348 is now known as evalbot_r14350
06:55 evalbot_r14350 OUTPUT[newline␤] Bool::True
06:56 audreyt gaal: we do, somethiing like nothingmuch_'s summarizer
06:56 audreyt gaal: so are you still up to chglog in the next, uhm, 48 hours?
06:56 audreyt win32 smoke all clean.
07:08 penk joined perl6
07:08 Southen_ joined perl6
07:11 penk left perl6
07:16 svnbot6 r14351 | audreyt++ | * Pugs.Embed.Perl5: Correctly find out the installed version
07:16 svnbot6 r14351 | audreyt++ |   of Pugs::Compiler::Rules after a "make install".
07:19 Southen joined perl6
07:21 svnbot6 r14352 | audreyt++ | * PIL2JS: Fix various outdated constructs.
07:21 mugwump joined perl6
07:30 svnbot6 r14353 | audreyt++ | * Pugs.Compile: Repair PIL1's handling of loops.
07:58 iblechbot joined perl6
08:01 foo\ joined perl6
08:05 gaal audreyt: unfortunately not :-(
08:06 svnbot6 r14354 | audreyt++ | * Pugs.Types: &Main::infix:<Moose> is canonical, _not_
08:06 svnbot6 r14354 | audreyt++ |   &infix:Main::<Moose>, which was confusing PIL2JS.
08:06 audreyt ok... I'll do it myself then :)
08:06 audreyt PIL2JS repaired!
08:06 gaal sorry :(
08:06 audreyt no prob at all
08:06 audreyt actually... any other volunteers?
08:06 gaal what can a summarizer help with?
08:07 audreyt group commit logs by dir
08:07 svnbot6 r14355 | audreyt++ | * Pugs.Compile: Stub out the "subst" and "trans" nodes.
08:07 audreyt and/or by file
08:07 gaal svk log -v in one window, $EDITOR in the other, and occasionally svk desc
08:07 gaal that's what I do
08:07 audreyt metoo
08:07 gaal hmm, that's a good idea
08:08 xinming hmm, I'll be volunteer If I can help. :-D
08:08 gaal i think the real solution si to pay the price at insert time, and ask everyone to edit changelog when they commit :)
08:08 audreyt or to use some sort of annotation that can be picked up
08:08 audreyt xinming: you can!
08:08 xinming audreyt: Ok, what do I need to do?
08:08 gaal we tried using prefix '* ' but that didn't really pick up, did it?
08:08 gaal xinming++
08:08 xinming summarize the mailing list? or summarize the changelog.
08:09 audreyt the changelog
08:09 audreyt r13262 ~ r14355
08:09 xinming Now, I have 3 hours to do...
08:09 xinming Ok, I'll try...
08:09 xinming *-*
08:09 xinming *_*
08:09 svnbot6 r14356 | audreyt++ | * Repair PIL2JS.  Largely it was because of the ::main ==> ::Main change.
08:09 audreyt just give it a cursory glance and put seemingly-relevant parts into the ChangeLog file under the appropriate secion
08:09 audreyt I'll do a full review after your 3 hours :)
08:09 xinming Ok, I will
08:10 xinming well, I think you have to review it tomorrow. :-P
08:10 xinming Now, I have to go out for something to eat.
08:10 audreyt ok. thanks!
08:10 xinming But I'll try my best to summarize the changelog
08:10 audreyt xinming++
08:11 audreyt meanwhile I'll do some dist preparation and post preflight #0.
08:12 gaal w00t
08:13 gaal I'm at $work ant can't moose that either :-(
08:13 * gaal doesn't feel very useful
08:13 audreyt you're useful on the #jifty front...
08:19 nothingmuch joined perl6
08:20 gaal eat and save the world at the same time
08:27 audreyt GHC 6.6 powerpc build: http://scsibug.com/ghc-6.6-ppc-apple-darwin.tar.gz
08:27 lambdabot http://tinyurl.com/yhn92c
08:27 audreyt (OSX)
08:44 gaal is a debian build planned?
08:45 gaal igloo mentioned that the 'ghc-cvs' packages were tricky to fix (they'd been broken for a yaer)
08:45 gaal *year. wow, latency kills spelling
08:46 audreyt well the bindist works
08:50 xinming hmm, I'll just start working. But I found, If you use 'svk log -r 13262:14355', It would get the revision locally, Do I need to calculate the revision number or, there is a convient way of doing this
08:51 audreyt svk log -r 13262@:14355@
08:51 svnbot6 r14357 | audreyt++ | * Makefile.PL: Don't include Pugs-Emitter-Perl6-Perl5-2 in
08:51 svnbot6 r14357 | audreyt++ |   the perl5/ builds.
08:51 gaal audreyt: yeah
08:54 svnbot6 r14358 | audreyt++ | * Pugs.hs: Make "pugs -BPerl5" invoke the correct helper in both uninstalled
08:54 svnbot6 r14358 | audreyt++ |   and installed positions.
09:12 svnbot6 r14359 | audreyt++ | * Bump version to 6.2.13.
09:15 svnbot6 r14360 | audreyt++ | * Add link to GHC 6.6.
09:19 xinming hmmm.
09:19 xinming * Don't parse :a as ::a.
09:19 xinming (orig r13326):  audreyt
09:19 xinming do I need to record these in to bug fixes section?
09:19 audreyt seems not
09:20 audreyt mostly it's during-development fix
09:20 audreyt i.e. ot actual fix compared to last relesae
09:20 audreyt if there's little or no context, don't bother recording that :)
09:20 xinming Ok.
09:21 nothingmuch joined perl6
09:21 audreyt changes topic to: Please test preflight! http://pugs.blogs.com/pre0/Perl6-Pugs-6.2.13.tar.gz | spec.pugscode.org | paste: http://sial.org/pbot/perl6 | pugs.blogs.com
09:21 audreyt Pre0 up: http://pugs.blogs.com/pre0/Perl6-Pugs-6.2.13.tar.gz
09:21 lambdabot http://tinyurl.com/y8ef5l
09:22 nothingmuch awwe, someone took out the nice paradigm slogan/motto thing from /topic
09:31 svnbot6 r14361 | audreyt++ | * Sort MANIFEST.SKIP.
09:33 xinming Pugs::Runtime::Regex - Deep recursion is made fatal
09:33 xinming Is this a bug for a feature?
09:33 audreyt neither; it does not even exist in the prev release
09:33 audreyt so it's skippable :)
09:33 xinming thanks
09:44 nothingmuch @seen gugod
09:44 lambdabot I saw gugod leaving #perl6 1h 36m 44s ago, and .
09:44 nothingmuch @tell gugod Crypt::OFB is a stream cipher mode, why does it need padding?
09:44 lambdabot Consider it noted.
09:44 Southen_ joined perl6
09:44 audreyt hmm pre0 is missing Pugs:compiler::Rules, oops.
09:52 audreyt changes topic to: Tester wanted: http://pugs.blogs.com/pre1/Perl6-Pugs-6.2.13.tar.gz | spec.pugscode.org | paste: http://sial.org/pbot/perl6 | pugs.blogs.com
09:52 audreyt pre1 up.
09:58 Alias_ joined perl6
09:58 Alias_ seen jhi?
09:58 jabbot Alias_: I havn't seen jhi, Alias_
10:03 nothingmuch audreyt: thanks for ghc 6.6
10:03 nothingmuch the link that is
10:03 audreyt np :)
10:03 * nothingmuch goes to get tangled in red tape
10:04 svnbot6 r14362 | audreyt++ | * Bumb version in Pugs.cabal.in.
10:04 audreyt Bump, even.
10:04 Southen joined perl6
10:06 nothingmuch ack
10:06 nothingmuch that build seems broken
10:06 nothingmuch i can't compile mo
10:06 nothingmuch moh
10:06 pasteling "nothingmuch" at 84.108.53.74 pasted "linkage errors" (6 lines, 120B) at http://sial.org/pbot/20444
10:08 audreyt hm are you on 10.3?
10:09 audreyt also try
10:09 audreyt sudo gcc_select 4.0
10:09 audreyt also try
10:09 audreyt sudo gcc_select 3.3
10:11 SubStack_ joined perl6
10:12 SubStack_ is now known as SubStack
10:17 Southen_ joined perl6
10:21 SubStack_ joined perl6
10:21 SubStack_ is now known as SubStack
10:26 buetow joined perl6
10:38 nothingmuch audreyt: ah, i was on 3.3, 4.0 fixed it
10:38 nothingmuch thanks
10:38 nothingmuch audreyt++
10:46 Southen joined perl6
10:47 audreyt np :)
10:47 audreyt while you're at it run a smoke on pre1?
10:47 audreyt we havn't had any GHC 6.6/OSX/PowerPC reports
10:50 nothingmuch okies
10:51 nothingmuch haven't done pugs yet since the disk crash
10:51 nothingmuch so it'll take a while
10:58 audreyt sure thing
10:59 dduncan joined perl6
11:01 audreyt dduncan: greetings. have you seen http://scsibug.com/ghc-6.6-ppc-apple-darwin.tar.gz ?
11:02 lambdabot http://tinyurl.com/yhn92c
11:02 dduncan I have not
11:02 dduncan I looked at the haskell site today, and it only showed an intel compiled version for Mac oS X
11:03 audreyt nod. it's a new build (as of today). will be on the haskell site soon
11:03 dduncan that said, I typically use the Mac OS X packages of GHC, for simplicity ... though if one isn't forthcoming soon, I'll try using the one you're pointing to
11:03 audreyt the installation for that is simple "sh configure; make ; sudo make install"
11:03 dduncan okay, I'll try that then, and then will try a clean Pugs build with it
11:03 audreyt try the pre1 posted on/tpopic
11:04 audreyt http://pugs.blogs.com/pre1/Perl6-Pugs-6.2.13.tar.gz
11:04 lambdabot http://tinyurl.com/yb3a52
11:04 audreyt it smokes clean on osx-intel and win32 here
11:04 dduncan sounds good
11:04 dduncan fyi, its already 4am here, so I won't be doing those tests until tomorrow
11:04 audreyt sure that's fine
11:05 audreyt a Tuesday release is good enough
11:05 dduncan I have a different question for the experts here, which concerns my QDRDBMS implementation in Perl 5 ...
11:06 dduncan as part of that implementation, I have made some strict data types for Perl 5 that are named after the Perl 6 ones ...
11:06 dduncan strict mainly meaning strongly typed
11:06 dduncan but they also use singletons, so the same value is only ever stored once, and so comparing refaddrs of 2 Value objects suffices to test that they are equal as well
11:07 dduncan construction is more expensive for simpler types, but cheaper for more complex types, and for all compares
11:07 dduncan now, the question ...
11:08 dduncan I have distinct Str and Blob types, which are basically wrappers around a Perl 5 string in character mode vs one in byte mode
11:09 dduncan now, when creating a variable from a native Perl string, I test the is_utf8 flag on the variable ... it must always be off in the Blob case, and sometimes has to be on in the Str case ...
11:09 dduncan to make things easier for users, I only want to complain about input for a Str if the flag is off and the bytes include 8-bit values
11:10 dduncan so byte mode with just ascii chars, or higher-chars in char mode are fine, ...
11:10 dduncan I would like to know the most performance-efficient way to test the 8th bit
11:10 dduncan ...
11:11 dduncan I thought a regexp may be more expensive, and I didn't see an Encode function
11:11 dduncan the regexp I came up with was: m/^ [0x00-0x7F]* $/x
11:11 dduncan if that's right, it should return true for ascii, and false for higher-bit values
11:11 audreyt you want /[^0x00-0x7F]/
11:11 audreyt and then negate it.
11:12 Alias_ So it shortcuts
11:12 Alias_ Although it's an error case, so in practice it shouldn't
11:12 dduncan note that I'm only doing that test if the flag is off
11:12 nothingmuch audreyt: trunk or 6.2.13?
11:12 dduncan so the one I did won't shortcut?
11:13 audreyt nothingmuch: should be the same thing but 6.2.13 is preferred
11:13 Alias_ It should to a degree, but it will do much more work to do so
11:13 dduncan okay, so I'll take Audrey's suggestion as a faster way
11:13 Alias_ Plus for the other one the range is shorter, which probably helps too
11:13 Alias_ yes
11:19 dduncan fyi, here is the current version of the Value code: http://darrenduncan.net/Value-20061014a.pm
11:20 dduncan also, if you change the 'Value' to 'Util', you will see that file's one dependency ... which currently just checks arguments
11:20 dduncan that is, http://darrenduncan.net/Util-20061014a.pm
11:21 dduncan it is very possible that the code contains a bunch of overkill that may slow it down a lot, or it may not, I have yet to know
11:21 dduncan a lot of it wouldn't be needed in a Perl 6 version, though
11:23 foo\bar joined perl6
11:24 Alias_ ok, best as I can tell, I need a cross-language serialization for an unknown infinite directed graph
11:24 dduncan would YAML work for this?
11:24 Alias_ That's what I'm thinking
11:25 dduncan YAML is what I plan to use for the ::Example Engine of my database
11:25 dduncan how its data is saved to disk that is
11:26 Alias_ Where are you up to in that regards
11:26 dduncan I hope to have it working in a few weeks
11:26 Alias_ You finished Rosetta, and now you are implementing a backend for it?
11:27 dduncan actually, Rosetta is stalled indefinitely, while the QDRDBMS project is meant to take its place in the short term
11:27 Alias_ stalled indefinitely?
11:28 Alias_ Why?
11:28 dduncan among other things, QDRDBMS is meant to be the non-over-engineered version that I use to play with ideas intended for Rosetta, and see what works out in practice
11:28 dduncan Rosetta is now pure Perl 6, and I want the language to develop a bit more before I resume it
11:29 dduncan mainly, the Perl 5 QDRDBMS is based on an unchanging foundation, and makes it easier to try out certain things, where I'm still trying to figure out how they should be working
11:30 elmex joined perl6
11:30 Alias_ ok
11:30 dduncan its hard to ask for help to get things working in Perl 6 when I can't effectively communicate what I actually want
11:30 Alias_ :)
11:30 dduncan the Perl 5 impl should help with that communication, having something more concrete to propose
11:31 dduncan also, I want to use some of this stuff in a $job right now, and Perl 5 is the best for that at the moment
11:31 xinming hmm, Changelog has been updated, Hope what I do is right. :-P
11:34 svnbot6 r14363 | yiyihu++ |  r14504@HomePc:  xinming | 2006-10-15 19:29:50 +0000
11:34 svnbot6 r14363 | yiyihu++ |  Change log update.
11:34 svnbot6 r14363 | yiyihu++ |  From r13262 to r14355. What I do mostly is just record changelog on features
11:34 svnbot6 r14363 | yiyihu++ |  and bugfix. I hope Changelog for Module and pX can be found in repository. But
11:34 svnbot6 r14363 | yiyihu++ |  no time for now, sorry.
11:42 buetow joined perl6
11:43 kanru joined perl6
11:44 dduncan incidentally, the Value module is intended for use in 2 main places; one is in the API that people use to put data in and get data out of a database, and the other is in the implementation of ::Example
11:44 dduncan it is mainly for the former that all the input checking is there
11:44 dduncan for the latter, most input checking isn't needed, as the only calling code is my own
11:45 dduncan in light of this, I may perhaps split that Value file in 2, one for the API, and one for the backend, as applicable, or maybe not
11:46 dduncan I'll just keep it together unless performance is a concern
11:59 Southen_ joined perl6
12:19 Psyche^ joined perl6
12:30 dduncan left perl6
12:31 Southen joined perl6
12:35 Psyche^ is now known as Patterner
12:51 chris2 joined perl6
12:51 Southen_ joined perl6
12:53 camenix joined perl6
12:55 DuskoDugousko joined perl6
12:56 wolverian DuskoDugousko, welcome! don't be afraid to ask things that are unclear to you. have fun :)
12:56 audreyt greetings DuskoDugousko :)
12:56 DuskoDugousko hi everyone
12:56 DuskoDugousko thanks ;)
12:57 wolverian audreyt, does the pre test and compile in 300mb HD space?
12:57 wolverian I know it sounds weird, but my laptop is pretty restricted right now ... :/
12:57 audreyt 358464k
12:58 audreyt er no
12:58 audreyt 278512k
12:58 audreyt so yes.
12:58 wolverian oh, that's fine :) thanks
12:59 audreyt np :)
12:59 wolverian I used roles for the first time ever tonight.. granted that was moose, but still pretty nice
13:00 wolverian I think I saw a light about the reuse bit there
13:00 wolverian they're components, so make them first class! instead of, you know, classes. cough.
13:01 DuskoDugousko is now known as Dugousko
13:01 audreyt well Pugs as of this release also supports simple roles...
13:01 wolverian oh.. I thought that was in the next next release
13:01 wolverian audreyt++
13:01 audreyt the MOP that allows you to introspect roles are
13:01 audreyt but "does" and "role Foo {}" is alraedy in
13:01 wolverian ah
13:01 audreyt so have fun with them :)
13:02 wolverian yay
13:02 wolverian can I pattern match on them?
13:04 audreyt ?eval role Foo {} class Bar does Foo {}; Bar.new ~~ Foo
13:04 evalbot_r14350 is now known as evalbot_r14363
13:04 evalbot_r14363 Bool::True
13:04 audreyt yes.
13:04 wolverian should have tried that. thanks!
13:04 wolverian ok, my ghc confused it
13:04 audreyt np :)
13:04 wolverian Could not find module `Distribution.Simple':
13:04 wolverian I've seen that before - my setup is probably borked
13:04 audreyt you didn't have Cabal?
13:05 wolverian I do, but I think I have remains of a ghc 6.5 somewhere
13:05 wolverian Cabal-1.1.4
13:05 wolverian ghc 6.4.1
13:06 audreyt that should be fine
13:07 wolverian yeah, I'm sure it's my setup that's borked. don't mind it
13:12 ndm joined perl6
13:22 nothingmuch joined perl6
13:36 sosko_ joined perl6
13:46 audreyt hm. 9 weeks. no blogs from me. oy.
13:46 * audreyt goes remedying that
13:46 wolverian audreyt++ :)
14:13 ludan joined perl6
14:18 diakopter joined perl6
14:18 audreyt journal up.
14:21 audreyt sprinkling it with hrefs...
14:23 justatheory joined perl6
14:37 Odin-LAP joined perl6
15:05 justatheory joined perl6
15:19 diakopter left perl6
15:20 xinming audreyt: please check my "homework" >_<
15:20 xinming :-P
15:20 xinming I don't know if the changelog I made is right...
15:21 ndm left perl6
15:21 xinming But don't have time really. sorry for that I didn't log the change on modules pX
15:21 xinming s/pX/and pX/
15:31 thepler_ joined perl6
15:37 araujo joined perl6
15:40 kanru joined perl6
15:42 mauke joined perl6
16:57 penk joined perl6
17:16 larsen joined perl6
17:27 b_jonas-2 joined perl6
17:27 b_jonas-2 left perl6
17:27 b_jonas joined perl6
17:29 b_jonas the numeric switch to regexps will be useful
17:30 b_jonas you can just write m:4/<[^|]>/ instead of (split "|")[4]
17:30 b_jonas or maybe m:5, I don't know if it's indexed from one or zero
17:32 TimToady bare number isn't supposed to be allowed.  Should be m:5th to get .[4]
17:33 b_jonas I see. :5th seems a very strange syntax though
17:33 TimToady or :5x to get the first five
17:33 b_jonas why is :5 not allowed? is that because of the other role of the colon?
17:33 TimToady it's syntactic sugar for :nth(5) and :x(5)
17:34 TimToady because it's ambig whether you mean nth or x
17:34 TimToady plus bare :16<...> is radix notation
17:34 b_jonas is it?
17:34 TimToady not that radix is meaningful on a regex
17:34 b_jonas so bare numeric switches are not allowed at all, not only for regexps?
17:35 TimToady ?eval chr :16<263a>
17:35 evalbot_r14363 "☺"
17:35 TimToady that's correct.
17:35 b_jonas then n:5th/<[^|]>/ isn't any shorter than (split"|")[4] though I still like it better
17:35 TimToady it gets the intent out front.
17:36 b_jonas oh, and a * is missing
17:36 TimToady particularly for s:2nd/foo/bar/
17:36 b_jonas m:5th/<[^|]>+/
17:36 b_jonas m:5th/<[^|]>*/ I mean
17:37 TimToady ^| is not correct if you're trying to negate
17:37 TimToady <-[|]>
17:38 b_jonas oh
17:38 b_jonas that's paralel to the <[a..z]> syntax which also has changed
17:38 TimToady yes.
17:38 TimToady gotta run &
17:39 b_jonas bye then
17:45 b_jonas left perl6
17:47 reZo joined perl6
18:18 Limbic_Region joined perl6
18:18 Limbic_Region salutations all
18:20 mako132_ joined perl6
18:23 Patterner "17954 test cases: 17954 ok, 0 failed" :)
18:23 Patterner Limbic_Region: moin moin
18:27 MacVince joined perl6
18:28 Limbic_Region I am testing from the repo - hope that is ok
18:33 Limbic_Region beginning smoke
18:34 mako132_ joined perl6
18:36 Aankhen`` joined perl6
18:38 * Patterner pours water over Limbic_Region
18:41 * Limbic_Region needed that to wake up
18:41 Limbic_Region thanks
18:44 zdeqb joined perl6
18:48 zdeqb left perl6
18:51 cjeris joined perl6
18:55 explorer joined perl6
18:56 ludan joined perl6
19:03 lollan joined perl6
19:06 araujo joined perl6
19:08 Limbic_Region smoke finished
19:11 Limbic_Region 1 failure
19:11 Limbic_Region 99.99% ok
19:11 Limbic_Region looking at what failed
19:14 Limbic_Region false alarm
19:14 Limbic_Region my local copy was b0rked
19:14 Limbic_Region @tell audreyt r14363 smokes 100% ok for me on Win32
19:14 lambdabot Consider it noted.
19:19 larsen joined perl6
19:36 penk joined perl6
19:37 penk joined perl6
19:39 tewk running debian sid on x86: Setup: cannot satisfy dependency pugs-fps-any
19:43 kcwu joined perl6
19:44 amv joined perl6
19:50 penk joined perl6
19:52 Limbic_Region tewk - are you running one of the binary builds?
19:55 tewk of ghc? no I'm using http://haskell-unsafe.alioth.debian.org/haskell-unsafe.html
19:55 lambdabot Title: Haskell Unsafe, http://tinyurl.com/y4lmhv
19:55 Limbic_Region well, that's probably the reason
19:55 Limbic_Region they removed some of the packages from the base
19:55 Limbic_Region not too hard to add though
19:55 Limbic_Region give me a second
19:57 Limbic_Region tewk - you may need to check with #haskell for help
19:57 Limbic_Region but
19:57 Limbic_Region http://paste.lisp.org/display/25124 is a related but not exactly the same problem
19:57 Limbic_Region I provided steps to use Cabal to install extra packages
19:57 tewk Thanks
19:57 Limbic_Region not sure about fps btw
19:57 Limbic_Region need to go
19:57 Limbic_Region good luck
20:01 penk joined perl6
20:14 penk joined perl6
20:18 penk joined perl6
20:19 mako132_ joined perl6
20:29 lollan joined perl6
20:29 lollan joined perl6
20:32 fridim__ joined perl6
20:40 markstos joined perl6
20:53 diakopter joined perl6
21:20 diakopter left perl6
21:21 reZo joined perl6
21:47 fridim__ is now known as fridim
21:49 nothingmuch is ther e a better way than $string .= $pad_char x ( ( 4 - length($string) % 4 ) % 4) ; to pad a string to a 4 bytre boundry?
21:49 nothingmuch the double modulu in there is annoying me
21:59 ingy joined perl6
22:00 nothingmuch ingy: ping
22:01 ingy hi nothingmuch
22:01 ingy ttul.org is down :(
22:02 nothingmuch ouch
22:02 nothingmuch M::C is also broken
22:07 ingy huh?
22:07 nothingmuch $foo <<= 4; # parsed as a heredoc
22:08 ingy ok
22:08 * ingy fixes
22:08 nothingmuch *hug*
22:08 * ingy is warmed
22:09 * nothingmuch 'smum knitted him a sweater
22:09 nothingmuch but it's not being used to help ingy's become warm
22:11 ingy I want to add a mode to M::C where the 'use Compile::Foo qw(a b);' is replaced by a compilation result.
22:11 nothingmuch hmm?
22:11 nothingmuch like, foo in expression?
22:12 nothingmuch s/foo/use/; # there was a siren at 5:20 am today, i'm tired
22:12 ingy well normally use Compile::Foo turns on Foo compilation
22:12 ingy but I ust want it to plop a chunk of perl in plae of that one line
22:12 nothingmuch ah
22:13 nothingmuch that would be cool
22:13 ingy so it acts like a function....
22:13 ingy my main use case is to create modules that act as templates
22:13 ingy for certain things
22:14 ingy just put the templates in the DATA section or something
22:14 ingy and qw(a b) can be template data
22:14 ingy or think of them as template function args
22:15 ingy anyway, what do I call the pmc_... method to enable this mode?
22:15 nothingmuch uh... i dunno
22:15 nothingmuch haven't looked into MC internals that much
22:15 ingy pmc_use_means_now?
22:15 ingy XD
22:15 ingy I like that
22:16 nothingmuch =)
22:17 ingy This will make Inline::Module much more trivial
22:19 ingy nothingmuch: do you have a test for <<=
22:19 nothingmuch no
22:19 nothingmuch ENOBRANE right now
22:20 ingy EDRUNK ?
22:20 nothingmuch no, read above
22:20 nothingmuch ESIRENATFIVEINTHEFUCKINGMORNING
22:20 nothingmuch a "we're getting nuked" siren
22:20 ingy are you?
22:20 nothingmuch nope
22:20 nothingmuch it was in error, it seems
22:21 ingy how did you find the <<= error?
22:21 nothingmuch i used <<= ;-)
22:21 ingy od
22:21 ingy n
22:21 * nothingmuch did some lightweight bit twiddling
22:30 dduncan joined perl6
22:41 justatheory joined perl6
22:41 nothingmuch ingy: night =)
22:42 ingy nothingmuch: fixed
22:42 ingy and tested
22:48 dduncan following up yesterday, I'm now downloading the GHC 6.6 binary for Mac OS X PPC, and once that's installed, I will do a clean/first compile of the Pugs preflight and smoke it ... and I'll see how much faster it is too
23:10 dduncan GHC now installed ...
23:19 dduncan hm, I notice that Pugs is building with a different GCC version than my Perl 5 was (4.0.1 vs 3.4); hopefully that won't be a problem for embedding
23:19 ofer0 joined perl6
23:31 explorer joined perl6
23:32 explorer is now known as JoaquinFerrero
23:32 bsb joined perl6
23:54 ofer0 joined perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo