pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
00:03 marcuse joined
marcuse hi there 00:03
is perl6 for real? is there any release date? 00:04
dduncan seen audreyt 00:05
buubot dduncan: Sorry, I haven't seen audreyt
blblack marcuse: christmas 00:07
marcuse and a lovely christmas will be 00:10
00:12 weinig_ joined
buu eval: 15 * 20 00:13
buubot buu: 300
buu eval: 3 * 13000 * .2 00:14
buubot buu: 7800
marcuse can parrot run .net bytecode?
00:16 yahooooo joined
marcuse is anyone here? 00:26
f**k you all 00:28
[particle] marcuse: you're more likely to get answers to parrot questions on irc.perl.org#parrot
00:36 Schwern joined 00:40 yahooooo joined 00:41 weinig joined 00:49 cmarcelo joined
pugs_svnbot r17041 | audreyt++ | trunk: 01:26
: * Pugs.cabal.in: Add the "time" package as dependency to migrate away from the
: archaic System.Time API.
diff: dev.pugscode.org/changeset/17041
lambdabot Title: Changeset 17041 - Pugs - Trac
pasteling "evalbot_r17040" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26282 01:27
meppl good night 01:28
pasteling "evalbot_r17041" at 193.200.132.135 pasted "Pugs build failure" (453 lines, 26.8K) at sial.org/pbot/26283 01:40
01:40 evalbot_r17041 joined 01:47 mako132_ joined 01:50 hkb_zz is now known as hakobe
pugs_svnbot r17042 | audreyt++ | trunk: 01:54
: * Switch from System.Time to the newer Data.Time modules
: for localtime(), time(), the :M filetest, etc.
: (No user-visible changes.)
diff: dev.pugscode.org/changeset/17042
lambdabot Title: Changeset 17042 - Pugs - Trac
01:56 colfax_ joined 02:00 sunnavy joined 02:03 Schwern joined 02:05 jedai joined
pugs_svnbot r17043 | audreyt++ | trunk: 02:05
: * Pugs.Eval: Eliminate all dead code paths discovered by the
: excellent ghc -fhpc (Haskell Program Coverage) too.
diff: dev.pugscode.org/changeset/17043
lambdabot Title: Changeset 17043 - Pugs - Trac
pugs_svnbot r17044 | audreyt++ | trunk: 02:26
: * Thanks to HPC, continue to eliminate a whole lot of
: can't-be-reached toplevel definitions:
: Pugs.Monads:
: tempVar
: Pugs.Parser.Literal:
: qInterpolateDelimiterMinimal
: (and its associated data constructor, QB_Minimal)
: Pugs.Parser.Operator:
: optPreSyn, optSymOps, _STATE_START_RUN
: Pugs.Parser.Types:
: getCurrCharClass
: Pugs.Parser:
: _dummyParam
: Pugs.Prim.Match:
: matchFromMR
: Pugs.Types:
: addNode
r17045 | audreyt++ | trunk:
: * Pugs.Parser.Program: Remove the unused s_freeVars field.
diff: dev.pugscode.org/changeset/17045
lambdabot Title: Changeset 17045 - Pugs - Trac 02:27
pugs_svnbot r17046 | Darren_Duncan++ | trunk: 02:32
: ext/Muldis-DB/ : in Core.pm, fleshed out the list of Text, Blob, Int operators
diff: dev.pugscode.org/changeset/17046
lambdabot Title: Changeset 17046 - Pugs - Trac
02:37 bmedley joined
pugs_svnbot r17047 | audreyt++ | trunk: 02:47
: * Final batch of HPC-inspired toplevel removals:
: Pugs.AST.Pads.diffPads
: Pugs.AST.Utils.getMapIndex
: Pugs.Internals.ID.hashList
: (moved to a local definition in Pugs.Parser.Charnames)
: Pugs.Lexer.balanced
: Pugs.Lexer.ruleWhiteSpaceLine (and ruleEndOfLine)
diff: dev.pugscode.org/changeset/17047
lambdabot Title: Changeset 17047 - Pugs - Trac
02:49 unobe joined
diakopter audreyt: welcome 02:50
audreyt hi :) 02:51
just doing some code cleanups that shouldn't affect smokes... mental space still not at real hacking level 02:52
diakopter audreyt: ah; I see you already have a commit bit!
:P
audreyt how surprising. 02:53
:)
the real value of HPC though is pointing out the codepaths that we _didn't_ test; those would make writing new tests easier
02:53 justatheory joined
audreyt but that requires a bit more Work, so maybe later... 02:53
02:54 bonesss joined
Aankhen`` Greetings, audreyt. 02:54
What is HPC? It sounds like something I need to add to abbr.dat
diakopter Aankhen``: probably Haskell P-something C-something :P
02:55 ewwr joined
audreyt Aankhen``: projects.unsafeperformio.com/hpc/ 02:55
lambdabot Title: Haskell Program Coverage
audreyt Haskell Program Coverage
ewwr when will perl6 be out ? just curious ?
audreyt equiv to Devel::Cover
diakopter ewwr: when you finish it! :)
would you like to help?
ewwr im just curious
when it was expected
audreyt ewwr: even perl 5 is not finished...
Aankhen`` audreyt: Ah, thanks.
audreyt perl 4 is finished, though :)
ewwr ah 02:56
when will they make it a bytecode thing ?
so it has a tad faster speeds etc
diakopter ewwr: here are some links for you
audreyt ewwr: we currently has a "bytecode" in YAML format that does speed up parsing 02:57
it's not as fast as it can be though...
emitting to parrot bytecode (or jvm or other vm's bytecode) is possible but not yet reliable
compiling directly to object code is another possibilty but of course not as portable 02:58
ewwr i see 02:59
audreyt so the current state is the less-than-speedy bytecode in .yml format. (if you compile pugs, blib6/lib/Prelude.pm.yml is the "bytecode" of src/perl6/Prelude.pm, for example.)
ewwr its just awstats takes like quite a bit of time to do its thing
would be cool if perl was done a different way that ddint take so much time
audreyt it might be that just by upgrading to perl 5.9.5, awstat may run a bit faster due to regex optimizations 03:00
but not guaranteed; YMMV
diakopter ewwr: there are plenty of applications or scripts written in Perl that I'm sure would see much improvement via refactoring and optimization while still remaining on perl5.
ewwr: wait, are you an/the author/maintainer of awstats? :) 03:01
ewwr lol i am not optimizign awstats i was looking for a quick heres this and that type of thing to parse something
diakopter wishes for magic, as well. 03:03
audreyt hands diakopter a SVt_PVMG 03:04
diakopter almost looks that up, then remembers where he saw that before. 03:05
03:07 drupek12 joined
pugs_svnbot r17048 | Aankhen++ | trunk: 03:11
: * [irclog] added HPC to abbrs.
diff: dev.pugscode.org/changeset/17048
lambdabot Title: Changeset 17048 - Pugs - Trac
03:12 mako132_ joined
dduncan audreyt, I'm getting an error on make: Could not find module `System.FilePath': 03:12
should I be doing any cleaning or explicit remakefiling?
audreyt dduncan: paste the context? 03:13
no you shouldn't be doing cleaning
dduncan one moment ...
pasteling "dduncan" at 24.69.53.198 pasted "Pugs build error" (66 lines, 4K) at sial.org/pbot/26284 03:14
dduncan the last few lines maybe all you need
audreyt dduncan: "make" again fixes this
it's a mystery to me too
still investigating
03:14 bmedley left
dduncan I've done "make" 3 times, no change 03:15
I can try running perl makefile.pl ?
audreyt no... a sec
03:16 bsb joined
audreyt try "./util/ghc-pkg-wrapper expose pugs-filepath" and then make 03:16
dduncan one moment ...
pasteling "dduncan" at 24.69.53.198 pasted "result of that last try" (73 lines, 4.6K) at sial.org/pbot/26285 03:17
dduncan so, that command said it did something, but I didn't notice any difference to "make" afterwards ... but you probably know what to look for 03:19
audreyt try r17049...
if it still doesn't work, give me a sec 03:20
pugs_svnbot r17049 | audreyt++ | trunk:
: * expose third-party modules before building pugs.
diff: dev.pugscode.org/changeset/17049
lambdabot Title: Changeset 17049 - Pugs - Trac
pasteling "dduncan" at 24.69.53.198 pasted "and following 17049" (77 lines, 4.6K) at sial.org/pbot/26286 03:21
audreyt fixing 03:22
paste your Pugs.cabal?
dduncan fyi, the last successful Pugs compile I had was r17019 03:23
looking for cabal ...
I assume you just mean Pugs.cabal and not the 'in' version 03:24
audreyt yes.
pasteling "dduncan" at 24.69.53.198 pasted "my Pugs.cabal" (23 lines, 3K) at sial.org/pbot/26287 03:25
dduncan incidentally, that reveals you probably want to update the: copyright: 2005-2006, Audrey Tang as well to say 2007 03:26
pugs_svnbot r17050 | audreyt++ | trunk:
: * Explicitly name required -package names in case Cabal doesn't
: grok build-depends.
diff: dev.pugscode.org/changeset/17050
lambdabot Title: Changeset 17050 - Pugs - Trac
dduncan pulling ...
audreyt k. try 17050. (17051 will be updating copyright)
?eval 1 03:27
pugs_svnbot r17051 | audreyt++ | trunk:
: * Pugs.cabal.in: Copyright years and metadata updates
: prompted by dduncan++.
diff: dev.pugscode.org/changeset/17051
lambdabot Title: Changeset 17051 - Pugs - Trac
evalbot_r17013 1
evalbot_r17041 1
dduncan make still fails ... will nopaste 03:28
audreyt :/
it's highly probable that rm -rf third-party/installed/ and then make will fix this
but that'd be just working around the problem
oh wait. 03:29
dduncan oddly, it didn't paste into the channel this time, anyway: sial.org/pbot/paste
whoops ...
sial.org/pbot/26288 03:30
lambdabot Title: Paste #26288 from "dduncan" at 24.69.53.198
audreyt is your Pugs.cabal.in verbatim?
dduncan on a tangent to that copyright thing ... there is some out of date docs concerning Pugs license matters
are you asking if it looks the same? 03:31
audreyt "svk st Pugs.cabal.in"
or svn
and see if it's modified "M"
dduncan entering that shows no results 03:32
audreyt in any case try r17052.
pugs_svnbot r17052 | audreyt++ | trunk:
: * Remove the comment line from Pugs.cabal.in as it's possible
: that GHC 6.6's Cabal chokes on it.
diff: dev.pugscode.org/changeset/17052
lambdabot Title: Changeset 17052 - Pugs - Trac
audreyt ok, that means it's verbatim
so r17052 should work. if it still doesn't, please nopaste your Pugs.cabal following "make" again
as there's an extra dash before "build-depends" that shouldn't be there.
dduncan I never cease to be amazed by just how fast you are 03:33
audreyt er, well, thanks
dduncan I at least couldn't move that fast 03:34
03:34 ewwr joined
audreyt not accurate enough though -- should've seen the anomaly before r17050 03:34
pasteling "dduncan" at 24.69.53.198 pasted "now with 17052" (68 lines, 4.1K) at sial.org/pbot/26289
dduncan perhaps its worth seeing if others have this problem? 03:35
does anyone else here have trouble building Pugs now?
for all I know, my machine could be acting up
audreyt much better.
no, it's just I tested with 6.6.1 and 6.7-HEAD but not 6.6.0
pugs_svnbot r17053 | audreyt++ | trunk:
: * Pugs.cabal.in: Cater to 6.6's deps.
diff: dev.pugscode.org/changeset/17053
lambdabot Title: Changeset 17053 - Pugs - Trac
audreyt (my machines doesn't have 6.6.0 installed atm)
dduncan that's a thought ... I think I have 6.6.0
audreyt you do. try again? :) 03:36
dduncan well, it did something different this time
audreyt good... 03:37
(I hope)
pasteling "dduncan" at 24.69.53.198 pasted "so 17053 fails differently" (91 lines, 5.4K) at sial.org/pbot/26290
03:38 rlb3 joined
audreyt try ls -l /usr/local/lib/time-*/ghc*/include 03:38
diakopter audreyt: if you have a second pull up svn.pugscode.org/ in a web browser that can do xml/xslt..
audreyt and see if you see HsTime.h
hey pretty 03:39
diakopter++
diakopter web-browser editing. you can commit via the browser lots of single-file types of changes.
audreyt what's this?
03:39 amnesiac joined
diakopter this swedish company donated a license for it 03:40
dduncan oooooo
diakopter it's obfuscated php - www.repos.se
audreyt very impressive!
diakopter I warned him that it'd get reimplemented in a FOSS fashion ;)
he didn't seem to find. I reported some bugs already and they fixed them, pronto. 03:41
audreyt nice
this is really pretty good.
diakopter mind, I mean
audreyt dduncan: do you see HsTime.h anywhere?
diakopter k; sorry to distract from exterminator-work
dduncan no
one moment ...
audreyt thinking
dduncan I don't have the folders it is in 03:42
my /usr/local/lib/ shows ghc-6.6, readline stuff, and perl5
that's all
trying /user/lib/ et al ...
audreyt did you install GHC 6.6 via PowerPC's bindist?
diakopter try the "Details" button in a folder view, when you get a sec 03:43
amnesiac there's no pugs package on DarwinPorts, right?
only the ghc
audreyt amnesiac: correct.
diakopter: cute
diakopter amnesiac: my port can't find pugs.
yeah
dduncan maybe its new for 6.6.1? 03:44
diakopter audreyt: more fun news: guess where perl6.com and perlsix.com|org point, now.
dduncan hm, I just realized I downloaded the ghc 6.6.1 tarball 2 months ago, not sure if I installed it though ... looking ... 03:45
no, I have 6.6 installed me thinks 03:46
I installed ghc-6.6-ppc-apple-darwin.tar.gz specifically, if that means anything
will try 6.6.1 if that helps
audreyt 6.6.1 will definitely work because 6.6 forgot to package HsTime.h 03:47
dduncan that one says ghc-6.6.1-powerpc-apple-darwin.tar.bz2
audreyt let's see if we can correct that
we can't. 03:51
we could arguably ship a copy of HsTime.h, but that'd be inaccurate
I wonder how much updating GHC dep to 6.6.1 may hurt people... 03:52
dduncan if 6.6.1 runs on everything 6.6.0 did, then I don't see why not
besides, Pugs 6.28.0 is a major update ... updated prereqs are more normal for such
audreyt dduncan: can you get ghc 6.6.1 installed and see if it works?
dduncan working on it ... 03:53
question: are things compiled by 6.6.1 binary compatible with 6.6.0, or should I do a make clean of pugs? 03:54
audreyt you should do a make clean.
dduncan okay
03:55 araujo joined
audreyt committing the 6.6.1 change. 03:57
it's in. 04:00
pugs_svnbot r17054 | audreyt++ | trunk:
: * Update GHC dependency to 6.6.1, since it comes with a
: (working) time package, a filepath package, and a Cabal
: package that has far less warts than before.
: * Remove third-party/filepath/ as it's part of GHC 6.6.1.
diff: dev.pugscode.org/changeset/17054
lambdabot Title: Changeset 17054 - Pugs - Trac
dduncan hm, for some reason plain 'make install' doesn't work citing permissions, so I have to 'sudo make install' 04:03
anyway, that may be it ... trying to use ...
audreyt or just
sudo chown -R dduncan .
oh you mean for ghc 6.6.1. 04:04
pugs_svnbot r17055 | audreyt++ | trunk:
: * build_pugs.pl: Chase the 6.6.1 change.
diff: dev.pugscode.org/changeset/17055
lambdabot Title: Changeset 17055 - Pugs - Trac
audreyt then yes, sudo make install is correct.
buliding r17055 should now work (I've verified locally)
Pugs's test coverage is now up to 60% (was 50%) 04:05
still not very good...
(the current coverage is at pugscode.org/~audreyt/hpc_index.html) 04:07
dduncan I was going to say earlier ... the 3 copies of docs/...01Overview.html refer to the license of Pugs ... I updated the english copy the other day, but couldn't read the 2 zh- versions to update theirs 04:08
in short, it should just say Artistic 2.0, and not Artistic 2 beta + GPL 04:09
this is the current/new english version: Is Pugs free software?</h2> 04:10
<p>
Yes. It is available under the <a
href="svn.pugscode.org/pugs/LICENSE/Artis...t;Artistic
License version 2.0</a>.</p>
the zh- copies should say that too
audreyt fixed
pugs_svnbot r17056 | audreyt++ | trunk: 04:11
: * docs/zh-*/: Reflect that Artistic 2.0 is now out. dduncan++
diff: dev.pugscode.org/changeset/17056
lambdabot Title: Changeset 17056 - Pugs - Trac
dduncan on invoking the Pugs makefile now, I get: darren-duncans-power-mac-g4:~ darrenduncan$ ghc -v 04:15
dyld: Library not loaded: GNUreadline.framework/GNUreadline
Referenced from: /usr/local/lib/ghc-6.6.1/ghc-6.6.1
Reason: image not found
Trace/BPT trap
maybe I need to install a newer readline?
audreyt no
dduncan or add something to my path
audreyt let's see... 04:16
add something, perhaps
a sec
www.informatik.uni-bremen.de/agbkb/.../mac_e.htm
lambdabot Title: Hets: libraries for Macintosh, tinyurl.com/22srs3
audreyt see the first para's instructions
if installed in /Library/Frameworks (I'd recommend that), you need to run the unzip using sudo 04:17
dduncan readline and GMP? okay, will do both 04:19
diakopter reads cpan6.org/papers/2007gpw.odp 04:23
pugs_svnbot r17057 | audreyt++ | trunk: 04:26
: * build_pugs.pl: Even more sequashing of obsolete cases
: for GHC 6.4.x.
diff: dev.pugscode.org/changeset/17057
lambdabot Title: Changeset 17057 - Pugs - Trac
04:33 takanori joined
dduncan so I installed readline 04:36
I started to install GMP too, but apparently GMP was already there, so I cancelled
audreyt ok
dduncan so ghc -v works now
trying to run the Pugs makefile complains that I should rerun ranlib(1) 04:37
audreyt cool... I just built pugs with r17058
pugs_svnbot r17058 | audreyt++ | trunk:
: * Pugs.Run, Pugs: Chase API change in System.FilePath (joinFileName -> combine).
diff: dev.pugscode.org/changeset/17058
lambdabot Title: Changeset 17058 - Pugs - Trac
audreyt and GHC 6.6.1
dduncan: ranlib?
sudo ranlib /usr/local/lib/ghc-6.6.1/*.a
like that?
weird, ghc 6.6.1's installer should do that for you 04:38
dduncan it says, eg: [1 of 1] Compiling Main ( Setup.lhs, Setup.o )
Linking ./Setup ...
audreyt ah, you installed readline _after_ ghc. then run ranlib by hand
dduncan I could reinstall ghc
audreyt nah, just run ranlib by hand :)
dduncan probably easier to repeat
audreyt and then pull r17058 and it'll work... I think
dduncan if its just that one line ... will do ...
audreyt probably.
what's the error message mentioning ranlib? 04:39
dduncan running the line you said fixed it 04:40
the makefile.pl now progresses beyond what it did before
but an example error was: /usr/bin/ld: table of contents for archive: /usr/local/lib/ghc-6.6.1/libHSCabal.a is out of date; rerun ranlib(1) (can't load from it)
4 lines like that
audreyt k, got it 04:41
dduncan the makefile finished
oddly enough, Pugs' makefile.pl ended to say Readline support disabled, though I just installed it
unless this one wants a different copy
audreyt mmm.
dduncan or I have to rerun something
audreyt ignore that for a moment and see if pugs builds first?
dduncan will do 04:42
04:46 ewr joined
dduncan now starting 'make' ... if like before, it'll take about an hour 04:49
audreyt "make soon" will do too just to ensure that it builds... 04:50
dduncan but that's a slower runtime? 04:54
I'll just wait for the normal
now at 34 of 109
audreyt ok... I'll bbiab 04:56
dduncan now at 41 04:57
50 04:58
warnings may be a non-issue, but I get: 05:00
[ 54 of 109] Compiling Pugs.Parser.Charnames ( src/Pugs/Parser/Charnames.hs, dist/build/Pugs/Parser/Charnames.o ) 05:01
warning: implicit declaration of function 'perl5_SvPV'
warning: implicit declaration of function 'perl5_SvIV'
[ 55 of 109] Compiling Pugs.AST[boot] ( src/Pugs/AST.hs-boot, nothing )
hm, some of that was cut out
nopaste ...
hm, I"m being told I can't nopaste for awhile, something about floods 05:02
diakopter dduncan: rafb.net/paste/
dduncan true, there are other pastebots ... 05:03
diakopter I mean, then just paste the nopaste url
dduncan rafb.net/p/YrMsLp11.html 05:04
lambdabot Title: Nopaste - warning when compiling pugs
dduncan those are probably known about and harmless, though 05:05
05:21 jisom joined
spinclad btw: debian etch has ghc 6.6, testing has 6.6.1, so this will push me into testing where i belong. it's been nice in a way staying in etch as it hit release and became stable, but it seems like time to reenter the stream. 05:22
and audreyt! nice to see you back! (even if lightly) 05:23
diakopter spinclad: why not apt-get install ghc/testing 05:24
(assuming you add the sources and a proper /etc/apt/preferences)
spinclad expect it'll want new libraries and such, pushing me most of the way into testing anyway. i'll check it out and see. 05:25
diakopter hmmm; someone must've fixed the autobuilds 05:26
(for evalbot)
yup: The following packages have unmet dependencies: ghc6: Depends: libc6 (>= 2.5-5) but 2.3.6.ds1-13 is to be installed 05:28
spinclad that's the big one. i could upgrade libc6 and leave all else old, and see how they get along. 05:29
dduncan 109 of 109 05:32
make is done 05:37
so it took about 45 minutes
now starting a smoke 05:38
last time, that was around 2 hours
note that the readline functionality doesn't seem to work ... hitting up-arrow doesn't do anything useful 05:41
so that makefile.pl note about it being disabled appears to be true
05:44 iblechbot joined 05:56 weinig joined 05:57 wer joined
diakopter sorry about http/svn being down for a while... apache2 got restarted while I was still in the midst of changing some config items. 05:58
(few minutes, it seems) 05:59
06:06 ashleyb joined 06:16 kanru joined
spinclad so: installing ghc6 6.6.1-2, i end up pulling in libc6*, binutils, tzdata, and the new libc6-dev wants linux-libc-dev 2.6.21-6, kernel headers from a later kernel than i'm running. makes you wonder a bit. i'll keep my eyes peeled for odd behaviors... 06:19
(and of course the new libghc6-*-dev's) 06:20
diakopter spinclad: interesting... 06:22
06:42 bsb left 06:51 penk joined
Aankhen`` diakopter: I keel you. 07:08
dduncan smoke done
07:09 wer joined
dduncan about 90 wallclock minutes it looks like 07:09
07:11 wer is now known as slacker403
dduncan 87.13 according to the smoke itself ... this is probably inline with the fastest smokes I've had 07:16
07:19 Psyche^ joined 07:21 rho joined 07:30 rintaro_ joined 07:37 Psyche^ is now known as Patterner 07:42 elmex joined 07:49 b00t joined 07:57 crashmatrix joined 08:02 iblechbot joined 08:17 monomorph joined 08:20 slacker403 joined 08:23 Entonian joined 08:49 dduncan left 08:52 masak joined
masak @tell moritz imressive abbr.dat list of abbreviations. though i have a feeling that many more of them could use the (?i:) modifier, on account of many people writing in all-lowercase, like I do now. (i see it in the logs quite a bit.) however, that may lead to false positives, as with 'tap' for example. it's a tough problem 09:00
lambdabot Consider it noted.
masak it's deeply reassuring to see audreyt committing to pugs again, even if it's just cleanups 09:04
09:14 Blwood joined
masak oh, new GHC version overnight :) 09:23
09:32 slacker403 joined 09:35 hakobe_ joined 09:37 barney joined 09:41 chris2 joined 09:49 ofer joined
meppl good morning 09:57
10:01 buetow joined
masak meppl: good day 10:20
meppl hello masak 10:21
10:32 barney joined
Aankhen`` masak: Indeed, false positives are the sole reason why they don't have the (?i:). 10:40
masak: There are only a few of them which it seemed to make sense to expand even in lowercase. Though I have to admit that the choice might have been arbitrary. 10:41
Aankhen`` heads off to bed.
10:43 slacker403 joined 10:46 jedai joined 10:55 kane_ joined 10:56 b_jonas joined 11:44 unobe is now known as unobe_away 11:55 ewr joined, fglock joined
masak @tell Aankhen` yes, the reason I came to think of it was that I saw someone writing first "OSCON" and then "oscon" in the logs, and only the former was abbr-underlined. I went in to add a (?i:), but then couldn't find a compelling reason to to it just to to the OSCON abbr 12:08
lambdabot Consider it noted.
12:52 rho joined, ilogger2 joined 13:07 wer joined 13:15 kroo joined 13:19 kroo left 13:25 theorb joined 13:44 snake__ joined 13:45 snake__ left, funktio joined 13:51 vel joined
masak I think I'll do that, on second thought. 'f anyone says oscon, they're talking about OSCON :) 14:02
pugs_svnbot r17059 | masak++ | trunk: 14:07
: OSCON is now case-sensitive
diff: dev.pugscode.org/changeset/17059
lambdabot Title: Changeset 17059 - Pugs - Trac
pugs_svnbot r17060 | fglock++ | trunk: 14:16
: [kp6] AST-dumper now shows a detailed ::Pad, including lexical declarations
diff: dev.pugscode.org/changeset/17060
lambdabot Title: Changeset 17060 - Pugs - Trac
14:18 ewr joined 14:20 cmarcelo joined 14:24 vel joined 14:39 polettix joined 14:47 prism joined 14:48 hanson122 joined, pifuruan joined 14:56 daxim joined 14:59 vel joined 15:04 hanson122 joined, rindolf joined 15:11 hanson122 joined 15:19 hanson122 left
obra TimToady: ping 15:26
pmichaud hello, obra 15:27
obra pmichaud: hi! 15:28
how's things?
pmichaud they're good
15:30 ewr joined
moritz masak: IMHO false positives are far worse thatn false negatives, so I'll leave it as it is exept on indivual request 15:31
lambdabot moritz: You have 1 new message. '/msg lambdabot @messages' to read it.
moritz masak: of course you are welcome to edit abbr.dat of course ;)
TimToady obra: pong 15:32
I slept longer than I thought...
obra TimToady: good morning! 15:33
TimToady blurg
obra take your time, but whenever you feel like it, we'd love to drag you out to brunch
TimToady I'm more or less ready 15:34
obra ok. I think we'd be good to go in ~10 minutes or so. Anything in particular you feel like, foodwise? Easy options at this hour include, crepes, eggs, waffles, pancakes, sandwiches, salads, burritos. 15:36
diakopter mmmmmm so hungry
15:39 offby1 joined 15:42 fglock left
TimToady Let's see, I'm allergic to crepes, eggs, waffles, pancakes, sandwiches, and burritos. That makes it easy. 15:44
Juerd Ruby's array operator "-" is useful here. Does Perl 6 have something like it yet? 15:50
['crepes', 'eggs', 'waffles', 'pancakes', 'sandwitches', 'salads', 'burritos'] - ['crepes', 'eggs', 'waffles', 'pancakes', 'sandwiches', 'burritos'] results in ['salads'] 15:51
integral set difference, or some other form of difference? 15:52
Juerd I'm not familiar with set jargon.
Does difference include addition?
integral I was just wondering if order is significant 15:53
Juerd It's not
Maybe @foo.without:LIST in Perl 6
@foo.=with:LIST would be the same as @foo.push:LIST :) 15:55
lambdabot Unknown command, try @list
wolverian Juerd, there are set operators already, I think.
Juerd wolverian: Sets don't have duplicates, right?
wolverian multisets do. normal sets don't. 15:56
Juerd Blergh.
I hate that I have to learn all this stuff to use Perl 6.
diakopter well, to follow maths-english, it would be written ['crepes', 'eggs', 'waffles', 'pancakes', 'sandwitches', 'salads', 'burritos']['crepes', 'eggs', 'waffles', 'pancakes', 'sandwitches', 'salads', 'burritos'] or $a$b if $a and $b are Array objects
Juerd wolverian: Is a multiset an array without order? 15:57
diakopter er, faulty paste there in the second []
wolverian Juerd, basically, yes.
Juerd, this probably falls under the underspecced libraries synopsis, though.. 15:58
Juerd wolverian: Is the lack of order ever needed or useful? (i.e. why not just use an array?)
wolverian Juerd, equality. 15:59
Juerd sort(@foo) ~~ sort(@bar)
Insufficient?
wolverian shrug. haskell doesn't have multisets either, I think. 16:00
Juerd A whole new datatype, with new operators, new everything, seems a bit much if it's just for equality testing. I can't remember when I last needed to compare two hashes or two arrays or two lists.
wolverian well, sets are pretty important, I think. multisets not so.
Juerd I don't see the point of sets either, if we already have hashes.
For all I care, we could have a kind of hash in which every value is forced to 1 :) 16:01
16:01 BinGOs joined
Juerd Until someone shows me why a set is more useful, of course :) 16:01
I'm willing to learn stuff, if it's pragmatic and not just academically more correct.
wolverian the keys of a hash are a set. sets are not more useful than hashes, they're a component of hashes. 16:02
or a way to analyse them.
wolverian shrugs again
Juerd wolverian: Does the Perl language need a datatype for sets, given hashes?
Not talking about internals, now.
wolverian perl5 didn't need 'em :)
s/didn't/doesn't/ 16:03
Juerd Sets and set operators, to me, look like language bloat. It's something that could and should be done in a module IMO.
wolverian I think that's the plan.
Juerd That's a relief
wolverian well, okay, 'Set' is mentioned in the built-in immutable types 16:04
perlcabal.org/syn/S02.html#Immutable_types
lambdabot Title: S02
wolverian yes, it's a more typeful approach than perl5 16:10
hopefully it won't be too awkward.. we still have implicit coercions, afaik :)
16:25 greentea joined 16:27 Eidolos joined
Gothmog_ A hashtable ist just a way to efficiently implement a set or multiset, with values in {0, 1} or {0, 1, 2, ...}, respectively. You may think of sets and multisets as special cases of a hashtable. 16:34
16:35 offby1 left
masak Gothmog_: the 0 values are not necessary if you delete kv pairs instead of zeroing them out 16:36
probably makes searches and stuff easier, too 16:37
16:42 ewr joined 16:51 vel joined
Gothmog_ Yeah, of course. I just thought it would be more clear. 16:58
17:01 vel_ joined 17:04 funktio joined 17:13 mr_ank joined 17:20 Blwood_ joined 17:22 mr_ank joined 17:37 mr_ank joined 17:40 Blwood joined 17:48 Psyche^ joined 17:54 ewr joined 18:02 Psyche^ is now known as Patterner 18:09 cmarcelo joined 18:10 bonesss joined 18:15 rlb3_ joined 18:17 aindilis joined, rlb3 joined 18:21 rlb3_ joined
TimToady See S02:829 for p6's take on Sets and Bags and their relationship to hashes. Pretty much exactly what you folks were reinventing here... 18:22
SamB wonders if that means it's a good idea or not 18:24
amnesiac TimToady, Bags as Counted Sets? 18:27
TimToady yes 18:28
amnesiac ok.
TimToady or Sets as Bags that can only count in Bool
amnesiac right 18:29
TimToady wonders if people can't even read as far as S03, how are they ever going to get to S13... 18:30
s/3/2
Juerd still has rereading synopses on his todo list. 18:31
amnesiac TimToady, because.... of lack of REAL interest, or probably too "deep" for them.
integral Tablet, capsule or syrup (or suppository for the continent) forms would be handy
TimToady course, if people actually did read them and comment, we'd have to change 'em all again :)
Juerd amnesiac: Or plain lack of tuits and concentration. 18:32
And the threshold gets higher as Perl 6 gets more complex
amnesiac concentration is part of interest, if you have no interest, reading them by "force" is having no concentration at all.
well at least, I can't concentrate if I'm reading something I don't want to.
18:33 Jmax joined
integral at somepoint "perl6doc Bag" "perl6doc Set" will be nice 18:33
Juerd amnesiac: If I don't want to read them, why would I have it on my todo list? :) 18:34
amnesiac Juerd, agreed, that's my point :P
Juerd What exactly is your point? :)
A question?
amnesiac integral, thing is... a Bag and a Set are abstract concepts, or you mean how they will be implemented?
18:34 pmurias joined
amnesiac Juerd, oh my answer is TimToady's question about why people doesn't read above S02. 18:35
integral amnesiac: I thought they were interfaces
amnesiac integral, A Bag and a Set is a form of representation of a data collection, with some constraints. 18:37
18:37 weinig joined
TimToady need to distinguish values from containers 18:37
amnesiac TimToady, as for the relation ship of values and Bag/Sets? 18:40
eerr, relationship of data collection and Bag/Sets? 18:41
TimToady sets and bags are immutable values. a KeySet can represent one set at a time, but when you change it, it's representing a different set
18:43 dduncan joined
wolverian yeah, I need to read more than just the names of the types sometime.. :) 18:46
I like Blob.. 18:47
TimToady, do you have any interface sketches for the basic types? 18:48
TimToady no, that would be a good thing for someone to work on
amnesiac sketches? 18:49
TimToady there might be some of it in Prelude.pm, but I think it's mostly in Haskell
amnesiac errr I mean, interface sketches? how's that?
TimToady to the extent that there is anythin
well, they'd be roles in p6
wolverian yeah, I meant role interfaces.. :)
TimToady nap & 18:52
amnesiac gnite 18:53
dduncan its a mid-day nap, afaik 18:54
18:59 jisom joined 19:05 ewr joined, Belaf joined 19:06 rashakil left 19:14 bonesss joined 19:24 Aankhen`` joined 19:51 fglock joined 20:16 idiotmax joined, ewr joined 20:17 idiotmax joined
pugs_svnbot r17061 | fglock++ | trunk: 20:21
: [kp6] added Visitor::Global - looks up lexical variables; undeclared vars are then looked up in GLOBAL (unfinished)
diff: dev.pugscode.org/changeset/17061
lambdabot Title: Changeset 17061 - Pugs - Trac
wolverian huh, svk 2.0.1's signatures file is corrupt, or I have a bad Module::Signature, or something 20:30
Aankhen`` Shhh.
wolverian well, sent off a cpan testers report. 20:32
those aren't very helpful if it's a local bug though :/ 20:33
Aankhen`` Shhhh!
20:44 vel joined 20:47 fglock joined 20:51 weinig joined 21:03 vel joined 21:06 aindilis joined
obra TimToady: what sounds tasty for dinner? 21:11
fglock obra: hey 21:15
lambdabot fglock: You have 3 new messages. '/msg lambdabot @messages' to read them.
21:17 turrepurre joined
obra hey fglock. how goes? 21:17
pugs_svnbot r17062 | fglock++ | trunk:
: [kp6] - GLOBAL/lexical lookup is decided at compile-time
: - say() works inside BEGIN blocks
diff: dev.pugscode.org/changeset/17062
obra how's the hacking?
lambdabot Title: Changeset 17062 - Pugs - Trac
fglock obra: now in kp6 phase 4 :) 21:18
obra nice.
roadmap updated?
fglock working on the runtime infrastructure
obra cool 21:19
fglock yes, I'd like to discuss it with you when possible
TimToady hmm, most things sound tasty. My wife married me because I'd eat almost anything... :/ 21:20
obra that's not cause for a ":/"
fglock: likely tomorrow or monday or starting tonight by email is best 21:21
TimToady: light or heavy?
"Would you prefer a light or heavy meal" to avoid the horrible pun potential.
TimToady water you talking about?
obra groans 21:22
obra remembers the mantra 'don't taunt the Larry. He taunts back'
TimToady oh, somewhere about medium-light would be good 21:23
obra nearby, we've got more japanese, thai, brazilian, bar-and-grill, burritos, sandwiches-and-salads. korean. decent chinese. decent american chinese.
I suspect we couldn't get into the good tapas place in finite time.
obra consults kaia 21:24
Mary Chung's could work. Slightly americanized chinese but quite tasty. ~famous in MIT circles. 21:25
TimToady that would be fine by me
obra excellent. should probably head out in 10-15 minutes to avoid being rushed. 21:26
unless that's too much of a rush on this end. ;) 21:27
21:28 ewr joined
fglock obra: mail sent 21:36
21:36 unobe_away is now known as unobe
pmurias fglock: hi 21:37
fglock pmurias: hi
got say() working inside BEGIN 21:38
GLOBAL is not imported; it's looked up instead
pmurias looking at the changes.. 21:40
looks good 21:41
21:44 Limbic_Region joined
fglock I wonder if "global" regexes like ws() should be defined in GLOBAL 21:45
pmurias i seem to remember vagly GLOBAL is organised into many sections... or was it the Prelude 21:46
fglock probably not becuse of method vs. sub
pmurias should ws be in a base Grammar? 21:47
fglock regexes should be defined in a class, under the current architecture
pmurias *shoudn't
fglock yes
pmurias sorry, i'm typing strangly because it's late for me 21:48
21:50 jisom_ joined
pmurias i'll better go to sleep than, bye& 21:50
21:51 penk joined 21:54 hakobe_ joined 22:05 Limbic_Region joined 22:10 Blwood joined 22:13 Blwood joined 22:32 fglock left 22:40 ewr joined 22:44 thorat joined
devbot6 dev: Ticket #26 (defect created): [BUG] 'dmake' broken on Win32 with GHC 6.6.1 <dev.pugscode.org/ticket/26> 23:13
diakopter maxcdo: you around? 23:14
23:16 mako132_ joined 23:23 bonesss joined 23:33 pro21ms4 joined 23:39 vel joined 23:51 ewr joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!