pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, smop: etc.) || We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by TimToady on 25 January 2008.
mncharity thing is it isn't really a syntax tree. abstract semantic tree? it's basically a kernel language. 00:02
00:02 monomorph left
mncharity /me tries to remember that thing clams grow perls around... googles... 00:03
not useful. 00:07
err, rather, didn't see anything. 00:08
00:08 wknight-away is now known as wknight8111 00:11 cmarcelo joined
mncharity re kernel language... or maybe not? some nodes like If could be calls to macros. it might be nice to have a language form where "is parsed" macro applications have been converted to normal macro-with-arguments form, but otherwise nothing has changed. so if can just be a macro prelude::if . 00:19
so (almost?) everything would be a Call node, and the rest would be up to the compiler+runtime. 00:20
00:21 hanekomu joined
mncharity ie, it would be scheme. :) 00:21
00:22 allbery_b joined
pugs_svnbot r19974 | lwall++ | [gimme5] alternatives within alternatives close to correct now 00:30
diff: dev.pugscode.org/changeset/19974
lambdabot Title: Changeset 19974 - Pugs - Trac
Juerd "Close to correct". Is that where the next 80% begin?
mncharity TimToady: woot. have 42's match yaml. curiously, if a lex/ directory doesn't exist, I instead get a segfault. :P 00:32
Tene sub node { lazy { (:value(80), :next( node() ) ) } }
TimToady mncharity: well, yes, and there's no provision for invalidating the lexer cache beyond rm lex/* :) 00:33
likewise yamlg5, but at least that one gets autocreated 00:34
mncharity :) rm redsix_cache/* too :) 00:35
TimToady but I must shut my eyes for a bit. that's why the checkin is only close to correct
it seems to do longest token right on 0xdeadbeef and then fail the match
probably because character classes are borken
mncharity np. just about end of hacking day for me. hope to bang on things tomorrow.
TimToady but nap &
mncharity obra: re meeting tomorrow, why don't we touch base tomorrow afternoon. depending on how the day goes, we might shorten it, or find it no longer needed. 00:41
obra sure
I'll be on irc and within 4 blocks of davis ;) 00:42
00:43 alc joined 00:44 hanekomu left
mncharity ok :) 00:48
ingy: re YAML::XS and no known bugs, there seem to be 3 in RT... :) rt.cpan.org/Public/Dist/Display.htm...ML-LibYAML fyi.
lambdabot Title: Active bugs for YAML-LibYAML
01:06 cnhackTNT joined 01:08 devogon left
mncharity ingy: also, trac.yaml.com/trac/wiki/PlYaml no longer works. 01:08
lambdabot Title: Com.org - Only the best links ...
ingy mncharity: oh yeah 01:13
that's #936 on my todo list.
thanks :)
mncharity: bugs in YAML::XS 01:14
hmm
Coke ~~ at ingy.
ingy that I do care about...
that reminds. 01:16
I could really use a Coke
Coke++
mugwump For cleaning engine parts?
mncharity re #936, ooo, a small finite number :) 01:17
ingy the engine btween my ears, yes
mugwump!
mugwump ingy! :)
ingy are you in .nz?
mugwump yup
ingy thanks god!
I was scared you might be stalking me again 01:18
ingy looks out his window...
mugwump hehe ... no, I'm honouring the restraining order
ingy lol
01:18 utopia_ joined 01:19 pbuetow joined
utopia_ Has Larry Wall stopped working on perl6? 01:20
TimToady Larry Wall never was working on perl6 01:21
utopia_ oh right
mugwump I saw some commits from him the other day to pugs though, are you sure TimToady ?
TimToady that's not perl6... 01:22
mugwump oh yeah
utopia_ Will there be a perl6?
TimToady there will be lots of them 01:23
utopia_ can you explain that a bit more?
mugwump perl6 is a language spec
sorry, "Perl 6" is a language spec
Rakudo was very briefly called perl6 01:24
wknight8111 the space is key
TimToady perl6 is probably just a symlink to something else that runs Perl 6
utopia_ ah ok :)
mugwump Much like Java isn't an implementation
mncharity good night & 01:29
01:29 mncharity left 01:32 thoughtpolice joined 01:37 Jamtech joined, Coke left 01:58 ting_ joined, ting left 02:03 ImustDIE joined
gwern hm. so the bot doesn't post new bug reports like dev.pugscode.org/ticket/32 ? 02:13
lambdabot Title: #32 (Using hsc2hs's #def feature leads to pain, don't use it) - Pugs - Trac
pugs_svnbot r19975 | particle++ | [spec] really fix pod this time 02:21
diff: dev.pugscode.org/changeset/19975
lambdabot Title: Changeset 19975 - Pugs - Trac
02:25 [particle] left 02:29 zakame joined 02:38 ImustDIE left 02:39 Schwern joined
pugs_svnbot r19976 | lwall++ | [gimme5] handles "our" package names more properly 02:48
r19976 | lwall++ | [Cursor5.pm] now does character classes much better
diff: dev.pugscode.org/changeset/19976
lambdabot Title: Changeset 19976 - Pugs - Trac
02:59 kanru left, lyokato_ joined 03:02 cmarcelo left 03:03 pbuetow left 03:05 Limbic_Region left 03:08 Psyche^ joined 03:10 fridim_ left 03:13 eternaleye joined 03:16 wknight8111 left 03:20 [particle] joined 03:27 Patterner left, Psyche^ is now known as Patterner 03:33 jhorwitz left 03:36 Auzon joined 03:37 alester_ joined 03:38 thoughtpolice left 03:39 SubStack left 03:40 SubStack joined 03:47 stevan_ left 04:04 justatheory joined 04:16 SCalimlim2 joined 04:17 SCalimlim2 left 04:18 RayMagini left 04:21 RayMagini joined
pugs_svnbot r19977 | lwall++ | [gimme5] stop trying when you succeed :) 04:31
diff: dev.pugscode.org/changeset/19977
lambdabot Title: Changeset 19977 - Pugs - Trac
04:34 justatheory left 04:45 jferrero left 04:53 justatheory joined 05:09 Schwern left
spinclad @tell mncharity i'm boston area and interested. could put my l33t kibitzing skills to good use. i'll check with you and obra and aim to come by. 05:38
lambdabot Consider it noted.
spinclad obra: see above ^ 05:39
05:39 justatheory left
spinclad @tell mncharity re seeds of perl^Wpearl: sand 05:40
lambdabot Consider it noted.
05:42 thoughtpolice joined 05:44 thoughtpolice left
spinclad @tell mncharity as [particle] says, i'd look at parrot's --target=past too for inspiration. 05:49
lambdabot Consider it noted.
05:49 lisppaste3 left, Zygo left 05:50 Zygo joined, njbartlett left 05:52 zakame left 05:57 lisppaste3 joined 06:09 DarkWolf84 left 06:11 zakame joined 06:24 devogon joined 06:28 Jamtech left 06:30 alester_ left 06:34 Schwern joined 06:35 BinGOs left 06:42 Zygo left 06:47 BinGOs joined 06:48 zakame left 06:51 BinGOs_ joined 06:53 zakame joined 06:54 zakame left, zakame joined 07:05 BinGOs left, hoowa joined
pugs_svnbot r19978 | lwall++ | [gimme5/Cursor5] now preserves orig without copying 07:11
r19978 | lwall++ | expect_term limps along, before/after work better
r19978 | lwall++ | removed redundant before around before
r19978 | lwall++ | various other cleanups
diff: dev.pugscode.org/changeset/19978
lambdabot Title: Changeset 19978 - Pugs - Trac
07:31 jisom joined, BinGOs_ is now known as BinGOs 07:36 njbartlett joined
pugs_svnbot r19979 | lwall++ | [STD] getting EXPR back up toward runnability 07:38
diff: dev.pugscode.org/changeset/19979
lambdabot Title: Changeset 19979 - Pugs - Trac
07:41 cnhackTN1 joined 07:46 cognominal_ left 07:48 cnhackTNT left 07:58 Aankhen`` joined 07:59 njbartlett left 08:14 iblechbot joined, razvanm02 joined 08:15 drrho joined 08:18 Grrrr joined 08:28 penk left 08:30 IllvilJa joined, penk joined 08:38 hoowa left 08:41 monomorph joined 08:42 xinming left 08:43 xinming joined 08:46 Aankhen`` left 08:55 zakame left 09:01 drrho left
rakudo_svn r25979 | kjs++ | [rakudo] update STATUS: until/while/repeat has been implemented. 09:12
09:21 Aankhen`` joined 09:24 masak joined
Juerd Maybe Perl 6 can prevent that cats die 09:33
masak the musical? 09:46
too late for that, I think...
09:55 cnhackTN1 is now known as cnhackTNT 10:01 cnhackTNT left 10:20 penk left 10:25 alc left 10:28 ebassi joined 10:44 jisom left 10:51 marmic left 10:55 marmic joined 11:01 cognominal_ joined 11:09 njbartlett joined 11:13 laye joined 11:33 njbartlett left 11:43 drrho joined 11:44 ebassi left 11:51 masak left 11:52 masak joined, Schwern left
pugs_svnbot r19980 | ruoso++ | [smop] Lowlevel method implementation. This will be the one used by the code generated in s1p. It is not "the" Method class, but it complies to the Method API, and thanks to the REPR polymorphism, we can have a lowlevel C implementation of Method. 11:52
diff: dev.pugscode.org/changeset/19980
lambdabot Title: Changeset 19980 - Pugs - Trac
11:53 Zygo joined 11:58 Schwern joined 12:16 Psyche^ joined, Patterner left, Psyche^ is now known as Patterner 12:19 njbartlett joined 12:27 zamolxes joined 12:36 smtms left 12:39 zamolxes left 12:40 zamolxes joined 12:41 njbartlett left 12:50 zamolxes_ joined 12:51 nipotan is now known as nipotaway 12:52 zamolxes left 12:57 avar left, avar joined 12:58 zamolxes joined 13:03 lyokato_ left 13:04 zamolxes_ left 13:06 cmarcelo joined 13:09 zamolxes_ joined 13:11 zamolxes left 13:15 pbuetow joined 13:17 zamolxes joined 13:20 pbuetow left, pbuetow joined 13:23 zamolxes_ left 13:28 zamolxes_ joined 13:31 zamolxes left 13:35 zamolxes joined 13:37 zamolxes_ left 13:41 zamolxes_ joined 13:44 zamolxes left 13:49 zamolxes joined 13:52 zamolxes_ left 13:56 alester_ joined 13:57 zamolxes_ joined 13:59 zamolxes left, alester_ left 14:05 awwaiid left 14:06 zamolxes_ left 14:16 Limbic_Region joined 14:37 chris2 joined 14:42 jhorwitz joined 14:45 Daveman left 14:49 ofer joined 14:57 cmarcelo left 15:06 orafu joined 15:08 kanru joined 15:16 cmarcelo joined
Tene the massive crossposting on the p6 lists is painful. 15:17
Limbic_Region Tene - why? 15:22
ruoso indeed...
Limbic_Region or is gmail's autothreading blanketing me ? 15:23
ruoso Limbic_Region, are you in perl6-all?
Tene Because I'm getting mostly the same messages in three different mailboxes.
But not all, so I have to read through the threads in all three mailboxes if I want to get all of the messages. 15:24
Limbic_Region ruoso - I belong to at least 2 of the 3 lists being addressed - not sure about perl6-all
but in any event, I am only seeing 1 copy of each
Tene - in that case, I have just sent an offlist email to a handful of individuals and you were one of them 15:25
as were you ruoso
Tene I got it.
15:27 mncharity joined 15:30 stevan_ joined
mncharity spinclad: great :) 15:31
lambdabot mncharity: You have 3 new messages. '/msg lambdabot @messages' to read them.
mncharity lambdabot: re sand and pearls, en.wikipedia.org/wiki/Pearl claims parasites and organic material are more often the cause. /sand/ 15:33
err, spinclad: ^
ruoso Limbic_Region, I've seen it already... and I support your idea... 15:35
I'm not sure what can I do right now for that matter... 15:36
Limbic_Region not everyone is a mover or a shaker
ruoso maybe I've fullfilled my shaking role already ;P 15:37
Limbic_Region on the other hand, I wanted to make sure that the people who were in a position to affect change realized their were people worth making that change for
in a nutshell, I am tired of people talking about how good an idea is or isn't an actually do something
and proof is in the pudding
[particle] there's pudding??! 15:38
Limbic_Region I assume you are still willing to accept funding for working on SMOP
[particle] - yes, but the cake is a lie
ruoso yeah... the time I was able to invest in SMOP is dropping... the funding would certainly help on that...
Limbic_Region btw, WTF is SMOP anyway? Is it the alternate runtime for kp6? 15:39
spinclad mncharity: re sand, well, yes, any irritant could be a seed. sensible thing to do to a parasite, cover it with nacre... i remember hearing pearl-culturers use bits of sand, but how reliable is that? 15:42
Tene Can I get TPF funds to use for pudding?
spinclad make it a figgy pudding, and bring it right here! 15:43
masak mm, figgy...
Limbic_Region anyone know how TimToad is coming along with the longest token matcher? 15:44
er TimToady that is
[particle] yes, it's mainly done 15:48
Limbic_Region oh great 15:49
I had seen in passing that it seemed like there was progress - I just wasn't sure if that was a false start
15:51 Limbic_Region left
ruoso Limbic_Region, www.perlfoundation.org/perl6/index.cgi?smop 15:51
spinclad he's been switching from running on pugs to perl5 -- see gimme5
lambdabot Title: SMOP / Perl 6
15:51 Limbic_Region joined
spinclad and STD5 15:51
ruoso Limbic_Region, www.perlfoundation.org/perl6/index.cgi?smop
lambdabot Title: SMOP / Perl 6
ruoso it seems you lost that link 15:52
Limbic_Region I will have to check the channel logs to see what I missed
Is TimToady happy with the longest token matcher he came up with or did he have to comprimise in undesireable ways?
spinclad still in progress i believe 15:53
he's talked with mncharity about it lately; see logs 15:54
Limbic_Region [particle] just said it was mainly done
[particle] yes, pugs was too slow, so he's now proceeding in perl 5 15:55
see src/perl6/gimme5 and Cursor5.pm
Limbic_Region IOW, the concept/algorithm is done and he is just toying with the implementation 15:56
[particle] well, it has forced some rethinking about STD, but it's pretty much details now
"well into the second 80%" 15:57
15:57 Lorn joined 15:58 jhorwitz_ joined
Limbic_Region 80% of the way through the second 80% 15:58
15:58 zamolxes joined
spinclad beginning the second 80% of it... 15:59
Limbic_Region wanders off to a meeting 16:00
16:00 Daveman joined 16:05 fridim_ joined 16:07 silug_ left 16:15 jhorwitz left, silug joined 16:25 meppl joined 16:29 Auzon left 16:38 TJCRI joined 16:39 barney joined 16:42 masak left 16:51 DarkWolf84 joined 16:53 utopia_ left 16:55 Limbic_Region left 16:57 ahmadz joined, chris2 left, ahmadz left 16:58 arguile joined 16:59 Aankhen`` left
mncharity Limbic_Region: [particle]: re 'anyone know how TimToad is coming along with the longest token matcher?', 'yes, it's mainly done'. I believe that is incorrect. Rather, it's a research project, very much a work in progress. Eg, a recent landmark was the ability to parse "42". 17:02
It has however recently become more active, switching from using pugs to using p5. 17:03
spinclad: re 'pearl-culturers use bits of sand', no. the wp article describes. 17:04
hmm... puuuddding.... (almost lunchtime)
17:04 razvanm02 left
[particle] mncharity: use.perl.org/~chromatic/journal/35657 "pretty much have the longest-token autolexer working, using tagged regular expressions" -- Larry 17:06
lambdabot Title: Journal of chromatic (983)
[particle] working != done, so i guess i screwed up
mncharity Is anyone around who understands kp6 internals? Specifically, a number of nodes in KindaPerl6/Ast.pm have fields with different types in their has() declaration, and their "method attribs" hash. Eg, Lit::Object's @.fields vs %.fields, and Decl's $.var vs @.var. Anyone have any idea what is going on? 17:08
(roundtrip-tests++)
Limbic_Region: [particle]: ah, ok. perhaps 'longest-token autolexer' vs 'STD5 using longest-token autolexer'? the first working, the second being worked on? 17:10
@seen fglock
lambdabot I haven't seen fglock.
TimToady the autolexer is pretty much working well, but the parser still has a number of kinks to work out. I may be running into a yaml bug or two as well. 17:11
there are large parts of STD that nobody has ever run on any platform, so of course there are bugs and thinkos 17:12
the leveraging thing here is that many of them can be fixed all at once by tweaking the p6-to-p5 translation when there is systemantic error 17:13
such as feeding the wrong arguments to some primitive
there are also still some spots where the parser feeds the wrong information to the autolexer
these show up as matching OOPS instead of the null string because TRE coredumps when it tries to match the null string 17:14
also, still haven't quite finished the transformation of a* b to incorporate b's longest tokens into the implicit alternation of (a|) implied by * 17:15
what does seem quite solid is the mechanism for determining which parts of the DFA matched, and what fate that maps to 17:16
and using that fate to call into recursive descent knowing which alternatives to take 17:17
but basically, yes, finishing the autolexer does not imply that STD5 works very well 17:20
it does imply that we can make it blow up much more quicly and consistently though :) 17:21
[particle] that's important.
mncharity lol
TimToady yes, when pugs was taking 10 minutes to compile one autolexer, it was very hard to get into a debuggin rhythm
mncharity also means translating the autolexer into p6 is becoming plausible... :) 17:22
TimToady well, I had a version in p6, which the p5 version is a translation of
so translating back won't be terribly difficult
the bigger problem for someone wanting to do autolexing in a parser is that they have to be able to call pass control back and forth between the parser and the autolexer in much the same way that STD5.pm and Cursor5.pm do 17:24
so I don't now how much work it would be to retrofit, say, PGE with it
[particle] pge uses coroutines now
TimToady especially since PGE doesn't even use methods yet
yes, philosophically it's quite different from the lazy gather/take approach I used 17:25
especially since the cursor objects are considered immutable 17:26
and mutable match objects are constructed only when needed for the user's point of view
[particle] yes, it's quite different.
i'm sure patrick will come up with something brilliant, as usual.
TimToady in all likelihood 17:27
17:30 jhorwitz__ joined, jhorwitz__ is now known as jhorwitz
[particle] is trying to decide whether Cursor5 is easier or harder to read than .l files 17:32
TimToady do "cheat STD.pm | gimme5 >STD5.pm" and then try to read that file :)
17:33 Lorn__ joined
[particle] more proof that perl 5 is write-only 17:33
17:33 justatheory joined
TimToady looking at STD5's expansion of even very simple regexen, it's really amazing that it runs as fast as it does 17:34
and, of course, debugging it is a complete brain pretzel
I've even had to resort to "perl -d" at times, which is a rarity for me :) 17:35
[particle] Cursor.pm is easier to read than .l, for sure
TimToady (despite having written the debugger myself in the first place)
[particle] heh.
17:35 blindfish joined, zamolxes left
TimToady I should probably have been tweaking Cursor.pm as I discover thinkos in Cursor5.pm, but I haven't been--didn't occur to me 17:36
[particle] looks like i need to fix my pugs dev env to include perl 5.01
TimToady I'll have to review the checkin logs at some point on Cursor5.pm, I guess
I use // freely 17:37
so late 5.9 would probably work too :)
[particle] easy enough, just change one config file and fire up another terminal
TimToady now all I have to do is rewrite Perl 5 to use Cursor objects :) 17:38
and then p5 can have regexen that automatically translate dfa-able prefixes to dfa matching :) 17:39
that conveniently say OOPS where it won't work. :)
17:39 Patterner left 17:40 Psyche^ joined, Psyche^ is now known as Patterner
[particle] hrmm, looks like i need to debug gimme5 17:42
Can't create tmpyaml/Perl::TOP.yml: Invalid argument at gimme5 line 375, <> line 1.
ah, it doesn't like the colons 17:44
TimToady error message is wrong, should say yamlg5 instead of tmpyaml
you'll also need to create a lex/ dir 17:45
I don't think it does that like it does yamlg5
[particle] i'll look into it 17:46
17:46 eternaleye left 17:48 jhorwitz_ left
TimToady also, if you want to set up any makely apparatus around the whole thing, feel free 17:48
[particle] when is the lex/ dir needed? the command seems to have succeeded 17:49
TimToady lex is used when you run STD5 17:50
well, I don't run STD5 directly, I run "try" that sez "use STD5;"
but it starts using lex/ as soon as you call into any rule in the parser 17:51
[particle] wow, this is some experiment in nested map!
ok, i mangleh :: to -- in filenames 17:52
TimToady I will paste a copy of my "try"
[particle] ...but obviously that'll cause problems later, haven't found where yet
pasteling "TimToady" at 71.139.15.175 pasted "current copy of "try"" (9 lines, 153B) at sial.org/pbot/30321
TimToady oh, are you on windows? 17:53
[particle] i are
TimToady then yeah, we can't use :
sorry 'bout that
[particle] no worries. should be an easy fix 17:54
pasteling "[particle]" at 24.19.3.148 pasted "gimme5 win fix" (16 lines, 662B) at sial.org/pbot/30322 17:55
mncharity easy fix for windows problems: www.ubuntu.com/getubuntu/download 17:58
lambdabot Title: Download Ubuntu | Ubuntu
[particle] hrmm, can't locate re/engine/TRE.pm
i run ubuntu, kubuntu, centos, fedora, haiku, and windows 17:59
mncharity :)
[particle] sometimes simultaneously 18:00
ruoso goodbye-microsoft.com/ 18:02
lambdabot Title: Say goodbye to Microsoft. Now.
18:03 tobeya joined
[particle] gah, i've got to update re::engine::TRE now :( 18:09
TimToady you need to be sure to get 0.02 of that, not 0.01 18:11
which perl are you using? and can you compile modules for it? 18:12
[particle] i'm using AS 5.010
i can compile modules
18:13 pack|eet joined
[particle] there's a number of deprecated functions used in TRE, strcpy, strncpy, etc 18:13
also, can't open 'alloca.h'
so, i've got to dig
avar: ping # is there an svn repo for re::engine::TRE?
mncharity err, with so many linuxen running... why sweat trying to get windows working...? 18:14
[particle] windows support is important to me 18:15
mncharity ok
TimToady long term we might have to fork tre and maintain our own version
maybe we can get permission to relicense under A2
or write our own tre-clone, which is likely harder
[particle] TimToady: looks like TRE-0.03 is out. you may try upgrading 18:16
TimToady re-engine-TRE?
18:16 rindolf joined
[particle] yep 18:16
search.cpan.org/~avar/re-engine-TRE-0.03/
lambdabot Title: &#198;var Arnfj&#246;r&#240; Bjarmason / re-engine-TRE-0.03 - search.cpan.org
[particle] tre days ago.
TimToady hmm, I wonder what chnaged...
mncharity I just note for those watching, the pain being experienced is an artifact of pursuing windows support, rather being inherent in getting started with STD5. 18:17
[particle] search.cpan.org/diff?from=re-engine...e-TRE-0.03
lambdabot Title: The CPAN Search Site - search.cpan.org, tinyurl.com/yrluyn
[particle] no longer requires libpcre
that's a very small change :)
18:18 justatheory left
TimToady ah yes 18:18
18:18 njbartlett joined
TimToady I made that change in my own personal copy, and forgot 18:18
18:19 simcop2387 left
ruoso beer & 18:20
TimToady I though it was girls and booze :) 18:21
*thought
ruoso heh
18:22 ruoso left
mncharity Sigh. If anyone would like to really help out with a small well-defined p5 task, here's one: kp6 -Cast currently returns an ast one can't actually eval(). find and change the dumper arguments so one can. eg: $ pwd # v6/v6-KindaPerl6 $ echo '\()'|./script/kp6 -Cast # output shouldn't include '...', as in "::Sub(...)". 18:24
mncharity bogs down, chased off the road by yak in need of shaving. goes to get lunch. 18:27
18:30 Lorn left, Lorn__ is now known as Lorn, Lorn__ joined 18:36 riffraff joined
[particle] hrmm, no 5.10 package for ubuntu? 18:44
rindolf [particle]: not even in Hardy? 19:03
19:04 drrho left
TimToady not fedora either 19:16
19:35 peeps[work] joined 19:36 riffraff left, riffraff joined 19:41 simcop2387 joined
rindolf Mandriva has 5.10.0. \o/ 19:42
19:45 barney left
obra mncharity: feelings on tonight? 19:48
mncharity obra: if it's easy for you, I'd still find it useful. though I suspect an hour is way overkill. 19:50
obra Would pushing it earlier work for you? 19:51
mncharity i'm only 15 min away, so flexible timing is fine 19:52
obra 3?
mncharity re earler, sure, whenever
:) sure. I'll head over now.
obra a few minutes after is fine ;) 19:53
mncharity spinclad: looks like we've rescheduled to 3 pm. I'd enjoy getting together any time to brainstorm. MIT has lots of nice whiteboards. :)
19:53 lichtkind joined
obra (I'm at the diesel but will head over in a couple moments. 19:53
mncharity pity the diesel doesn't have wb's.
_that_ would be a nice coffee house. 19:54
rakudo_svn r25994 | pmichaud++ | [rakudo]:
r25994 | pmichaud++ | * Get postcircumfix < > to use <quote_expression> instead of <anglewords>
r25994 | pmichaud++ | (especially since quote_expression already knows how to handle
r25994 | pmichaud++ | splits and other coding issues.
pmichaud TimToady: any specific features or items you'd like me to be sure and hit in my Perl 6 talk at FOSDEM tomorrow? 19:55
19:57 fridim_ left
spinclad mncharity: sure thing. there's no way i would get there today in less than an hour, by which time you'll be all done. have fun! 20:01
i could get to MIT anytime (pretty much) within about a half hour. 20:02
20:08 riffraff left 20:14 buchetc joined 20:18 rindolf left
buchetc I'm quite desperated and need help. I've broken my pugs environment two days ago. 20:22
make clean && perl Makefile.PL && make ends in
Setup: HsSyck.cabal:14: 'Executable' stanza starting with field 'license'
Installation failed for HsSyck at util/build_pugs.pl line 235.
20:22 Schwern left 20:23 cathyal joined
buchetc I'm ready to reinstall ubuntu, but perhaps there are better ideas available 20:23
gwern strange. what version cabal you using? 20:33
20:35 cathyal left
buchetc I've installed libghc6-cabal-devel, but no programm called cabal 20:36
gwern see what ghc-pkg list|grep Cabal gives you
wolverian (there isn't supposed to be a cabal binary, yet.) 20:37
buchetc Cabal-1.1.3, Cabal-1.1.6.2, base-2.1.1, binary-0.3, filepath-1.0,
gwern hm. 1.1. I guess you're not using ghc 6.8.x? 20:38
well, try cd'ing to thirdparty/hssyck and doing the cabal installation by hand 20:39
it might be the scripts fault
buchetc no, 6.6.1 but with this version I was able to build pugs 3 days ago
gwern yes, well, that's probably my fault
gwern did a lot of messing around with hssyck and company recently 20:40
buchetc manual installation means calling somthing like __~/dev/pugs/third-party/HsSyck$ ./Setup.hs build__ 20:42
Yeah, problem seems to be solved. I've download older version HsSyck-0.4 and are now able to run ./Setup.hs configure && build 21:08
gwern: many thx
gwern buchetc: you should've just been able to do the 'runhaskell Setup configure 21:11
with the svn one
but oh well
21:12 jhorwitz left
spinclad trying that (the ./Setup.hs one), i still get "Setup.hs: HsSyck.cabal:14: 'Executable' stanza starting with field 'license'" on configure, failing. 21:13
(but that's with 0.42, from the repo. will try reverting changes...) 21:15
gwern could you paste line 14?
spinclad License: OtherLicense
gwern or just hpaste the whole thing, I wonder if we're looking at the same thing sometimes 21:16
oh. hm/ you know, audry had that commented out.
spinclad svn r19980
gwern I wonder if maybe that constant was only added in 1.2?
try just commenting that out. cabal might complain, but it won';t be an error
spinclad will do 21:17
buchetc I'm now stuck with pugs-hsregex. Similar message ": pugs-hsregex.cabal:13: 'Executable' stanza starting with field 'build-type"
gwern build-type isn't an important field either, you can cm it
buchetc hmmm... Setup: pugs-hsregex.cabal:13: 'Executable' stanza starting with field 'cabal-version' 21:19
* away because my childs are crying & 21:23
spinclad if i leave any field uncommented after the Description: field, it complains about "'Executable' stanza starting with field <whatever>"
if i comment them all out, it complains about various things those fields would have told it. 21:24
gwern weird. I'd say personally just upgrade to cabal 1.2.3.0. I think it's backwards compatible with 6.6.2 21:25
spinclad (6.6.1 here)
gwern oh. maybe it too 21:26
spinclad i'd consider upgrading to 6.8.2, but i'd like to make it work again with 6.6.1 if possible
gwern pugs is kind of broken with 6.8.2. I followed the example diff, but never managed to make it work 21:27
spinclad ok, GHC newbie: how to upgrade?
(cabal, that is)
gwern offhand, I'd visit hackage.haskell.org/cgi-bin/hackage...al-1.2.3.0 21:28
lambdabot tinyurl.com/2eey3b
spinclad will do
meanwhile i see that the License: field used to be BSD3, i'll try that now. 21:33
gwern it's not actually BSD3, of course, but MIT
spinclad right, it was always a lie
21:39 justatheory joined
spinclad hm. back to complaining about the 'license' field itself, of course. 21:41
aha! older cabal doesn't like blank lines; they seem to mark beginning of an 'executable' stanza. commenting out... 21:43
and... "Setup.hs: cannot satisfy dependency bytestring>=0.9.0.1" -- i seem to lack bytestring entirely (ghc-pkg list doesn't mention it) 21:49
a 6.8.x thing? 21:50
buchetc I've got the same result... 21:51
spinclad looks to me i could either downgrade HsSyck or upgrade ghc
gwern spinclad: or update bytestring. I had to add in that version constraint because the bytestring api changed 21:52
and I obviously had update the code to use the new api 21:53
so using an older bytestring would break
21:54 blindfish left
spinclad i don't seem to have it at all at 6.6.1; is it a new dependency? 21:55
buchetc so I will going to install ghc 6.8.1 ( and not reinstall ubuntu :) 21:56
gwern bytestring was bundled in as part of the default install, iirc. 6.8.x kicked it out as a separate library, along with a bunch of other stuff
spinclad ok, i'll fetch it in then. 21:57
pmichaud in Perl 6, is "1E6" a valid shortcut for 1_000_000 ? Or does it have to be written "1.0E6"? STD.pm seems to indicate the latter. 22:26
TimToady fixing
22:29 TJCRI left 22:31 kyrbe joined, kyrbe left, kyrbe joined, meteorjay left 22:32 meteorjay joined 22:34 kyrbe left 22:38 BinGOs left
pugs_svnbot r19981 | lwall++ | allow 1e10 as valid numbers 22:38
diff: dev.pugscode.org/changeset/19981
lambdabot Title: Changeset 19981 - Pugs - Trac
22:38 BinGOs joined
TimToady but note that 1.e10 is a method call on 1 22:40
22:43 BinGOs left, BinGOs joined 22:46 Jamtech joined 22:48 thoughtpolice joined, BinGOs left 22:49 ruoso joined
xinming TimToady: Then, How to represent for the real 1.e10 ? 22:50
pmichaud 1.0e10, or 1e10
TimToady either 1e10 and 1.0e10 work
we actually decided to outlaw 42. several years ago for that reason 22:51
22:53 Limbic_Region joined
xinming TimToady: Ok, thanks 23:01
23:12 Jamtech left 23:13 peeps[work] left 23:14 buchetc left, silug left, dalek left, cognominal left, Eidolos left, jrockway left, nothingmuch left, Maddingue left 23:16 dalek joined 23:19 Eidolos joined, nothingmuch joined 23:20 jrockway joined 23:23 cmarcelo left 23:33 cognominal joined 23:35 alester left 23:38 lumi left, GeJ left, yahooooo left, yves left, nipotaway left, kcwu left, buu left, cookys left, ead left, pasteling left, kcwu joined 23:39 cookys joined, buu joined 23:40 GeJ joined, lumi joined 23:41 yahooooo joined 23:45 lichtkind left 23:51 yves joined 23:52 jjore-w joined, pasteling joined, nipotaway joined 23:54 ead joined 23:57 cathya joined, pack|eet left

Powered by Cro and the Raku® Programming Language.

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

Thank you!