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.
duzy hi, all 06:21
Tene Hi!
duzy :) 06:22
Tene Welcome to #perl6 ! 06:24
agentzh heh 06:26
duzy thanks, do u open this channel?
agentzh duzy: audreyt++ opened this channel in the year 2005, iirc ;)
Tene No, I don't have any particular position in this channel. I'm just friendly.
duzy :) 06:29
thank all
agentzh: how do you bold you states? 06:30
agentzh duzy: ? 06:31
duzy i am a new guy using irc, using emacs, i can see your words come out in bolded and colored font 06:33
ahmadz hi 08:06
agentzh hey, ahmadz
clkao agentzh: hola! 08:07
agentzh hi, clkao :) 08:10
fglock hi 13:54
moritz_ hi all ;) 13:58
masak hi 13:59
fglock I'm working on a workflow language for $work 14:02
it looks like cobol :P
moritz_ fglock: here on #perl6 it looks like most of your $work is writing compiler/interpreter to do something else for $work 14:04
fglock: is that your normal workflow? ;-)
fglock moritz_: yes :)
moritz_ fglock: sure looks like a cool job ;) 14:05
fglock and I can use Perl 6 for that
moritz_ that's cool 14:06
which implementation do you use?
fglock I'm mostly using v6.pm, 14:10
it is fast enough, it works with Perl 5, and regex works reasonably 14:11
pugs_svn r18837 | fglock++ | [kp6] turned the roadmap upside-down, as suggested by nferraz++ 14:26
jMCg Is it a bug or a feature, that one cannot compile pugs? 14:33
moritz_ jMCg: I wouldn't call it a feature ;) 14:34
jMCg: but usually it can be compiled, provided that you have enough RAM and all the prerequiries
fglock I compiled Pugs a few days ago, on a new computer (installed ghc and such) - no problems here 14:38
moritz_ starts a 'dpkg-buildpackage -rfakeroot' 14:39
pugs_svn r18838 | fglock++ | [kp6] v6/docs/ reorganization 14:44
jMCg moritz_: I pasted the output of a compile this morning, I can do it again. 14:47
jMCg I can do that again, [I just did an svn up 14:48
moritz_ jMCg: ah, just saw it in the backlog... looking
jMCg uuh.. fails with the same issue... 14:49
moritz_ jMCg: could you please do a 'make realclean; perl Makefile.PL; make' and post the full output?
jMCg: it could be a missing dependency that isn't caught by Makefile.PL 14:50
but I'm neither a build nor a haskell expert, so I can't promise to help you ;)
jMCg moritz_: where's the difference between make distclean and realclean?
moritz_ jMCg: dunno - make realclean always worked for em ;) 14:51
s/em/me/
pugs_svn r18839 | fglock++ | [kp6] moved some implementation notes from docs/ to misc/ 14:59
jMCg O_o 15:00
I'm doing something wrong.
Ah.. I should put () around the stuff.. 15:01
pugs_svn r18840 | fglock++ | [kp6] TODO update 15:11
moritz_ d'oh, even more TODO ;) 15:13
jMCg Ah.. right. I copied the stuff to a server... now maybe I should've made it accessible as well ~_~ 16:12
jMCg brainsware.org/jmcg/pugs.make.output 16:16
moritz_ jMCg: what kind of unix/linux is this? 16:18
moritz_ you need ghc6-network, ghc6-mtl and ghc6-time 16:19
the development packages of each
jMCg moritz_: this is ArchLinux, I'm just checking how it's been built, so I can tell you if those are included. 16:22
jMCg considers rebuilding ghc 16:23
Or not.
moritz_: those are all available. 16:24
ghc-pkg: cannot find package pugs-hsregex 16:25
*** Finished building dependencies.
That's quite a definite error to me.
moritz_ aye 16:28
but I don't know how to fix it
jMCg Setup: Error: Could not find module: RRegex with any suffix: ["hi"] 16:30
Maaaaaybe.. that's just a syntax error.. 16:31
moritz_ jMCg: you could try to send a mail to [email@hidden.address] describing your problem... maybe a lambdacamel still reads it and has a good idea
jMCg moritz_: I shall check in a few days, maybe it's fixed until then.. right now I'm actually quite busy, and doing this only as a distraction while waiting for things to finish left and right, so I can pick them up. 16:36
moritz_ jMCg: ok
jMCg I wouldn't want to fire anything up prematurely, before not having researched it properely
pmurias jMCg: do you have cabal package bytestring? 16:47
jMCg pmurias: what's 'cabal package bytestring'? 16:49
pmurias i'm not sure ;)
pugs.make.output suggest you have hidden cabal packages 16:50
Could not find module `Data.ByteString.Char8':
it is a member of package bytestring-0.9.0.1, which is hidden
jMCg pugs.make.out* sounds perverted, I just realized....
pmurias see www.haskell.org/cabal/proposal/index.html 16:51
Aankhen`` Making out sounds perverted to you? o_O
pmurias . Exposed packages and hidden packages
jMCg 29 import qualified Data.ByteString.Char8 as Buf
Aankhen``: making out with pugs does. 16:52
Aankhen`` Fair enough.
pmurias jMCg: that's the line which is causing the error 16:53
jMCg pmurias: great thanks. 16:54
pmurias you are welcome ;) 16:55
jMCg Do I have to build this stuff as root? 17:01
moritz_ no
jMCg Right, shouldn't be, it supplies a prefix.. 17:02
moritz_ it might be that you need to be root for installing 17:03
spinclad fglock, other kp6camels: i'm doing some rework to [kp6]'s Makefile.PL and script/kp6, to allow a multi-staged bootstrap like gcc's build does. 17:42
moritz_ YaY
spinclad that is, have your previous version in one directory (say, stage0, where we use compiled now), and compile to stage1/; 17:43
fglock spinclad: how does it works?
fglock spinclad: cool - have you seen mp6 bootstrap script? (maybe change that too) 17:44
spinclad ... and compile to stage1 (where we compile back to compiled/ now, stepping on our compiler for make -j);
fglock re mp6: v6-MiniPerl6/util/bootstrap-perl5.sh 17:45
spinclad then you could compile from stage1/ to stage2/, and make sure they are identical, before updating changed files in your master copy.
fglock YaY indeed :) 17:46
spinclad i haven't looked at it yet; changing it similarly sounds a good plan; and likewise for onion/garlick/shallot/whateveritis 17:47
moritz_ onion/garlick/shallot sounds like "veg" or "spice" ;)
spinclad seasoning 17:48
maybe
moritz_ the compiler named seasoning ;)
spinclad after all, natural alium knows how to bootstrap itself from season to season... 17:49
so ours should too
spinclad i'm still contemplating how to rework the Makefile so you could just 'make stage1 && make stage2 && make compare && make registerchanges' 17:52
but i've got changes to script/kp6, giving it a new argument '--stage=stage0', 17:54
moritz_ spinclad: when you comit your changes, please /msg or ping me so that I can update the library path for evalbot 17:55
spinclad and a new env var KP6_STAGE=stage1 for Makefile.PL .
spinclad we could still use compiled/ for the master copies, writing them back there, in which case evalbot shouldn't need a change. 17:57
but i still have to put together some compare and write-back scripting (for the Makefile, eventually). 17:58
moritz_ spinclad: but only if kp6 bootstraps itself successfully, right?
spinclad that would be good practice: 'make stage1 stage2 compare test write-back' or such 17:59
with a 'make bootstrap' that depends on that sequence maybe. 18:00
moritz_ sounds good 18:02
spinclad so this has been a heads-up and request for comments. i've been trying it out in a playpen/ directory, and i'll try to make sure the process is working sensibly before checking things in. 18:02
moritz_ spinclad++ 18:04
spinclad ty :) 18:13
pmurias spinclad: stage2+ should be indentical, as the previous version is allowed to give different results then the new current one 18:14
s/new current/new/ 18:16
or do you mean comparing test results 18:17
spinclad no, the compiled compiler-library files. stage1 and stage2 might differ, but stage2 and stage3 should be identical (almost always), since stage1,2,3 should all be the new compiler (compiled with previous stages). 18:20
so i agree, stage2+ should have reached a fixed point, which can then run the test suite as a final check. 18:21
pmurias spinclad: it might be hard to try it out now as kp6 dosn't correctly compile itself 18:24
spinclad i've tried it out so far with KP6-MP6, which worked (converged, anyway). haven't tried KP6-KP6 yet; though KP6-MP6 at least compiles its runtime with its kp6. 18:27
pmurias fglock: using ExtractRuleBlock might be used for the delimiters inside code block problem with perl5rx 18:31
s/using//
fglock pmurias: yes, it's a possible solution 18:32
I was trying to fix it in the perl5rx emitter
spinclad: kp6-kp6 doesn't work yet 18:33
but you can leave the solution in place, for when it works
spinclad i'm hoping this may help kp6-kp6 converge, too 18:34
spinclad (in fact that's my prime motive for this) 18:45
jMCg made an update to brainsware.org/jmcg/pugs.make.output adding the output of ghc-pkg list 19:17

Powered by Cro and the Raku® Programming Language.

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

Thank you!