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.
diakopter maybe line endings were changed. 00:02
Limbic_Region ah, that could be 00:04
unobe Limbic_Region: svn diff -r PREV:HEAD works for me 00:12
is that just a shortcut for -r PREV? 00:13
nope, it's different
00:13 nperez joined
Limbic_Region unobe - I am only interested in a specific file 00:20
unobe unobe: then just append the filename: svn diff -r PREV:HEAD filename
hehe oops
Limbic_Region: then just append the filename: svn diff -r PREV:HEAD filename
Limbic_Region right
svn diff file -rPREV is shorter however 00:21
unobe ah, well -r PREV will show the current changes that aren't versioned yet
Limbic_Region I had just checked out HEAD though
unobe o
Limbic_Region but of course you are right
[19:56] <Limbic_Region> I just checked out and saw that it had been modified in the current version
unobe oops, my bad 00:22
sorry
unobe needs to read instead of scan
diakopter uses of 'scan' have shifted from antonyms of 'skim' to synonyms of 'skim' 00:23
00:24 Schwern joined, nekokak joined
unobe hmm...sorta like peruse then 00:28
00:30 aindilis joined
aindilis ls 00:41
diakopter I'm here 00:45
Aankhen`` . .. MAKE_MONEY_NOW.EXE 00:49
daxim @tell juerd who renamed unitut to unifaq in 5.9.5? 00:52
lambdabot Consider it noted.
diakopter Aankhen``: lol 00:55
00:58 nipotaway is now known as nipotan 01:13 BooK_ joined
ingy hi Schwern 01:20
stevan_ hey ingy 01:22
ingy hiya stevan_ 01:23
stevan_ ingy: so what is PMIR?
buu ingy: Did you want something?
ingy stevan_: PMIR is a kind of preinstalled CPAN over svn 01:24
stevan_ no shit
hmm
ingy buu: yes. I want to host pmir.org on your box :)
buu Uhoh.
What do you need? 01:25
ingy just for bootstrapping...
buu: just a couple apache restarts :)
buu And dns I suppose.
ingy no I got that handled
just like all my stuff, it's all in my homedir 01:26
stevan_: I'm using YAML descriptions to gen big/smart Makefiles 01:27
Limbic_Region yawns
ingy stevan_: where did you see PMIR anyways?
stevan_ ingy: cpan recent
stevan_ is reading the ingy-blog right ow
whats up with the wrist jewlery
stevan_ is gonna skip to the bottom
ingy my wrist is all operated on 01:28
that blog needs an update
I'm just finally getting off the slowdown drugs
stevan_ what happened ? 01:29
ingy stevan_: yeah I posted a placeholder module. as you can see...
Limbic_Region ingy go fast, ingy go boom
ingy hi Limbic_Region
stevan_ wow .,.. shit man,.. I been outta touch 01:30
ingy: glad you are okay,.. a belated get-well :)
ingy thanks bro
just need to get my wrist working over the next many months
stevan_ practice practice practice,.. just try not to go blind or grow too much hair ;) 01:31
I have to say,.. thats one badass coffee cup holder :)
you should get a removable model
ingy when I can play my strat again, it's all good 01:32
stevan_ nice
uhm.,.. wait,.. use your teeth there hendrix
ingy stevan_: I had that remove with a hand-drill and no anesthetic
stevan_ makes a mental note to walk when in Taiwan 01:33
ingy i rode that sme highway every day for months
I just got too wild
it won't happen again
01:39 weinig joined 01:47 mako132_ joined
dmwaters {global notice} Good day folks, we have some maintenence about to happen to 2 rotation servers. These boxes have been out of rotation for a while but we'll still lose some users. 02:10
02:14 dduncan joined 02:18 sunnavy joined
Schwern ingy: So you're tiling? That mean you're staying in Seattle? 02:22
02:23 nothingmuch joined 02:24 TreyHarris joined 02:31 thepler joined 02:33 Alias_ joined
diakopter Alias_: hi 02:35
Alias_ heya 02:36
02:39 clkao joined, gugod joined, Caelum joined, ft joined 02:40 offby1 joined 02:44 Tene joined 02:47 drupek12 joined 02:49 ft joined 02:50 drupek12 joined 02:53 Tene joined 03:03 bonesss joined 03:04 offby1 joined 03:07 rashakil joined 03:11 stevan_ joined 03:14 dolmans joined, mako132_ joined 03:53 drupek12 joined 04:00 DarkWolf84 joined 04:12 Alias_ joined 04:45 amnesiac joined 04:46 mtve joined 04:55 Alias__ joined 04:59 monomorph joined 05:02 jiing joined
pugs_svnbot r16987 | Darren_Duncan++ | trunk: 05:04
: ext/Muldis-DB/ : this is the Perl 6 equivalent of what will end up on CPAN as Muldis::DB version 0.1.0 for Perl 5
diff: dev.pugscode.org/changeset/16987
lambdabot Title: Changeset 16987 - Pugs - Trac
05:04 SubStack joined, justatheory joined 05:05 weinig joined 05:13 devogon joined
diakopter dduncan: have you excluded Muldis-DB from the pugs distribution builder? 05:25
merely curious
dduncan what do you mean?
build-wise, it is treated like all the other ext/
if you are thinking it should be excluded because I do separate CPAN uploads, then that isn't right 05:26
the Muldis-DB in the Pugs repo is written in Perl 6 ... the one on CPAN is written in Perl 5 ... they are different
that said, at such time it becomes normal to distribute Perl 6 code separately from Pugs or Parrot et al, then I will do that with Muldis-DB for Perl 6 05:27
but it isn't normal yet
diakopter, does that answer your question?
until further notice, MDB for Perl 6 is shipped with Pugs 05:28
diakopter heh; ok. yes, that answers my question :) 05:29
dduncan on that note, since a lot of the pod is identical for the Perl 5 and 6 versions, I may split that into its own distro as well ... I"m thinking mainly of Language.pod and friends ... which are to Muldis-DB proper what the Perl 6 synopsis are to Pugs or Parrot 05:30
but for now they are kept together to reduce certain matters of complexity ... 05:31
probably a good tipping point might be if I get around to implementing Muldis D as a parrot-hosted language
diakopter do they fulfill the same test suite? 05:32
dduncan fyi, Muldis D has 2 forms, called Concrete and Abstract ... the former is identical regardless of implementation, some details of the latter change depending on the implementation ... the test suite is currently written to the latter, so it differs between impls 05:33
but yes, it is the same test suite, aside from perl 5 vs perl 6 syntax
eventually there'll be a Concrete test suite, which can then be shared 05:34
diakopter ponders wordless notions of some sort 05:37
dduncan to further explain, Concrete Muldis D is for all intents and purposes the *normal* programming language, like Perl 6 itself 05:38
Abstract Muldis D is what I call the meta-model by which Muldis D code can produce or update new Muldis D code ...
and it is also the best thing to use for bridging if Muldis D is embedded in a host language, like Perl 05:39
eg, if Perl code defines Muldis D code using the Abstract form, then various native data types can carry over without conversion, eg numbers stay numbers 05:40
whereas if only the Concrete form was used ... which is analagous to how SQL is used, then any data would have to be converted to character strings and back again ... bind parameters aside
diakopter thinks he might have to RTFM 05:41
dduncan if you prefer HTML, wait a few hours for v0.1.0 to reach search.cpan.org et al ... 0.1.0 adds a lot of what I just said, which wasn't there previously 05:42
actually, nevermind
its already there
thankfully, pause et al is much faster now than a few years ago
docs be at search.cpan.org/dist/Muldis-DB/ as html 05:43
lambdabot Title: Darren Duncan / Muldis-DB - search.cpan.org
diakopter cool
dduncan note that any .pod files are identical in the perl 5/6 versions
only the pod in the .pm files may differ, and currently that is maybe 90% the same
diakopter I don't prefer reading HTML over POD, but I do like how my browser renders the HTML. 05:44
dduncan that's what I mean
diakopter I know :P
pugs_svnbot r16988 | moritz++ | trunk: 06:14
: [perlhint] a few more captials in the data
diff: dev.pugscode.org/changeset/16988
lambdabot Title: Changeset 16988 - Pugs - Trac
moritz moritz.faui2k3.org/perlhint/?q=* <-- *very* first prototype, nearly no data by now
lambdabot Title: Perlhint: *
pasteling "evalbot_r16987" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26232 06:16
06:22 vsoni joined
moritz sorry, the right URL is moritz.faui2k3.org/perlhints/?q=* 06:23
lambdabot Title: Perlhint: *
06:24 vsoni left
Alias_ purl, seen ask? 06:24
er
wrong channel :)
moritz svn sucks 06:29
I svn mv'ed a directory, and now on commit it complains that the old dir is missing :(
svn--
wow, a 'svn revert --recursive' delete the .svn directory, now no updates/commits are possible 06:35
06:39 rho joined
pugs_svnbot r16989 | moritz++ | trunk: 06:41
: renamed 'perlhint' to 'perlhints' - no more renames soonish, I promise ;)
diff: dev.pugscode.org/changeset/16989
lambdabot Title: Changeset 16989 - Pugs - Trac
06:47 Schwern joined, Eidolos_ joined 06:49 iblechbot joined 07:05 elmex joined 07:16 dduncan left 07:19 jiing joined 07:25 jisom joined
meppl good morning 07:25
07:43 newbie joined 07:47 zamolxes joined 08:11 chris2 joined 08:12 pmurias joined 08:29 goloo joined 08:39 xinming joined 08:57 |Jedai| joined 09:08 pnu joined
moritz moin 09:11
bloonix moin moritz :)
pugs_svnbot r16990 | pmurias++ | trunk: 09:15
: kp6:
: added sleep
: a basic draw a blue rect sdl script works
diff: dev.pugscode.org/changeset/16990
lambdabot Title: Changeset 16990 - Pugs - Trac
pasteling "evalbot_r16989" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26236 09:16
09:25 andara joined 09:48 xinming joined 09:54 ruoso joined 09:56 hkb_lab joined
pugs_svnbot r16991 | pmurias++ | trunk: 10:24
: missing file & uncommited change
diff: dev.pugscode.org/changeset/16991
lambdabot Title: Changeset 16991 - Pugs - Trac
10:30 Casan joined 10:33 lichtkind joined 10:38 ruoso joined 10:44 buubot joined 10:49 masak joined 11:25 nipotan is now known as nipotaway 11:27 meppl joined, scw joined 11:31 trippeh joined, xern joined
moritz ping * 11:31
this being a perl channel, maybe I should ping .* ;-) 11:34
11:34 cmarcelo joined
moritz basically, I just want to hear opinions on how to annote STD.pm 11:34
what POD tags to use, and if I should use URIs and the like 11:35
pmurias moritz: pong $moritz 11:42
moritz pmurias: do you read perl6-users? 11:43
11:48 fglock joined
pmurias fglock: hi! 11:48
fglock pmurias: hi
pmurias moritz: just looked at www.nntp.perl.org/group/perl.perl6.users/ 11:49
lambdabot Title: perl.perl6.users archive - nntp.perl.org
pmurias fglock: what is missing for bootstraping kp6? 11:50
fglock since kp6 is written in mp6, all you need is to emit mp6-level code 11:51
which is not much interesting 11:52
you can almost reuse the mp6-perl5 emitter
moritz can kp6 already compile all of mp6?
fglock sure
pmurias added a bit Time::HiRes::time's and the only thing taking noticable time while compiling kp6 with itself is the parsing 11:53
* a bit of
moritz I wasn't sure because you mumbled something about "missing runtime" every so often ;)
fglock moritz: that's for running Perl6; mp6 is simple enough
moritz fglock: ok 11:54
fglock and kp6 is finally entering the develop-the-runtime phase :)
pmurias fglock: is the perl5rx emitter ratchety? 11:56
fglock pmurias: developing a mp6-on-p5rx backend for kp6 would probably provide a faster kp6 compilation
pmurias: p5rx is backtracking by default; there are no :ratchet switch processing yet 11:57
pmurias: I'd rather proceed with the project, instead of optimize 11:59
because we can optimize p6-in-p6, instead of p6-in-mp6
pmurias is mp6 a kp6 subset? 12:04
mp6 and kp6 the languages (not implementations) 12:05
fglock mp6 is a Perl6 subset; kp6 is not a separate language, it's a Perl 6 compiler
meant to be
moritz fglock: will kp6 always be in mp6, or will you at some point use more advanced p6 features? (with a bootstrapped kp6) 12:07
fglock mp6 is an mp6 implementation written in mp6; kp6 is a Perl 6 written in mp6; 6-in-6 is written in Perl 6
moritz: that will be called 6-in-6
moritz ok, that answers my question ;)
fglock TimToady is writing a 6-in-6 parser; kp6 is trying to catch up 12:08
pmurias fglock: re "can kp6 already compile all of mp6", so it's just bugs stoping kp6 from compiling itself? 12:09
masak fglock: thx, that was the cleares explanation of <mp6 kp6 6-in-6> I've seen so far 12:13
12:15 fglock_ joined
fglock_ the kp6 runtime and algorithms are reusable for 6-in-6; so the actual 6-in-6 final development is expected to be fast, and will likely focus on performance 12:16
masak fglock_: maybe you answered this already, but what's the easiest way to help you along with kp6? 12:18
write tests?
fglock_ masak: there is a project plan: svn.pugscode.org/pugs/v6/docs/kp6-roadmap.pod 12:20
one second, I'll commit an update
12:20 Blwood joined
pugs_svnbot r16992 | fglock++ | trunk: 12:21
: [kp6] v6/docs/kp6-roadmap.pod - phase 4 placeholder
diff: dev.pugscode.org/changeset/16992
lambdabot Title: Changeset 16992 - Pugs - Trac
12:21 Blwood joined
masak fglock_: yes, kudos on the project plan, it's impressively current/comprehensive 12:22
fglock_ masak: discussing the plan is a great way to help
masak fglock_: oki. I will read it carefully and get back to you
moritz updates the html version of the project plan 12:23
fglock_ dead-ends are very time-consuming, and actual development has been quite fast
pasteling "evalbot_r16991" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26238 12:25
fglock_ brb # coffee 12:26
moritz p6l # irc log test 12:30
masak moritz: cool. maybe even a link to it? 12:34
moritz masak: that'swhat I'm implementing atm (if you mean a link to p6l)
masak moritz: yes 12:35
moritz++ # making irc.pugscode.org perl6-specific
12:35 Blwood joined
pugs_svnbot r16993 | fglock++ | trunk: 12:42
: [kp6] v6/docs/kp6-roadmap.pod - added links to POD and HTML versions
diff: dev.pugscode.org/changeset/16993
lambdabot Title: Changeset 16993 - Pugs - Trac
12:44 TJCRI joined
moritz ok, implemented ;) 12:45
pugs_svnbot r16994 | moritz++ | trunk:
: [irclog]
: * fixed a lot of whitespaces
: * added a list of words that are turned into links when mentioned verbatim
: (like p6l -> www.nntp.perl.org/group/perl.perl6.language/)
: needs some refactoring though, because it contains duplicate code with
lambdabot Title: perl.perl6.language archive - nntp.perl.org
pugs_svnbot : Aankhen++'s abbreviation parsing.
diff: dev.pugscode.org/changeset/16994
lambdabot Title: Changeset 16994 - Pugs - Trac 12:46
moritz hm, maybe the abbr.dat and links.dat should be channel specific 12:50
pugs_svnbot r16995 | moritz++ | trunk:
: [irclog] deleted p6l and p6c from abbr.dat since they are links now
: and thus never reached
diff: dev.pugscode.org/changeset/16995
lambdabot Title: Changeset 16995 - Pugs - Trac
pmurias from a quick skim of the s12, simple multis would require only type narrownes comparision from the meta model 12:51
fglock_ wonders how to implement Signature, Capture, and Match in plain mp6 13:00
I'll try something and see if it works
hmm - I'd better finish the plan first - this is too distracting 13:02
pmurias: there is an old prototype for "Code" in lib/KindaPerl6/Runtime/Perl5/Code.pm 13:05
this could probably be reimplemented in mp6;
one problem is the access to Pad, but this can be implemented as a nativa subroutine call 13:06
native
masak hm, the multi method call order never was completely worked out, was it?
fglock_ no, we need to reinvent :)
masak fglock_: not sure I understand what 'reinvent' has to do with the spec not being ready... 13:08
fglock_ masak: we implement it, and have the implementation be specced, than reimplement if necessary 13:09
masak fglock_: ah.
fglock_ and then the new implementation brings new ideas, which are then re-specced 13:10
masak the way I understand it, the problem with speccing it was that nobody could even concieve of a good spec. so this is probably a good idea
(implement <-> spec)++
13:11 Boolahman joined
pmurias masak: it was quite worked out 13:11
s12:line 905 13:12
masak pmurias: hm, yes. but I remember at least TheDamian voting against the MMD proposal and advocating simplicity
moritz: does 's12' become a link, too? :) 13:13
moritz masak: atm only S12:905 becomes a link...
13:13 Boolahman joined
pmurias all the ';' stuff was delegated for later 13:13
moritz masak: but if you want it, takes just two minutes to implement ;) 13:14
masak moritz: no, S12:905 becoming a link is cool too
moritz: was just wondering if you'd considered it at all
moritz masak: I did consider it, and until now I'm not entirely sure if I should do it or not ;) 13:15
13:16 agentzh joined
pmurias we could probably assume type narrownes meaning subclassing relationship for now 13:16
without ';' the MMD is IMHO is really simple to implement so the AST & grammar parts would be the difficult ones 13:17
haircut& 13:19
fglock_ pmurias: the AST is just 'Signature' and 'Capture' 13:21
pugs_svnbot r16996 | moritz++ | trunk:
: [irclog] synopsis links with line numbers are now processed very early, S\\d\\d
: for which synopsis exists are now turned into links as well
diff: dev.pugscode.org/changeset/16996
lambdabot Title: Changeset 16996 - Pugs - Trac
fglock_ and annotations
moritz synopsis links should now work for upper case S02 13:22
perl++ # making irc logs such a fun ;)
[particle] which logs show this? 13:23
moritz irc.pugscode.org 13:24
[particle] nice. 13:25
if you're in the mood for adding more, processing S05:546-549 would be grand. as well as S05/"Nothing is illegal" 13:29
agentzh [particle]: S05:129 is already supported, no?
lambdabot agentzh: You have 1 new message. '/msg lambdabot @messages' to read it.
agentzh @messages
lambdabot moritz said 1d 5h 47m 6s ago: moritz.faui2k3.org/tmp/irclog.sql.bz2 should contain all you need
[particle] oh, i should say, *highlighting* the text and processing S05:546-549 13:30
agentzh wow...
moritz: already downloaded, will look tomorrow :)
moritz: thanks :)
13:30 jiing joined
moritz agentzh: you're welcome 13:31
[particle]: I'll start a TODO list ;)
[particle] moritz++, this is cool stuff. todo++ too :)
agentzh you're discussing a Perl 6 regex annotator ?
agentzh backlogs. 13:32
moritz are we? not at the moment ;)
agentzh sorry, lacking the context...
13:32 rashakil left
pmurias moritz: you couls annote STD.pm with the sort of tags the smartlinks use 13:34
moritz pmurias: yes
[particle] STD should only be annotated by perl 6 pod and perl 6 comments imo 13:35
moritz what I'm unsure about atm is how the enclosing POD6 tags should be called
[particle] is there a pod 6 parser in perl 6 yet?
pugs_svnbot r16997 | fglock++ | trunk:
: [kp6] v6/docs/kp6-roadmap.pod - started detailing phase 4
diff: dev.pugscode.org/changeset/16997
lambdabot Title: Changeset 16997 - Pugs - Trac
fglock_ hmm - how about a kp6-based syntax highlighter 13:36
moritz [particle]: not that I know of
fglock_: that what be way cool ;)
[particle] fglock_: that would rock
fglock_ I think a html-backend would do
moritz Juerd: btw the money left my account today, it should arrive at yours soonish
fglock_ oh my, I'm so easily distracted 13:37
moritz fglock_: I think you should take one intermediate step, like returning a lol like so [["keyword", "my"], ["space", " "], ["variable", '$foo']...] 13:38
fglock_: then some HTML slaves like Aankhen'' or me could write the HTML emitter ;)
[particle] [[<RO>,"TF"]] 13:40
fglock_ moritz: it's as easy as copying Visitor::EmitPerl6.pm to Visitor::EmitHTML.pm and add html tags
moritz fglock_: I'll take a look at it 13:41
fglock_ it will not resemble the source code, becuse it is re-emitting - but this can be fixed later
13:42 penk joined
fglock_ moritz: re intermediate step - agreed, but we could have a working prototype as a motivation 13:42
moritz fglock_: Aye 13:43
fglock_ kp6 currently throws away the pos() information
moritz just works on importing the old colabti irclogs
finished downloading of 2005 ;) 13:44
fglock_ moritz: I'll start it # no rush for kp6 planning :) 13:45
moritz fglock_: -Ofun is very important after all ;)
diakopter fglock_: may I ask your opinion on making 'proto' the default (implicit) routine modifier (instead of 'only')? My argument was that why would you need the 'only' modifier: to declare that a routine (sub/method/operator) can't be partially-overridden? 13:46
13:47 iblechbot joined
pmurias avoiding accidential merges is probably the reason 13:48
diakopter pmurias: an example?
fglock_ diakopter: the 'only' idea is to forbid multi-ing, isn't it?
Juerd moritz: Wonderful! 13:49
daxim: It wasn't renamed. I split perlunitut to perlunitut and perlunifaq
pmurias diakopter: take a look at en.wiktionary.org/wiki/take 13:50
diakopter fglock: yes, so it seems. but why make that the default?
fglock_ hmm - because Perl 6 is strict by default? 13:51
pmurias imagine importing several modules which each implement a multi doing one of the meanings
you would get a very strange routine then 13:52
fglock_ diakopter: but I don't really have an opinion for or against it
diakopter pmurias: how would you get a strange routine? if any of the multis' signatures collided, it'd be a compiler error 13:53
I mean, how would *6 know which of the editions with identical signatures to use? 13:54
pmurias if you have a specific routine and a general one which to unrelated things you woudn't get a error 13:55
13:55 drupek12 joined
pmurias found en.wiktionary.org/wiki/have as a better example :) 13:56
lambdabot Title: have - Wiktionary
moritz diakopter: I do hope you have the means to /op yourself again ;)
diakopter pmurias: apologies; I'm still not understanding. Perhaps a concrete example of a specific one vs a general one where both had identical sigs
moritz diakopter: a simple "yes" would have been sufficient ;) 13:57
diakopter sure, sufficient, but not efficacious toward settling all curiousity 13:58
:D
pmurias meaning To possess, own, hold with sub have(Any $thing) is diffrent from To give birth to. sub have (Child $offspring) {}
13:59 Blwood joined
diakopter pmurias: ok. in that example, those don't have identical signatures 13:59
pmurias if they would have indentical signatures you would get an compiler error with either default 14:00
* had identical
diakopter right... I wasn't claiming otherwise. 14:01
pmurias it's probably historical reason which dictate the default
diakopter my point was that why bother with the 'only' keyword as default, since it's redundant
I mean, it's either the only routine with that signature or it's not.
pmurias if you use only you forbid all others with the same short name 14:02
diakopter exactly... I'm not suggesting removing 'only' as an option.
moritz pmurias: only if exported to the same namespace, of course 14:03
pmurias if somebody else imposed proto on you withought your consent
14:03 bonesss joined
diakopter but how would *6 know whether to obey the proto or the only? 14:04
order of parsing appearance?
the proto, or the implied 'only', I mean. 14:05
pugs_svnbot r16998 | fglock++ | trunk:
: [kp6] added a very rough syntax highlighter
: $ perl kp6-perl5.pl --do EmitHTML < t/01-sanity/01-tap.t > x.html
diff: dev.pugscode.org/changeset/16998
lambdabot Title: Changeset 16998 - Pugs - Trac
fglock_ yay!
pmurias not sure
moritz fglock_++ #coding for -Ofun ;)
pmurias fglock_: nice
fglock_ code is blue, strings are green - that's all :P
moritz fglock_: no CSS? 14:06
diakopter fglock: what would it take to get kp6 to where I could set up autobuilds and evalbot for it?
moritz I can refine that later on ;)
fglock_ moritz: please do - it's a proof-of-concept :)
diakopter fglock: or is it already there?
fglock_ hmm
moritz I'm aboutto import some 800k lines into the irclogs - I don't know if it will work ;) 14:07
diakopter if you tell me it's possible, I'll go for it...
(without bugging you with questions about it, I mean) :)
fglock_ diakopter: it's possible :)
[particle] a kp6 evalbot? heh.
moritz ?kp6 3+4 14:08
lambdabot 7
moritz lambdabot: do you parse kp6?
fglock_ diakopter: kp6 has several compilation options - not sure how to handle that
diakopter moritz: er, what? :)
moritz diakopter: just /ignore me ;)
diakopter no.. :P
fglock_ but maybe --perl5rx is good enough 14:09
[particle] &kp6 :rx<perl5rx> ...
diakopter since pugs has the 'safe' filtering, I'll have to run it through pugs first... (same for the p6P(arrot) evalbot)
[particle] eh, just use a sensible default for now. 14:10
working > good
diakopter puts on the sysadmin hat again 14:11
moritz how do I construct a timestamp with p5 I have have date and time given?
s/I have/if I have/
diakopter Juerd: did you see the thing I posted about svn.pugscode.org/pugs 14:12
moritz diakopter: it looks quite cool
agentzh moritz: what's the default encoding settting in you mysql database?
diakopter moritz: I agree. probably won't be terribly useful to many people for now. 14:13
moritz agentzh: how do I find out?
diakopter although
combined with firefox' plugin to edit a textarea externally, and multi-user editors such as subthaedit, it could be way cool.
as in... 5 people working on the same file at once... then someone yells "Checkpoint!", and it gets committed. 14:14
but of course that can be done without the browser interface.
I guess previously I was imagining making a browser-multiuser-editor.
pugs_svnbot r16999 | fglock++ | XXX-HACK-SVNBOT-SUBST-0.23065042210059483:
: [kp6] highlighter - variables are orange :P
diff: dev.pugscode.org/changeset/16999
lambdabot Title: Changeset 16999 - Pugs - Trac
[particle] unless the multi-user editor is browser-based.
diakopter so nm
[particle] right.
diakopter wasn't someone mentioning an opensource multiuser editor? 14:15
fglock_ moritz: this is the command line for compiling the highlighter: $ perl mp6.pl < lib/KindaPerl6/Visitor/EmitHTML.pm > lib5/KindaPerl6/Visitor/EmitHTML.pm
diakopter or did I imagine that too.
[particle] gobby?
diakopter yeah, that.
diakopter adds it to the list of things to learn about 14:16
[particle] oh, i think there's writewith.com
diakopter I just thought that having the multiuser editor in the browser would make it easier for casual spectators to observe/onramp.
pmurias mortiz: or add it to the Makefile and call make :) 14:17
* moritz
moritz pmurias: ... hm, what? did you mean fglock_ ?
diakopter watching skilled language implementors would be quite instructive.
moritz ETOOMANYTHREADSINIRC
Tene Tag the different threads with different prefix characters! 14:18
[particle] diakopter: how about saving it as a screencast, with the ability to playback later :)
moritz Tene: actually I thought about something like that...
Tene I've done that several times before.
moritz Tene: with thread merging and grafical display in client+logs
diakopter [particle]: good point. now to convince all the committers to record all their editings... ;)
fglock_ pmurias: right :) 14:19
Tene There was a service that recorded terminal sessions for later playback... either through a web gui or on a terminal.
[particle] diakopter: if "capture for playback" is built into the editing tool, juts convince them to use the tool :)
pmurias fglock_: did it 14:20
diakopter oh. I'm pretty sure Subthaedit has that. but not everyone has a mac.
agentzh moritz: dunno. 14:21
[particle] heck, "capture" is basically "commit". "playback" could use javascript to replay the diffs char-by-char.
pugs_svnbot r17000 | pmurias++ | trunk:
: kp6: lib5/KinaPerl6/Visitor/* in the Makefile
diff: dev.pugscode.org/changeset/17000
lambdabot Title: Changeset 17000 - Pugs - Trac
agentzh moritz: the bytes in the dumpfile seems to have been malformed...
moritz agentzh: I feared that :| 14:22
agentzh moritz: afaik, both the mysql database and the mysqldump utility uses an encoding... 14:23
i see a lot of bug reports over the net regarding mysql's encoding problems...
14:23 ProperNoun joined
diakopter LOLz at [particle]'s suggestion to commit every single-char diff 14:23
pasteling "evalbot_r16998" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26240
[particle] :)
diakopter oh yeah, the pugs build 'failures'
14:24 polettix joined, jhorwitz joined
moritz agentzh: and if you let my_decode in its current form run over the lines in the sql dump - does that produce sensible output for you? 14:24
14:25 Blwood_ joined
agentzh moritz: no, not all. even utf8 chars become a mess. 14:26
the mysqldump does an encoding already?
pasteling "fglock" at 201.25.37.25 pasted "highlighted t/01-sanity/01-tap.t" (1 line, 1.1K) at sial.org/pbot/26242
agentzh *encoding convertion
14:26 Blwood__ joined
agentzh mysqldump destroys everything. 14:27
moritz agentzh: probably, yes
14:27 Coke joined
agentzh moritz: have you specified an encoding setting while calling mysqldump? 14:27
moritz agentzh: should I write a simple mysqldump using perl and DBI?
agentzh moritz: that may help and may not :) 14:28
14:28 lisppaste3 joined
moritz agentzh: then you could be sure that the data arrives exactly as IrcLog.pm sees them 14:28
Coke Can someone look at the 4 tickets at rt.perl.org/rt3/Search/Results.html...'stalled') and let me know if they are resolved, obsolete, etc?
agentzh moritz: but why not have a try?
moritz agentzh: no, I didn't
diakopter Coke: welcome
agentzh moritz: yes, it makes sense :)
moritz: be sure to use proper line breaks...while writing you own dumpers ;) 14:29
mysqldump makes my life harder while reading it using Perl ;)
very very long lines in the .sql file... 14:30
diakopter Coke: the first one is definitely long stale
moritz agentzh: let me first finish the colabti import (estimated 15min), then I'll write the dumper
agentzh moritz: no hurry :)
moritz: i'm going to bed anyway :)
moritz: see you tomorrow &
14:31 agentzh left 14:34 amnesiac joined
diakopter Coke: the 2nd one's diff is empty, to me... 14:34
Coke: and the 1st thing that guy uploaded uses tons of pasm, and isn't usable for the current languages/perl6 14:36
so I say close #1,#2
#3 can be closed 14:37
#4 uses the old P6Grammar (replaced by PGE, I thought)... so it can be closed... 14:38
14:39 pmurias left 14:40 pmurias joined, prism joined
pmurias fglock_: font tags are a thing of the past ;) 14:44
diakopter fglock_: the other reason I was suggesting making proto the default/implied is that nearly everything in a distribution of a *6 implementation will need to be proto. 14:46
thus, proto proto everywhere
14:51 unobe is now known as unobe_away
pmurias if you use the multi keyword you don't need a proto 14:51
fglock_ pmurias: I wonder if there is an existing css we could reuse 14:53
diakopter pmurias: are you suggesting all the code objects be marked 'multi'?
14:54 rlb3_work joined 14:56 daxim joined
fglock_ hey - we could have a kp6-AST highlighter too :) 14:57
[particle] fglock_: check with one of the nopaste engines
(ast highlighter)++ (ast highlighter)++ (ast highlighter)++
fglock_ [particle]: ok - but that will have to wait until after lunch 14:59
[particle] lunch++ :)
pmurias fglock_: I wonder if it is possible to have the <span class="AstNodeName">...</span> added automaticly
[particle] something has to emit it 15:00
pmurias its generic for all the AST nodes 15:01
bbs 15:02
fglock_ pmurias: sure 15:03
Juerd diakopter: Where did you post that? (No, I didn't see it)
moritz on IRC 15:04
moritz.faui2k3.org/irclog/out.pl?ch...11#id_l987
lambdabot Title: IRC log for #perl6, 2007-07-11, tinyurl.com/252nfb 15:05
moritz now imports old logs from colabti 15:07
let's see if it slows down mysql too much
15:08 unobe_away is now known as unobe 15:09 Coke left
diakopter wow; I somehow entirely missed audreyt being here yesterday 15:10
15:17 unobe is now known as unobe_away
pugs_svnbot r17001 | moritz++ | trunk: 15:21
: [irclog]
: * added TODO file (mostly with suggestions from [particle]++)
: * added links to the colabti.de irc logs
diff: dev.pugscode.org/changeset/17001
lambdabot Title: Changeset 17001 - Pugs - Trac
pugs_svnbot r17002 | moritz++ | trunk: 15:26
: [irclog]
: * added pugs_svnbot to the list of bots
: * whitespace fixes in cgi/out.pl
diff: dev.pugscode.org/changeset/17002
lambdabot Title: Changeset 17002 - Pugs - Trac
15:26 yum_ joined
pugs_svnbot r17003 | moritz++ | trunk: 15:35
: [irclog] hilighting synopsis links with ranges like S01:12-23 (untested)
diff: dev.pugscode.org/changeset/17003
lambdabot Title: Changeset 17003 - Pugs - Trac
moritz ouch, it (or something else) broke S01:123 links
pugs_svnbot r17004 | moritz++ | trunk: 15:38
: [irclog] fixed previous commit
r17005 | moritz++ | trunk:
: [irclog] real fix this time, moritz--
diff: dev.pugscode.org/changeset/17005
lambdabot Title: Changeset 17005 - Pugs - Trac
diakopter Juerd: will feather's IP address change when you VM-ize it? 15:43
pugs_svnbot r17006 | moritz++ | trunk:
: [irclog] another fix of S\\d\\d:\\d+ links, tested this time, moritz--
diff: dev.pugscode.org/changeset/17006
lambdabot Title: Changeset 17006 - Pugs - Trac
diakopter moritz++ :P 15:44
moritz anyway, now it works
both locally and on the server
15:49 |Jedai| is now known as jedai 15:50 htmldeveloper joined, wilx joined 15:51 Boolahman joined 15:53 vel joined
moritz could somebody ( diakopter ?) please kickban Boolahman ? 15:54
it's a damned spam bot
diakopter moritz: thanks for the hilight 15:55
lumi Looks like it's channeling Zippy
moritz just cleared the logs
diakopter welcome, Vaclav 15:57
welcome Veljko
pmurias fglock_: the question is how to do it
15:59 htmldeveloper left 16:07 integral joined, weinig joined 16:08 Eidolos_ is now known as Eidolo, Eidolo is now known as Eidolos
pmurias should i copy the mp6 test suit to kp6 some of it fails and it tests the runtime only as whole of it parses 16:10
?
avar pmurias: (more testing)++ I suppose 16:11
pmurias: Working on kp6 packaging?
16:13 buetow joined
fglock_ pmurias: mp6 has some grammar implementation errors; it's documented in the roadmap under 'KindaPerl6 syntax review' 16:13
16:13 ProperNoun left 16:14 offby1 joined
moritz fglock_: once I have compiled EmitHTML.pm, how do I execute it? 16:14
fglock_ $ perl kp6-perl5.pl --do EmitHTML < t/01-sanity/01-tap.t > x.html
moritz fglock_: thanks
fglock_ pmurias: so the syntax in the tests needs fixing, too 16:16
16:17 justatheory joined
fglock_ but this can be postponed to phase 5 16:17
which is to import Perl6-STD
otoh, it is forcing the developers to use wrong syntax 16:19
avar someone please upload a sample html emit 16:20
16:20 cernd left, hkb_lab is now known as hakobe
moritz avar: just a sec 16:20
avar: moritz.faui2k3.org/tmp/out.html (no <head> etc) 16:21
fglock_: in EmitHTML.pm, return '$self->{' ~ $.name ~ '}' 16:23
isn't that p5?
fglock_ yes :)
it's a bug
16:24 stevan_ joined
fglock_ moritz++ # joined the kp6 crew 16:25
moritz does mp6 do ineterpolating strings? 16:29
"<br />\\n" and the like
fglock_ no, you need to call a subroutine for "\\n" 16:31
or maybe handle it with css
moritz I want the HTML code to look OK ;) 16:32
not just the rendered ;)
fglock_ '{ package ' ~ $.name ~ '; ' ~ Main::newline()
for perl5 code, I just run through perltidy
moritz aye 16:33
there's tidy for html as well ;)
pmurias avar: packaging is an area i'm totaly ignorant in so no :) 16:38
perltidy++ #really good code formating 16:39
16:41 rintaro joined
pmurias moritz++: congrats on joining the kp6 team :) 16:41
moritz pmurias: I'm more on the HTML team ;-) 16:42
avar k 16:44
pmurias we use perltidy on the ast too 16:46
16:46 Blwood joined 16:49 kanru joined
pmurias regarding implementing Int,Str,etc. in perl6 this would push a lot of things into the emitter 16:50
16:52 bonesss joined
moritz look at moritz.faui2k3.org/tmp/out.html again - now it's with <span class="keyword">module</span> and the like, and the corresponding CSS 16:52
lambdabot Title: Auto Generated P6 code
moritz the header of the html file was manually added by me
should I commit the changes to EmitHTML.pm? 16:53
pmurias fglock: is p6-in-p6 meant to be a complete rewrite or a evolution of kp6?
moritz: yes
moritz where should I add the corresponding style.css?
[particle] moritz: as a .pm file, or as a method in the emitter 16:54
i think we need an 'axle' language 16:56
ast transformation language
16:57 elmex joined
[particle] or whatever, like xslt 16:57
maybe astt is a better name
ah, nevermind. 16:58
moritz [particle]: or "TGE"?
;)
[particle] yeah. or kp6 :)
[particle] is currently struggling with xpath 16:59
moritz I don't know what TGE exactly is about, but I have 'easy AST transformation' in the back of my mind
fglock_ pmurias: an evolution, I guess 17:01
pugs_svnbot r17007 | fglock++ | trunk:
: [kp6] Visitor::EmitAstHTML.pm - AST highlighter; pmurias++ # css idea
diff: dev.pugscode.org/changeset/17007
lambdabot Title: Changeset 17007 - Pugs - Trac
[particle] tge is a way to transform trees in parrot. it's designed for PAST, but should work with any tree structure
17:01 dduncan joined
[particle] hrmm... i suppose you could emit the ast as xml and then use existing tools to xform to html, pdf, etc 17:03
fglock_ [particle]: nothingmuch has some ideas on adding attribute grammars to perl6
[particle] ah, great!
fglock_ I don't know how to get to it - I'd appreciate if somebody could write about that 17:04
pugs_svnbot r17008 | moritz++ | trunk:
: [kp6]
: * modified EmitHTML.pm to output semantic HTML, eg <span
: class="keyword">module</span> and the like
: * added method css (which is not yet called)
diff: dev.pugscode.org/changeset/17008
lambdabot Title: Changeset 17008 - Pugs - Trac
fglock_ [particle]: is there a particular xml format for that, do you have an example? 17:08
[particle] fglock_: i don't have an example. standard tools like xslt allow you to manipulate xml and make it into html for example. 17:09
one needs to define a dtd that describes what the node hierarchy looks like 17:10
foo contains 0 or more bar... bar has optional attributes baz and quux, etc
the ast already has a well defined format 17:11
fglock_ moritz: take a look at EmitAstHTML output? I used the node class name as the span class name, is this ok?
[particle] it should be straightforward to translate that format to a matching xml representation
moritz fglock_: I thinks \\w+ is allowed as class names 17:12
pugs_svnbot r17009 | pmurias++ | trunk:
: kp6:mp6 test suit
diff: dev.pugscode.org/changeset/17009
lambdabot Title: Changeset 17009 - Pugs - Trac
pasteling "evalbot_r17008" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26245
moritz fglock_: when running kp6-perl5.pl on t/01-sanity/02-counter.t I get a syntax error
fglock_: is that normal?
fglock_ moritz: looking 17:13
pmurias moritz: yes
fglock_ moritz: only t/kp6/ tests are supposed to run 17:14
pugs_svnbot r17010 | moritz++ | trunk:
: [kp6] EmitHTML.pm: a complete HTML file (with header and CSS) is emitted
diff: dev.pugscode.org/changeset/17010
lambdabot Title: Changeset 17010 - Pugs - Trac
pmurias and t/insanity
moritz fglock_: ok
pmurias except the last one 17:15
;)
[particle] hearts the diff browser 17:16
fglock_ pmurias++ # Quality Assurance Head
pugs_svnbot r17011 | moritz++ | trunk: 17:23
: [kp6] EmitHTML.pm: more tweaks, more newlines in output
diff: dev.pugscode.org/changeset/17011
lambdabot Title: Changeset 17011 - Pugs - Trac
17:27 ujwal joined
ujwal Hi any one can help me to help :) 17:31
on pugs and perl6
diakopter ujwal: what are your skills, and what would be fun for you to do? 17:32
17:32 devogon_ joined
moritz ujwal: just ask, maybe we can help you 17:33
pugs_svnbot r17012 | fglock++ | trunk:
: [kp6] added Main::mangle_ident() for mangling '::' out of namespaces
: - EmitAstHTML now uses mangle_ident() for emitting valid css
diff: dev.pugscode.org/changeset/17012
ujwal I just checked out pugs code
lambdabot Title: Changeset 17012 - Pugs - Trac
diakopter moritz: ujwal wants to contribute
ujwal I on win32
moritz diakopter: good to know ;)
fglock_ moritz: ::Val::Buf is now Val_58__58_Buf
ujwal rx:g
is not working
my tests are not working 17:34
moritz fglock_: that should do for class="..." attributes
ujwal: what do you mean with "not working"? it's normal that some (actually quite a lot) don't work 17:35
ujwal ok
moritz that's why we are still working on different perl6 implementations ;) 17:36
ujwal #is $str.subst(rx:g:i/L/,'p'), 'heppo', '.. with multiple adverbs';
rx:g is a infinite loop 17:37
i tried it in pugs.exe interactively
moritz ujwal: which backend are you using? native haskell?
ujwal GHC 6.6.1 17:38
I have followed instruactions in the INSTALL 17:39
fglock_ moritz: we can now cross-link the highlighted code to the highlighted AST :)
s/can/could/
diakopter ?eval $str = 'foo'; $str.=subst(rx:g:i/o/,'a'); say $str; 17:41
17:41 evalbot_r16971 is now known as evalbot_r17012
evalbot_r17012 Error: No such method in class Str: &subst 17:41
pugs_svnbot r17013 | pmurias++ | trunk: 17:42
: kp6:
: $object->method -> $object->{_dispatch}($object,...} sugar
diff: dev.pugscode.org/changeset/17013
lambdabot Title: Changeset 17013 - Pugs - Trac
ujwal its line 13 from pugs\\t\\builtins\\strings\\subst.t 17:43
17:43 weinig joined
fglock_ pmurias: you could call that something like MOBridge 17:43
pmurias++ 17:44
moritz ujwal: it produces an error for me, but it doesn't loop
ujwal my compiled version is in a loop 17:45
fglock_ or Perl5Bridge, depending from which side you look
moritz ujwal: wich pugs revision are using?
appears in the output of pugs -v 17:46
ujwal svn 16996
moritz ouch, I just noticed that my last pugs built is about 100 revisions behind 17:47
fglock_ regex loops are most likely a perl5-side fault (PCR)
moritz fglock_: aye
should I @tell agentzh?
fglock_ my pugs is 2500 revisions old :P 17:48
yes!
moritz @tell agentzh there's a new regex bug most likely introduced by PCR modifications: t/builtins/strings/subst.t test 3 is looping
lambdabot Consider it noted.
moritz ujwal++
fglock_ MagicFixingGnomes++ 17:49
moritz *lol*
echo "everyhting else \\t\\t\\t\\tMagicFixingGnome" >> SLAVES 17:50
ujwal error in line 125 of pugs\\t\\data_types\\array.t
moritz ?eval my $a = "hello"; $a~~s:g:i/l/p/; $a 17:51
17:51 evalbot_r17012 is now known as evalbot_r17013
evalbot_r17013 \\"heppo" 17:51
ujwal is @array12[*-1],'e', "indexing from the end [-1]";
moritz ?eval my $a = "hello"; $a.susbst(m:g:i/l/, 'm'); $a
evalbot_r17013 Error: No such method in class Str: &susbst
moritz ?eval my $a = "hello"; $a.subst(m:g:i/l/, 'm'); $a
evalbot_r17013 Error: No such method in class Str: &subst
moritz ujwal: aye, that's a newer change that is NYI in pugs 17:52
ujwal: you're welcome to fix them, of course ;) 17:53
ujwal which one is the stable version with i can compile and tweek
moritz well, current pugs should compile
it's just the test suite that fails
ujwal ya
any pointers to understand the code 17:55
moritz dev.pugscode.org/ has a few pages on the interals 17:56
17:56 Psyche^ joined
lambdabot Title: Pugs - Trac 17:56
moritz s/interals/internals/
and there's docs/Pugs/Hack*
ujwal: you can just comment out the looping tests and run the rest 17:57
ujwal ya i did that and tests stoped line 125 of array.t
moritz sorry, gotta go % 17:58
s/%/&/ 17:59
ujwal thanks and I am it ...
s{it}{dev.pugscode.org/wiki/} 18:00
diakopter ujwal: do you have a pugs commit bit for subversion? 18:11
pugs_svnbot r17014 | pmurias++ | trunk:
: kp6: DispatchSugar: sugar_off
diff: dev.pugscode.org/changeset/17014
lambdabot Title: Changeset 17014 - Pugs - Trac
18:11 Psyche^ is now known as Patterner
pmurias fglock: i wrote DispatchSugar with the intention of using it to make MO.pm more readable 18:12
i hope it's ok if i do it
ujwal diakopter: no I don't have
diakopter ujwal: send me a private message with your desired username and an email address, please, and you'll receive the invitation 18:26
pugs_svnbot r17015 | pmurias++ | trunk:
: kp6: DispatchSugar gets test via run_tests.sh
: sprinkled a view sugar() calls in MOP.pm
diff: dev.pugscode.org/changeset/17015
lambdabot Title: Changeset 17015 - Pugs - Trac
18:31 BooK joined
diakopter ujwal: invitation sent! welcome aboard 18:33
ujwal thank u :) 18:34
18:37 rindolf joined
diakopter moritz: you still around? 18:39
18:54 buetow joined 19:03 beppu joined
moritz diakopter: for just one minute, yes 19:06
now gone for good & 19:10
nothingmuch [particle]: the basic idea i had is that an attribute grammer is kinda like a context sensitive role 19:26
they inherit other attribute grammers in the same way that roles inherit roles
each class can do an attribute grammar in a similar way that a class does a role 19:27
however, these methods are never accessible on the outside
you need to instantiate the attribute grammar over a root object
e.g. AverageDifference->new( :root($tree) );
and then you can call the attributes of the root on that 19:28
and all the methods defined in the attribute grammar instances of the classes that play part in $tree are visible within that call chain
a somewhat crude perl 5 example is available here: nothingmuch.woobling.org/MO/t/ag.t
Juerd diakopter: No, feather's IP will not change when I virtualize it. 19:29
[particle] nm: i see. so that's the semantics, and then a syntax layer needs to be defined to use it. 19:32
well, to make it sugary, anyway
nothingmuch: what is 'bestowed'? 19:33
nothingmuch it's the other side of inherited 19:34
a child has an inherited attr
so it calls it on itself
which calls back to the parent to give it the value
the instance of the method in the parent's metamodel that has that body is under that category 19:35
19:35 drupek12 joined
nothingmuch also note that parent<->child relations are *very* crude 19:35
due to lack of typing
to search for a parent the call stack is walked
and each object is asked, like in the dr. seuss book, are you my mother?
the default impl is to search inside the top level attrs of the parent for the child 19:36
and this parent <-> child relationship is cached in the attribute grammar instance
[particle] i see: "here, son, program this vcr for me to tape tonight's pbs special."
nothingmuch fror the syntax i figured something like this:
(writing nopaste) 19:37
19:37 bonesss joined
[particle] so, with roles, they're only important during construction, not runtime. what's the equivalent for an ast? is an ast always in the construction phase, so attribute grammars are always available? 19:39
nothingmuch ast? you mean ag? 19:40
fwiw, i designed the MO metamodel so that adding this functionality is a matter of a simple set of macros to decalre an AG, and some intermediate meta programming 19:41
it's fairly well supported to invent your own concept of OO
the way it's done is that the responder interface is wrapped (in my implementation, could be augmented in any other way) in a context sensitive one
which checks whether or not we are currently in the scope of an attribute grammar
and if so, it makes the extra methods visible 19:42
the other bits that had to be programmed is the code gen for bestowing attrs
the parent finding logic
and the AG merging (reuses most of the role code)
Sial.org
eep
pasteling "nothingmuch" at 84.110.255.235 pasted "ag mockup" (1 line, 9B) at sial.org/pbot/26246
nothingmuch shyte
pasteling "nothingmuch" at 84.110.255.235 pasted "ag mockup" (27 lines, 463B) at sial.org/pbot/26247
nothingmuch there we go
Tree should has $.value; 19:43
wait, 'm actually missing some more code
pugs_svnbot r17016 | pmurias++ | trunk:
: kp6: fixing kp6 back after my previous commit
: TODO() - confesses
diff: dev.pugscode.org/changeset/17016
lambdabot Title: Changeset 17016 - Pugs - Trac
[particle] do you need to specify in class Tree that global_average is an inherited method, or is it implicitly inherited and you're just putting it there for clarification. 19:45
i suppose "has $.value;" belongs there somewhere 19:47
pasteling "nothingmuch" at 84.110.255.235 pasted "ag mockup" (70 lines, 1.8K) at sial.org/pbot/26248 19:50
19:50 ujwal is now known as [ujwal]
nothingmuch here, that'w a much better example 19:50
method global_average : inherited; 19:51
that's what I figured
19:51 [ujwal] is now known as [_ujwal_]
nothingmuch [particle]: k, i have to go shower real quick 19:52
i hope that makes more sense
but what I really care about is the metaprogramming level
any syntax will do
i'll make an example of an inherited ag 19:53
later
that's really funky because you get a dual tree of inheritence
and it still works out cleanly
19:54 [_ujwal_] left
[particle] thanks, that helps. 19:55
20:13 xerox joined
awwaiid whats going on in here? 20:16
20:19 DarkWolf84 joined
fglock_ pmurias: re MO.pm more readable - I think the best approach would be to port MO to mp6 (when/if possible) 20:20
offby1 nobody here but us zombie processes.
awwaiid ooo... zombies... 20:25
20:26 Aankhen`` joined
pmurias fglock_: i'm not really sure about it 20:27
it would at large be pushing stuff into the emitter
* largly 20:28
silug so is Pugs::Emitter::Rule::Parsec just expected to be broken?
it does use Pugs::Grammar::MiniPerl6; 20:29
as near as i can tell, that doesn't exist
unless i'm just being slower than usual
pmurias Pugs::Emitter::Rule::Parsec is in my knowledge abandoned 20:30
fglock_: i think we should add bootstraping kp6 into the roadmap
fglock_ pmurias: that's ok 20:31
when in the timeline do you think it should be done?
pmurias it would allow using cool stuff from previous milestones in the next ones
fglock_ silug: it's been abandoned for some time 20:32
pmurias: but it would also result in a very slow compiler, since kp6 is very unoptimized
silug i wonder why my fedora package just picked up a new dependency then
pmurias i don't think speed of anything other then the parser would make a difference 20:33
silug so i should be able to just remove it from the package and nobody will notice?
pmurias fglock_: --perl5rx emitts some code faster 20:34
fglock_ silug: yes; it is there waiting for someone to further develop pcr-on-haskell
pmurias: hmm - makes sense 20:35
pmurias any way we could cheat a bit and cache parse results 20:36
would make a huge difference on some file like EmitPerl5 which are composed of many different compliation units 20:37
[particle] md5 each compunit and compare against cache key? 20:38
pmurias yes 20:39
fglock_ hmm - caching is tricky, we are allowed to modify the grammar
caching must be lexical, probably 20:40
[particle] that would have to invalidate the cache
pmurias caching it's not nessesary for doing the bootstrap 20:43
compiling Grammar.pm takes over 3 minutes on my computer waiting another minute wouldn't make a big difference :) 20:44
i'll polish and checkin the version of kp6-perl5.pl shows which which parts of the compiler use up the most cycles 20:46
what is the recomended way of conditionaly use'ing a module? 20:47
(in p5)
Time::HiRes 20:48
[particle] BEGIN { require 'foo.pm' if bar; foo->import; }
pmurias used if ($profile) {require Time::HiRes;Time::HiRes->import(qw(time));} 20:52
dosn't seem to work :( 20:56
pugs_svnbot r17017 | pmurias++ | trunk: 21:01
: kp6: a version of kp6-perl5 with profiling
diff: dev.pugscode.org/changeset/17017
lambdabot Title: Changeset 17017 - Pugs - Trac
21:02 Blwood joined
pasteling "evalbot_r17016" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26252 21:03
21:03 weirdo_ joined
weirdo_ can i run p5 code under haskell? 21:03
uh, sorry 21:04
can i run p5 code under pugs?
diakopter weirdo_: sure
but it effectively just passes it to p5
./perl5.x.x I mean
weirdo_ that's bad :(
diakopter do you mean *only* p5 code? or a mixture of p5 and p6
Tene weirdo_: what's bad about that? 21:05
weirdo_ i need to work around my code's circular references
pmurias did i implement the profiling right?
weirdo_ p5 is pure-refcnt
guess i'll have to learn p6 :) 21:06
bye
21:06 weirdo_ left 21:08 lumi joined, fglock joined
pmurias fglock: re bootstrap i suppose doing it before or after phase 5 are the options 21:09
fglock agreed - after phase 5 would be the right time 21:10
pmurias althought the bootstrap is trivial compared to phase 5 21:13
and it could sort of test the compiler and give handy things such as multis to work with 21:14
fglock how about this:
create a directory for 6-in-6, and (mostly) duplicate the kp6 structure there 21:15
then use kp6 to compile 6-in-6
which is almost the same as bootstrapping, but it doesn't matter if it fails
21:16 pmurias joined
pmurias sounds the only resonable way 21:17
fglock cool
21:18 ofer joined
pmurias modifing the compiler in place using itself was pretty horid 21:18
fglock i know :) 21:19
pmurias it still bootstrapping but with a rather nicer pair of boots 21:21
fglock (still here) 21:22
heh
pmurias why do you use multiple nicks at a time
?
fglock i guess it's an auto-reconnect thing, i'm on wifi and it changes networks somtimes 21:23
sometimes
21:24 jedai joined
pmurias i see 21:26
fglock brb 21:32
21:39 cernd joined
pugs_svnbot r17018 | pmurias++ | trunk: 21:41
: kp6: a bit more sugar Method ->'ed
: infix:<!=>
diff: dev.pugscode.org/changeset/17018
lambdabot Title: Changeset 17018 - Pugs - Trac
21:42 mr_ank joined
pmurias will finish changing MOP.pm tommorow 21:46
sleep&
21:46 pmurias left 22:14 fglock left 22:20 unobe_away is now known as unobe 22:29 rashakil joined, rlb3 joined 22:32 ofer joined
nothingmuch [particle]: will you be attending YAPC::EU? 22:32
[particle] not this year :(
nothingmuch bummer 22:33
[particle] i'll be attending trek::nepal::annapurna
nothingmuch oh wow
better choice =)
how high will you be going?
[particle] 16 or 17k ft 22:34
nothingmuch do you need a special permit for that?
is that already glaciated, btw?
[particle] no. there's a circuit trail.
no, too tropical for that. the mountains are much taller than the valleys we'll be walking in :) 22:35
nothingmuch how long?
[particle] 25 days total
nothingmuch cool
22:35 ofer joined 22:36 justatheory joined
nothingmuch either way, my talk 22:37
when it's written
should be interesting for you =)
i'll be sure to email you a link when I'm done
[particle] thank you 22:38
nothingmuch though I thought it'd be fun to talk IRL, too
[particle] yeah, it would. maybe next year.
nothingmuch what are you kidding me? perl 6 will be out by then!
[particle] sorry, next christmas :)
we need to start planning the release party. where is the most perlish place to have it? 22:42
22:55 marmic joined, ofer0 joined 22:59 ofer0 joined 23:01 Blwood joined
meppl good night 23:02
nothingmuch [particle]: in an onion field in the sahara 23:05
23:18 justatheory joined 23:26 xern_ joined 23:29 thorat joined 23:31 fglock_ joined, Limbic_Region joined
fglock_ $neighbour++ # open wifi :) 23:32
pugs_svnbot r17019 | Darren_Duncan++ | trunk:
: ext/Muldis-DB/ : updated Core.pod to add the generic ordinal operators
diff: dev.pugscode.org/changeset/17019
lambdabot Title: Changeset 17019 - Pugs - Trac
pasteling "evalbot_r17018" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at sial.org/pbot/26258 23:33
23:43 devogon joined 23:48 rlb3 joined 23:59 drupek12 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!