svn switch --relocate svn.openfoundry.org/ svn.pugscode.org/pugs/ | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com
Set by audreyt on 7 November 2006.
eviltwin_b I have a vague recollection of their trying to convert their entire repo and having the conversion script choke, but that was several months back 00:00
cmarcelo Aankhen``: indeed it is.. gnome folks recently tried migrate everything to SVN twice and are still on CVS.. 00:01
Aankhen`` Ouch.
clkao really. i wonder why kde and gcc actaully migrated then
audreyt reinvite all done. 00:03
00:03 gnuvince joined
gnuvince audreyt: thanks for the commit bit :) 00:03
audreyt :) 00:04
gnuvince Quick question though, the email mentionned that my username was gnuvince, but I chose vfoley in the registration dialog, does that change anything?<
awwaiid what are you all doing, some sort of pyramid scheme? :) 00:05
audreyt that's fine
you can change nick anytime
eviltwin_b probably depends on the complexity of the revision history
audreyt awwaiid: yes!
awwaiid hah
audreyt awwaiid: want a commit bit? what's your email?
awwaiid [email@hidden.address]
clkao audreyt: the commitbit runs on weird 9999 port 00:06
gnuvince cool domain name
audreyt clkao: I know. jifty wants to be toplevel and I lack the fu to setup virtualhost on apache1 just today
or, actually: how do I convince jifty to not be toplevel?
i.e. /css/ and /__jifty/ etc 00:07
eviltwin_b Pyramid of Unusual Global Size?
awwaiid audreyt, also see perlbal for fancy virtualhost++ :)
audreyt awwaiid: welcome aboard :) please add yourself to AUTHORS and commit
that isn't a bad idea :)
but, not today
awwaiid right
clkao audreyt: what do you use for diagram these days? 00:08
audreyt clkao: OmniGraffle
only piece of proprietary software I depend on, if you don't count the macintel kernel
smash_ nice top 00:09
(ups)
clkao :/ 00:10
obra clkao: rumor is of a core duo to macbook this month 00:12
want to slip the customer hsip date? ;)
audreyt repairing evalbot 00:13
syncd will have to wait
00:13 evalbot_r14675 joined
audreyt ?eval 1+1 00:14
evalbot_r14675 2
clkao hsip?
00:21 dduncan joined 00:22 Khisanth joined
[particle] ship 00:22
00:28 drbean joined 00:31 electrogeek_brb is now known as electrogeek 00:32 electrogeek is now known as electrogeek_off
svnbot6 r14676 | Sage++ | Adding a todo on the Perl5 to Perl6 translation. Since my work has slowed down because of school work, I thought I'd add some thoughts on where the code is going, to show what I'm considering for the future. 00:36
audreyt journal and p6a posts up 00:37
have fun, I'll fade now. see you tomorrow... & 00:38
smash_ Compiling Pugs.AST.Internals # is completely trashing my system 00:42
00:48 mAdzzzzzz is now known as GabrielVieira
Patterner hmm... "unknown host commitbit.pugscode.org" 00:48
[particle] it should forward you to svn.pugscode.org:9999 00:50
00:55 monkey___ joined 00:56 diakopter joined 00:57 lyokato joined
cj audreyt: just got an email from you about a commit bit... 01:02
Firefox can't establish a connection to the server at svn.pugscode.org:9999.
sage: M-q 01:03
diakopter dj: works for me. 01:05
cj: 01:06
mugwump did somebody just send out admin bits to everyone?
cmarcelo mugwump: yes 01:07
eviltwin_b pugs just relocated, and everyone's been re-invited
cmarcelo there's a blog post with details on how change svn/svk to the new repo url.. 01:08
I think it was intentional giving meta-commits (admin) to everyone... 01:10
eviltwin_b looked to me like everyone was an admin on openfoundry too 01:11
obra not everybody got admin bits 01:12
01:12 cjeris left
cmarcelo obra: apparently diakopter is author but can see admin screens.. maybe its read-only? 01:15
diakopter I guess I could try to update something temporarily just to test it.
obra diakopter: change the punctuation of the pugs description 01:16
cmarcelo diakopter: there's a people menu where you can delete people from project?
obra diakopter: change the blurb on svn.pugscode.org:9999/admin/project/Pugs
lambdabot Title: Login, tinyurl.com/yyyr6c 01:17
diakopter oh good, it says permission denied. 01:20
heh.
all alarmed about nothing, I guess I was. 01:21
obra oh good :)
01:22 dmq joined
mugwump is there going to be svnweb support or similar soon, obra ? 01:24
oh, there's a url entry for it
01:28 kanru joined 01:34 ajg joined 01:35 ajg left 01:39 ajtgurney joined 01:46 ajtgurney left, fayland joined, baest_ joined 01:50 duend joined
Limbic_Region question - Perl6::Grammar has been usurped by Pugs::Compiler::Rule but has PCR been usurped by anything? 01:57
and why is PCR called PCR if it is pure perl 5? 01:58
SamB because it is in the pugs repo? 02:01
avar Limbic_Region: Why do you think something being called "Pugs" and being written in p5 is mutually exclusive? 02:02
Limbic_Region avar - I don't 02:03
it is an open question, actually asked by merlyn 02:04
SamB - but it is also on CPAN and it might be misleading to someone not in the know that they don't need Pugs to use it
namespaces can be confusing because ppl read more into them than is there
SamB so? 02:05
Limbic_Region SamB - I am not suggesting anything is wrong or that anything change
I am merely asking why it is the way it is (if there is a reason)
SamB hysterical raisons? 02:06
Limbic_Region yep, would be my guess
before it was completely p5
SamB or before it was on CPAN
Limbic_Region right 02:07
thanks 02:10
02:16 araujo joined
araujo thanks audreyt for invitation :-) 02:18
02:30 geoffb joined
araujo audreyt, you and TimToady will make me screw up Pugs and then don't tell me i didn't warn you :-) 02:32
cj araujo: we've got good docs about how things get screwed up, so we should be fine :) 02:33
araujo cj, hah 02:34
02:41 thepler_ joined 02:43 boo is now known as beu
geoffb audreyt: Just enabled my admin account under the new repo. Noticed that the user management interface of CommitBit doesn't have a "bulk add", like many email addresses in one text box -- that might be a useful thing when you want to hand out bits to a several people. 02:44
Oh, and thanks. :-) 02:45
Limbic_Region geoffb - she handed out lots and lots of bits at once - she scripted it 02:51
araujo Limbic_Region, don't say that ... i wanted to feel *special* 02:55
buu You are special! 03:08
cmarcelo @tell audreyt in src/MO/Compile/Class.hs all_class_methods isn't getting the attribute accessors, I suppose this is intended but is /= from the conclusion I got talking to you yesterday (or before). just checking. (6.4.2 compatibility on the way...)
lambdabot Consider it noted. 03:09
03:11 justatheory joined 03:12 justatheory joined 03:35 mako132_ joined 03:36 bsb joined
cj prepares to push some code 04:01
04:14 scw joined 04:19 SubStack joined
svnbot6 r14677 | cmarcelo++ | * src/MO: allows compilation in GHC 6.4.2. 04:19
pasteling "cmarcelo" at 200.232.237.143 pasted "re MO(H): redefining "data NoResponse" to "data NoResponse m" .. check line 43 of src/MO/Run.hs" (28 lines, 784B) at sial.org/pbot/20966 04:21
cmarcelo @tell audreyt check this paste when you can: sial.org/pbot/20966 04:23
lambdabot Consider it noted.
cmarcelo sleep & 04:26
04:38 eggzeck joined 04:46 justatheory joined 05:05 autark_ joined
svnbot6 r14678 | zgh++ | - set %*ENV<REQUEST_METHOD> (for CGI) 05:05
gaal oi. the rss feed is regurgitating very old commits... 05:16
05:26 drrho joined 05:53 duend is now known as duend`sleeping 05:57 fayland joined 06:14 baest_ is now known as baest 06:15 BooK_ joined
geoffb @tell Limbic_Region Yep, I knew that ... still, is there ever a bad reason to say "thank you"? :-) 06:20
lambdabot Consider it noted.
06:25 bsb left 06:37 buetow joined 06:49 marmic joined 06:54 kanru joined 07:05 eggzeck joined 07:20 Psyche^ joined 07:26 lisppaste3 joined 07:31 shar joined
svnbot6 r14679 | zgh++ | [CGI.pm] 07:31
r14679 | zgh++ | - added self. to internal method calls in load_params
07:33 Psyche^ is now known as Patterner 08:12 nipotan is now known as nipotaway 08:29 christopher joined
christopher Now that the SVN server has relocated, is there a place where I can browse the revision log, like rt.openfoundry.org/Foundry/Project/...it=Show%21 ? 08:30
lambdabot Title: OpenFoundry&#40;č‡Ŗē”±č»Ÿé«”é‘„ōæ½xE9ōæ½x80 ōæ½xE5 ōæ½xB4-å°ˆę”ˆé–‹ē™¼å¹³å°&#41; -- OSSF, tinyurl.com/y2536v
masak just wanted to report that the svn switch as advertised in the channel topic didn't work out for me 08:34
however, a checkout from the new location did 08:35
christopher masak: I think there should be "pugs" at the end of the first URL in the topic
masak christopher: you might be right. can anyone (1) confirm that the current command in the topic doesn't work, but that christopher++'s addition does, and (2) change the topic? 08:40
08:49 Psyche^ joined 08:52 dakkar joined
dakkar did something change in the hosting of Pugs? I just received a message asking me to set a password... but I already had an account on openfoundry 08:53
ok, the topic tells something...
lumi dakkar: The repository's been moved 08:54
08:54 kensanata joined
dakkar lumi: I assumed as much ;-) reasons? 08:54
lumi Apparently that command in the topic doesn't work (I use svk so I didn't use it)
openfoundry goes down too much, I think?
dakkar well, that command looks like it needs something after the Ā«svn.openfoundry.org/Ā» part 08:55
lumi It probably wants a /pugs, yeah 08:57
If you're in svn, and adding that bit works, we can change the topic 08:58
dakkar I don't have a worknig copy right now... it's in a backup disk, I rebuilt my machine last week
lumi No worries then, but the password invites are real 09:00
09:01 Schnueff joined
christopher lumi: worked for me with "/pugs" when I tried with svn 4-5 hours ago 09:01
09:05 Psyche^ is now known as Patterner
lumi I'll try it.. 09:05
svn switch --relocate svn.openfoundry.org/pugs svn.pugscode.org/pugs/ | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com 09:06
xinming lumi: How about svk? 09:11
:-p
lumi svk mi --relocate //mirror/pugs svn.pugscode.org/pugs/ # is what I did 09:12
lambdabot Title: Revision 14679: /
lumi Um thanks, lambdabot
09:12 drrho joined 09:29 kanru joined 09:42 ludan joined 09:53 fayland joined 10:15 ofer0 joined 10:19 chris2_ joined
Coleoid Is anyone in channel able to help with a make failure on Win32? 10:20
10:20 chris2_ is now known as chris2
Coleoid NMAKE : fatal error U1073: don't know how to make 'Spec/_svn/tmp/Functions.pod.tmp' 10:20
10:30 Odin- joined 10:36 ruoso joined
Coleoid With a bit of hacking on Makefile.PL and Makefile, I got it to build... 10:41
Though I know what I've got is only 20% of the correct answer. 10:42
The short story is that on Win32, dot prefixed directories are incorrectly handled by Visual Studio... 10:44
So a SVN client option is to have the .svn directories as _svn instead...
And the makefiles have magic to avoid .svn/ wherever, but not _svn. 10:45
Anyone with makefile fu, I'd appreciate an approach that might be worth committing. 10:51
work &
dduncan while I've already moved over to the new repo, I notice the old one was missing most of its commits 10:53
just having 343 10:54
dakkar they're probably still mirroring 10:59
dduncan nothings moving ... the old repo is staying at 343 11:02
dakkar but the new repository has the most recent version?
11:09 fglock joined 11:10 folli300 joined
svnbot6 r14680 | rootmj++ | [PaP6-testing] 11:12
r14680 | rootmj++ | * switch URL for pugs repository
11:14 xinming_ joined 11:16 elmex joined
elmex hmm 11:16
why was i sent a pugs commitbit? 11:17
dakkar elmex: you probably had a commit bit in the old openfoundry repository 11:18
elmex ok
that was it
dakkar the repository has been moved
it caught some of us by surprise ;-) 11:19
11:24 folli300 is now known as follil 11:50 andara joined 11:57 frederico joined
fglock no commits today - I'm stuck upgrading OS X - I found some links about it but it didn't help - forum.insanelymac.com/index.php?showtopic=17584 12:02
lambdabot fglock: You have 1 new message. '/msg lambdabot @messages' to read it.
Title: Perpetual OS X Installation - InsanelyMac Forum, tinyurl.com/y83wpb
fglock ashleyb: thanks! 12:04
audreyt greetings lambdacamels 12:07
lambdabot audreyt: You have 2 new messages. '/msg lambdabot @messages' to read them.
audreyt @massages
lambdabot cmarcelo said 8h 58m 23s ago: in src/MO/Compile/Class.hs all_class_methods isn't getting the attribute accessors, I suppose this is intended but is /= from the conclusion I got talking to you
yesterday (or before). just checking. (6.4.2 compatibility on the way...)
cmarcelo said 7h 43m 33s ago: check this paste when you can: sial.org/pbot/20966
audreyt fglock: oy, your macbook locked? 12:08
fglock: put DVD#1 into the cdrom; shut down the computer; press C and don't relesae that key; press power button 12:09
fglock: then wait a while when it boots into DVD; from there you have a Terminal
and you can hunt down the cause by either doing an upgrade install or removing the offending files 12:10
it also helps to first boot with the apple key and V (Cmd+V) pressed so you can see at which step it's hanging
12:12 Psyche^ joined
audreyt Coleoid: can you paste the patch to sial.org/pbot/perl6 ? 12:12
12:14 eggzeck joined
fglock audreyt: will do - thanks! 12:16
12:18 jferrero joined 12:19 boo__ joined 12:20 Psyche^ is now known as Patterner 12:21 rodi joined, boo_ joined 12:22 boo_ left, boo_ joined
nothingmuch mooooosen 12:26
audreyt nothingmuch: hi moose! 12:27
nothingmuch hola
audreyt nothingmuch: how's life? 12:28
nothingmuch sleep is screwey, skin is unhappy
but asiide from thatall good
clkao moose 12:29
audreyt: new repository so far so good?
nothingmuch how's life on the other side of the planet? 12:32
audreyt clkao: so far so good
openfoundry got stuck autosyncing 12:33
so I guess we'll have to push to it manually
svnbot6 r14681 | audreyt++ | * Remove the irrelevant parts of early sketches from src/PIL/.
audreyt I think svnsync is the way to go
I think.
clkao right, it could sync to ancient server 12:34
fglock audreyt: I guess I'll reinstall from scratch, after saving yesterday work on MP6 12:36
audreyt fglock: oy :/ so sorry to hear that
I think you want /opt and /usr/local backed up 12:37
that will save a lot of downloading time
as well as /Applications and your /Users/fglock
try boot to dvd and rsync them into your other computer 12:38
fglock yes, I'll do that
audreyt sorry for messing up your /usr/lib -- I think that's the root cause
fglock np :)
12:42 kanru joined 12:46 bonesss joined
nothingmuch audreyt: oh btw 12:46
in Invocation::Method, the caller thing was a hack
a really really relaly ugly hack =)
i'm refactoring that so that the responder interface gets, alongside the invocation, extra params 12:47
and the ones that require caller simply depend on getting the "stack" parameter
which is an object that can be queried for stack frames
i don't believe there are anymore yucky yucky hacks except the AG stuff, which you haven't ported anyway 12:48
12:49 alexe joined
nothingmuch workout & 12:49
12:50 Juerd_ joined
audreyt nothingmuch: ok cool, looking forward to refactoring 12:50
nothingmuch: I think in Haskell we'll put the stack into a monad anyway
so it's not carrying it thru invocation
and also instead of an Arguments abtraction we'll just use concrete Capture callconv 12:51
while providing ways to cast other callconvs into Perl6 Captures
it's assuming of course that it can subsume all other callconvs.
(but I think it can.)
12:53 dduncan left
nothingmuch yes, arguments is a list in p5 but a capture in p6 12:57
invocations that don't get subsumed by captures are simply going to use something else ^_^ 12:58
audreyt :D
in any case the returning continuation will be put into the monad 12:59
so you can query the Signature of the retcont
otherwise known as the context
it amounts to the same thing as passing everything with the invocation
which is the same as your stackobj. 13:00
clkao: openfoundry has no revpropchg hook 13:02
clkao: so, what do we do. can pushmi help? 13:03
clkao right
not immediately. pushmi has to be installed in slave site
audreyt i.e. pretend we are slave to openfoundry
and relay commits to there
but script a replay
of prev commits 13:04
clkao won't work when openfoundry is down then
audreyt not realtime
i.e. async not sync
clkao well, then it's the same as svnsync.
audreyt but it wouldn't require revprop?
clkao you need prop changes to make the author stuff happy
audreyt chghook, that is
what if I don't care about that?
clkao one way export? mnmmm 13:05
audreyt I can already do that with handwriten svkmerge 13:06
clkao ya
audreyt but it's not as reliable as a dedicated replay
clkao but you still want repository mostly as replicate right?
so you can do sw
audreyt well no
not in the case of openfoundry
clkao then just sm -I all the time
audreyt or svm ;)) 13:07
13:07 Limbic_Region joined
clkao but this is going to be a bit tricky for svk 13:07
because the two repositories have the same uuid
why don't you commit a revision in openfoundry to delete everything except a README saying this repository has moved
audreyt no we changed the uuid
clkao on openfoundry? or the dump i had? 13:08
audreyt on openfoundry
clkao good
so sm -I would work
nothingmuch audreyt: yes, the stack obj is for p5 land 13:10
audreyt nothingmuch: if you carefully erase all calls using &goto 13:12
then you can make caller() be your stack obj in p5 land :)
may be insane, though. nvm
nothingmuch yes 13:13
that is the plan for the p5 "emitter"
audreyt oh, so you anticipated the insanity
nothingmuch that all hooks are wrapped in a secret gotoing frame
audreyt good good.
yup.
nothingmuch for e.g.private methods
audreyt btw, p5 does not distinguish clsmeth with instancemeth
nothingmuch clkao's local-stack fudge may be very veyr useful for this toio
audreyt neither does p6
nothingmuch i know 13:14
audreyt how do you forsee we make use of separate slots?
nothingmuch wait
p6 doesn't?
audreyt no it does not
nothingmuch in p5 the plan was to just mush them together on the responder interface level
yuck =(
audreyt why yuck?
nothingmuch with a hook-frame that does if ref blah
audreyt ::Point.x will simply respond with an error of uninitialized x in the slot
i.e. the layout created for ::Point is nil 13:15
nothingmuch because object methods want instance data, they don't make sense as class methods
audreyt while ::Point.new will have a proper layout
clkao audreyt: did you see my selfless hack
nothingmuch and class methods as instance methods are not completely bork, but also quite silly
audreyt but you never know; maybe they don't want instance data
maybe derived class use methods to emulate instance data
nothingmuch then they should be a module, not a class ;-)
13:15 Debolaz2 is now known as Debolaz
audreyt clkao: url? 13:15
clkao audreyt: p5p
nothingmuch i don't mean to say it's useless 13:16
audreyt it's just very p5ish :)
nothingmuch it's just not as useful as a separate namespace
but whatever
audreyt yeah. I initially find it surprising as well
but then ify ou think about prototype OO systems
then it makes perfect sense to have an unified method table
clkao audreyt: groups.google.com/group/perl.perl5....4cfb580d09
lambdabot Title: Google Groups: perl.perl5.porters, tinyurl.com/tfupd 13:17
nothingmuch that's cruft from prototype systems, IMHO ;-)
prototype systems make sense that way when the prototype is indeed a prototype
but the mother of all prototypes is just nil
audreyt clkao: nice :) 13:22
nothingmuch: I don't disagree -- but S12 has a consistent design and I'd like to go with it for a while and see how it breaks :) 13:23
@tell cmarcelo indeed; we should probably just remove the class_methods etc from the typeclass Class. 13:24
lambdabot Consider it noted.
13:24 kanru_ joined
audreyt clkao: deploy that on jifty any time soon? :) 13:25
clkao: also, I think Spiffy has the right idea
any subs that has a prototype is exempted from selflessing
13:25 GnuVince joined
audreyt becuase in p5 prototypes doesn't work on methods 13:25
so it's a clear sign that it's Not A Method.
makes sense to you? 13:26
clkao audreyt: i know. those are the trivial bits. right now my version segfaults for unknown reasons
audreyt you just need to install more segments!
clkao or more faults
13:30 GabrielVieira joined 13:36 nusgnaf joined
audreyt it appears I broke svn.openfoundry.org/pugs entirely... 13:36
requesting technical assistence from openfoundry team :/
13:37 kisu joined
clkao how long is it gonna take? 13:44
13:46 rodi joined
audreyt I don't know. worst case I can rm pugs and start anew there 13:51
14:00 nusgnaf left, cmarcelo joined 14:05 alexe left 14:06 justatheory joined
fglock not sure if I got it right - is chromatic volunteering to implement a Parrot backend? 14:14
that would be nice
for MP6 :)
audreyt fglock: yes. get in touch with him and point out what he needs to do :) 14:15
smash_ good afternoon 14:17
audreyt smash_: heya :)
Limbic_Region so out of curiosity, should MO landing in the repo result in more tests being enabled and passed? 14:18
lambdabot Limbic_Region: You have 1 new message. '/msg lambdabot @messages' to read it.
Limbic_Region @messages
audreyt gives up fighting with svn.openfoundry.org...
lambdabot geoffb said 7h 57m 33s ago: Yep, I knew that ... still, is there ever a bad reason to say "thank you"? :-)
audreyt Limbic_Region: yes, once it's hooked up with our calling convention, t/oo/ should all pass
Limbic_Region woooot
smash_ audreyt: hiya
audreyt Limbic_Region: I plan to do that over the next week :)
but again no promises -- though the hard part is done over the last 2 months
Limbic_Region audreyt - when are you traveling?
audreyt the glue is much easier.
I'm in Rio 14:19
leave brazil in 4 days
smash_ implement a Parrot backend? for MP6.. hmm
audreyt so need to finish the glue sketch in 4 days
otherwise from past experience I'll be disabled by 7 days following flying back to .tw
otoh, it may not be so bad, as $job is not demanding anymore
Limbic_Region audreyt - on the off chance you get some free time, please post pics somewhere (outdoor pics)
audreyt and I have another month ahead before flying to france. 14:20
Limbic_Region: heh. you don't like the two I posted? :)
Limbic_Region has been wanting to visit SA for some time now
audreyt ah. brazil pictures
I don't have a camera.
Limbic_Region :-(
audreyt and macbook taking pictures outdoors is... fragile
Limbic_Region steel er um, borrow someone elses
audreyt there is plenty of Rio pictures on youtube and flickr...
not to mention National Geography :)
Limbic_Region not quite the same as virtually knowing someone who was there
clkao audreyt: france?
audreyt clkao: Nice France. work with rgs on unified runtime. 14:21
clkao oh right.
14:21 vel joined
clkao why is rgs in Nice 14:21
audreyt apparently it's his parent's place and he grew up there.
so coming there for weekend hackathon = no problem
clkao whoot. what date?
rgs right 14:22
audreyt Jan 20th
19~21 actually for me
not sure what for rgs
rgs not sure yet
clkao going anywhere else in europe?
i will be in taiwan that time
audreyt no :)
clkao sugoi
audreyt oh cool.
Limbic_Region audreyt - perlmonk's homenode pic is nice, where is the other one?
audreyt Limbic_Region: the other one is pugs.blogs.com/photos/audrey/20061021.html from schwern's place 14:23
lambdabot Title: Audrey: 20061021, tinyurl.com/yzcpk4
14:23 bpphillips joined
rgs audreyt: probably 20-21st 14:23
audreyt the homenode one is in amazon.com
rgs: ok, checked 14:24
rgs or maybe 14-15. I'll know that in a few days
audreyt ok
I'll not announce on pugs.blogs.com before you confirm :)
rgs since I'm trying to get another job, it's a bit difficult to make plans.
clkao audreyt: you should get rtw ticket 14:25
audreyt: consult bergman travel inc.
Limbic_Region very nice 14:26
rgs audreyt: that said, if you have a specific list of topic, or only ideas, you're welcome to sent them in my general direction
audreyt rgs: ok, will do
clkao: rtw?
clkao round the world
fglock audreyt: unified runtime = P5+P6 ? 14:28
nnunley Careful. Bergman travel inc. requires first class legs all the way around.
audreyt fglock: yes 14:29
and perl4
nnunley And may choose longest path travel.
audreyt you'll note perl4 runs on that runtime as well
clkao more hackathon venues then ;)
audreyt lunch, bbl
14:31 weinig is now known as weinig|bbl 14:36 idiotmax joined 14:42 rhizo joined 14:48 duend`sleeping is now known as duend 14:56 iblechbot joined, rlb3 joined 14:57 HiTech69 joined 15:00 shigeta joined 15:16 [particle] joined, nusgnaf joined, hexmode joined 15:20 cjeris joined
Limbic_Region buu when you are about ping me - I think I figured out how to get ssh from work without changing the port ssh is listening on - requires an apache hack and a special ssh client dag.wieers.com/howto/ssh-http-tunneling/ and proxytunnel.sourceforge.net/ 15:43
lambdabot Title: DAG: Tunneling SSH over HTTP(S), tinyurl.com/z5emd
Limbic_Region oops
nevermind the man behind the curtain 15:44
15:56 bpphillips joined 16:01 ofer1 joined 16:21 nusgnaf left 16:24 hugues23 joined 16:36 seano joined
seano hey -- anyone know the proper process-coding-system to feed pugs non-ascii characters inside emacs? 16:38
audreyt seano: utf8? 16:43
@tell cmarcelo let's just forget GHC 6.4 ever existed :)
lambdabot Consider it noted.
audreyt bbiab... &
avar be back in a blip? 16:44
fglock how to fix "Cannot relocate uninitialized path /mirror/pugs at .../SVN/Mirror.pm" ? 16:45
after "svk mirror --relocate //mirror/pugs svn.pugscode.org/pugs" 16:46
lambdabot Title: Revision 14681: /
seano audreyt: hmph. Tried that (for both input and output coding, then for terminal coding system), but it still seems to misbehave. I'll poke around a bit more. 16:47
16:53 buetow joined
lumi fglock: svk mirror --list 16:54
ofer1 what's svk ? 16:55
Subveriok ?
Subversiok ?
dakkar ofer1: it's a sort of super-powered front-end to Subversion
written in Perl by clkao 16:56
ofer1 dakkar, cool, where can I get it from ?
awwaiid cpan :)
lumi Subversion on Steroids
[particle] svk.bestpractical.com/view/HomePage 16:57
lambdabot Title: HomePage - SVK Wiki
ofer1 thanks!
clkao dakkar: :) 16:59
fglock lumi: --list gives me "/pugs/trunk svn.openfoundry.org/pugs" - should I use "svk mirror --relocate //pugs/trunk ... " instead?
dakkar fglock: most probably
clkao fglock: yes, read audrey's blog entry 17:01
fglock clkao: I get a "cannot relocate uninitialized path" error 17:03
clkao what command did you use ? 17:04
hmmm i see it. 17:05
oh, //pugs/trunk instead
fglock same problem with //pugs/trunk 17:06
clkao does svk log -l1 //pugs/trunk shows you a "orig r..." ? 17:10
fglock I get "Filesystem has no item: File not found: revision 0, path '/pugs/trunk' " - I think I lost some files while restoring my backup 17:12
I guess I'll do a new checkout 17:13
17:18 cmarcelo joined
monkey___ what's going on with the pugs webserver? 17:19
monkey@cowu:~$ telnet www.pugscode.org 80
Trying 213.84.134.174...
telnet: Unable to connect to remote host: Connection refused
:(
cmarcelo @moosages 17:20
lambdabot audreyt said 3h 56m 19s ago: indeed; we should probably just remove the class_methods etc from the typeclass Class.
audreyt said 36m 34s ago: let's just forget GHC 6.4 ever existed :)
[particle] are there any writeups on using MP6 yet? 17:21
cmarcelo @tell audreyt re 6.4: hehe I was expecting that :).. it's a matter of reverting except for the 'NoResponse vs NoResponse m'-thing.. (did you saw it?) 17:23
lambdabot Consider it noted.
fglock [particle]: MP6 is written in MP6 17:27
there is a small text file in docs/notes/mini...
and AST.pm and BOOTSTRAPPING in v6/v6-MiniPerl6 17:28
[particle] ok, great.
17:28 DHGE joined
[particle] new PAST will probably be committed today 17:28
fglock [particle]: MiniPerl6 is "PIL-3", kind of 17:29
[particle] so it'll be time to start doing transformations soon
17:29 justatheory joined
nothingmuch mooo! 17:29
fglock [particle]: transformations? 17:30
[particle] so, there's no diff between assignment and binding... is everything binding, or assignment?
fglock binding
[particle] fglock: tree transformations parse -> past -> backend stuff
great
and named args exist, but don't map to positional params automatically? 17:31
TimToady no named, just { foo => '123' } [note curlies] 17:32
no variadic args, even
[particle] ok, that's great
actually, it'll mean a lot of parrot's calling convention magic can be avoided 17:33
of course, that's assuming a p6 built on mp6
fglock yes, and the same for P5 and Haskell - so we can get a fast implementation
clkao GnuVince: you pinged me the other day? 17:34
[particle] sure, but probably slower runtime performance
...which is fine for now.
dmq cool. i just got cutgroup working in perl5.
[particle] yay 17:35
fglock [particle]: P6-full will not be implemented in MP6
TimToady It's just (?()true|false), really.
if by cutgroup you mean P6's ::
[particle] fglock: that's what i figured.
fglock rather, MP6 is a subset, so that we can share some code using it 17:36
dmq cutgroup is what i understand :: to be. 'on failure try next alternation'
17:37 mosca___ joined
dmq but i dont really get what you mean by "It's just (?()true|false), really." 17:37
TimToady it really comes out to the closest thing regex has to if/then/else
[particle] dmq: did you and pm solve the impedence mismatch between (*CUT) and <cut> yet?
...and <commit>...
dmq yes, sortof. im going to rename my CUT to be 'SKIP', and not implement a 'CUT' (since it doesnt fit into the p5 framework from what i can tell). 17:38
TimToady but :: generalizes what can go into the condition over (?()|)
dmq (*COMMIT) in p5 is afaict the same as <commit> in p6. 17:39
[particle] dmq: great, the new name seems more appropriate, and helps distinguish
TimToady fglock: I would not be *suprised* if someone implements a P6 interpreter in MP6 eventually, just so they can run P6 on their nanobots
dmq TimToday: im probably being dense, but i dont see the connection between conditional patterns and ::
17:40 bernhard joined
TimToady [ condition :: true | false ] 17:40
[particle] i figured there would be a p6-on-mp6 impl, too
dmq oh.
oh right.
heh.
TimToady but it stacks up a lot nicer than nested (?()|(?()|))
[ cond1 :: true1 | cond2 :: true2 | cond3 :: true3 | otherwise ] 17:41
dmq hrm, they are implemented totally differently tho. :-)
fglock TimToady: yes, that would be nice :) 17:42
17:42 mwh joined
dmq and they dont require the inside of the (?(...) to be a lookahead. which is cool. 17:43
nice way to think about it.
TimToady *curtsies*, er, *bows*
fglock [particle]: mp6-on-mp6 is almost ready - I'll get back to it when I finish recovering from the system reinstall
[particle] what's the best way to see mp6 in action? 17:44
dmq one minor issue with them is that they impose a speed/space issue as BRANCH_next and TRIE_next have to be treated as "yes-states". But thats not such a bad thing and the pernalty is per regex only.
fglock then it can be extended by programming in mp6
[particle] install v6, then goto v6/v6-MiniPerl6
dmq so i think this means that the only perl6 regex construct that you flat out cannot represent in perl5 is the p6 <cut> 17:45
GnuVince clkao: Yes, I saw that you were the original author of Benchmark.pm in pugs and I wanted to know if you wouldn't mind if I added some features (and fixed the TODO)
fglock the t/* files work with some errors - I'll commit some fixes asap (as soon as I finish recovering svk) 17:46
[particle] ok, great.
i'm relocating my pugs repo atm, then i'll sync/install/play
i think elephant would be a good name for p6-on-mp6 17:47
it's strong, slow, and is self-aware :)
dmq TimToady: im assuming that since you havent made any negative comments that you dont have too much of an issue with what im doing... I hope im right.
fglock clkao: how do you bypass "svk up -s" to take only the latest revisions?
clkao svk sync --skipto HEAD //path/mirror; svk up 17:49
PerlJam dmq: you mean that you can not now do perl6's <cut> in perl5 or that it's impossible to do?
fglock [particle]: mp6 = lrep + PILN; p6 = pugs + v6 + something else
dmq i dont see how its possible to do. 17:50
17:50 mosca___ is now known as mosca
clkao GnuVince: please do! 17:50
dmq the soon to be called (*SKIP) IMO gives you something close enough. i think.
TimToady well, it's easy enough to pull the beginning of the string up to the current location. question is whether regex represents all positions as pointers or some of them as offsets from front of string. 17:51
which would be thown off if the front of the string moved.
dmq: my main qualm is that so far it seems like what you're putting in is rather difficult to explain, and if it's difficult to explain, it probably means we're missing some important generalization. 17:52
dmq or im a sucky explainer. 17:53
:-)
which almost certainly correct, regardless as to what it says about the generalization.
PerlJam It's too bad there's no way to insert markers in the RE that say "tag this location in the string with this little bit of information" and then be able to get that stuff out again in some structure. 17:54
fglock [particle]: see also pugs.blogs.com/photos/visiolization...tstrap.png
lambdabot tinyurl.com/yzjxzo
dmq actually, thats almost what does happen PerlJam.
TimToady it's like you want something that generalizes "last" into "last MyLoop".
PerlJam ah, then how do you access the structure?
because I was thinking that once the locations are tagged, you just use a substr() to do the <cut> 17:55
(among other things)
dmq like for instance you can say: /A+(*MARK:FOO)B+(*CUT:FOO)/
TimToady substr can certainly be used to pull up the front of the string. the question is whether the regex engine will coredump after you do that.
PerlJam TimToady: not if you switch out the string it's operating on :-) 17:56
TimToady then it's not <cut>...
dmq I more meant that it doesnt really fit into the expected behaviour of a perl5 pattern.
PerlJam TimToady: no, no, you copy, cut and forget about the original
TimToady and all the little pointers currently running around the regex engine can just go hang? 17:57
fglock clkao: I get "REPORT request failed on '/pugs/!svn/bc/14681' ... 200 OK
PerlJam Well, I was thinking that at the copy it would be more like a "reboot" on the RE engine.
dmq and you can do it with /(?{$Position=$-[0]})..../ afaict.
TimToady well, really, to fulfill the intent of <cut>, you have to get the regex engine to throw all that info away, or it's rather pointless. 17:58
clkao fglock: you need latest svn::mirror from svn://svn.clkao.org/member/clkao/modules/SVN-Mirror
GnuVince clkao: cool, thanks :) I know what I'll be doing tonight.
clkao i will release soon
TimToady whole point of <cut> is to be able to do continuous matching on an infinite stream.
dmq right which IMO is out of perl5's scope. at least insofar as I doubt ill try to implement it. 17:59
Perl6's regex engine has to be better at something eh? ;-)
TimToady :D 18:00
PerlJam I've never needed <cut> in perl5 so far ;)
dmq ill put it on my speculative todo list tho.
TimToady but how many times have you heard people wish for the ability to run a regex against an input stream?
18:00 GabrielVieira joined
PerlJam I've never heard anyone wish hard enough. 18:01
dmq many many times.
mostly because they want to use a regex for $/
TimToady well, until they cotton to the fact that they've got a figure/ground issue with that.
PerlJam dmq: these changes you've hacked in, are they going towards 5.10 or what? 18:02
TimToady just as most splits really want to be combs
dmq yes, pretty much.
PerlJam excellent
dmq most of them are in perl5.9.5 already in some form or another.
im still tweaking.
:-)
TimToady anyway, <cut> is just one of those things that's in there to make sure P6 has a completely different set of FAQs than P5. 18:03
dmq like id love a better name for cutgroup than cutgroup.
:-)
its kind long to type.
[particle] if
dmq oh, and i wanted to ask, my cutgroup doesnt actually care about group.
TimToady then
dmq its any branch. so (?:foo(*IF)|...) and (foo(*IF)|...) would behave the same assuming the (*IF) (cutgroup) worked. 18:04
s/worked/applied/
i hope that matches. it would be pretty tricky to only make it work inside of a capturing parens. 18:05
TimToady when we say group with respect to :: in P6 we also mean inside (...) as well as [...] 18:06
dmq so, Ill call it (*IF) and (*THEN) :-)
TimToady only there is no *IF
:: is then
regex already iffifies the left side.
dmq oh my.
TimToady the fact that you get to the :: means the conditional passed. 18:07
[particle] tt: you had me at 'then'
TimToady sorry, to lazy to type what I mean. :) 18:08
*too even
18:09 mwh left 18:12 elmex joined
nnunley I suppose it would be confusing to use prolog's cut operator ; 18:17
TimToady cut WHAT is the question
18:18 andara left 18:19 kisu joined 18:25 vell joined 18:29 rlb3 joined
fglock what packages should I install to get "make" in OS X? 18:31
dmq returns from afk. 18:32
clkao the apple dev package
dmq TimToady: so I'll use (*THEN)
fglock is that Xcode? it does not seem to be free
nnunley It's a free, but large, download from Apple.
developer.apple.com/tools/xcode/ 18:33
lambdabot Title: Tools - Xcode 18:34
18:34 rlb3 joined
fglock I get "you must be a member of the ADC to download this item" 18:37
ashleyb fglock: it's on the os x install cd 18:38
but you have to do a "custom" install
awwaiid and doing the ADC membership is free
fglock ashleyb: thanks 18:39
1st-day newbie here
ashleyb x11 should be on that cd too -- but it to is not a "default" install 18:40
*too
18:41 SubStack joined
ashleyb you can get these packages by using the installer on the CD -- or by just opening the pkg's in the directory on the CD -- I'd suggest using the installer. 18:42
fglock ok - found it 18:43
nothingmuch lllllllllllllllll 18:47
ah
18:57 frederico joined 18:58 boo_ is now known as boo_away
cj audreyt: how interested are you in going to boston to work with the mono folks? 19:26
fglock I can commit again! 19:27
svnbot6 r14682 | fglock++ | MP6 - compiles MP6 "token" to MP6 "method"; started "method" impl
audreyt fglock++ # wooot! 19:28
lambdabot audreyt: You have 1 new message. '/msg lambdabot @messages' to read it.
audreyt cj: very interested. but my schedule is full until at least March.
clkao audreyt: i have a flight ticket to boson in may 19:29
audreyt clkao: oh nice. maybe May then?
clkao sounds almost like a plan.
fglock audreyt: if you have time, try MP6 t/04-regex.t 19:30
cj audreyt: okay. Early, Mid or Late march?
audreyt clkao: early, mid or late may?
clkao the flgiht is early may iircf
audreyt cj: if clkao is also around that can easily 5x productivity 19:31
cj audreyt: okay. May then?
clkao but wait
audreyt would be nice if we also have putter kolibrie etc around
clkao is the project still going on after 8 weeks from now ?
cj audreyt: a large group would take more planning
audreyt clkao: huh? 19:32
clkao it's christmas!
aren't we all done by christmas?
audreyt I think this christmas is about perl6-alpha and pugs 6.28.0 and MP6 bootstrap
I don't think anyone has promised anything perl6-final on 2006 christmas :)
clkao ;) 19:33
where's 6 on 5
audreyt er v6-alpha.
sorry, in my brain the unified perl runtime is already firmly taking shape
I no longer distinguish perl6-alpha with v6-alpha
:)
clkao what's the unified runtime plan
19:33 boo_away is now known as boo_
audreyt you know as much as anyone; try to find encodings of all pugs/mp6 primitives as perl5 ops. 19:34
that's all actually :)
clkao sounds like nameshub 19:36
audreyt cj: so, may 10ish?
cj audreyt: I can't commit to it, but I'll aim for it. 19:37
I haven't been here long, so folks might not take me seriously yet. Ask TimToady how it worked last time :)
audreyt cj: muito obrigada.
cj no problemo 19:38
19:38 [particle] joined 19:39 TSa joined
cj audreyt: who usually foots the bill? TPF? 19:40
audreyt cj: no... I havn't been paid by TPF to do Pugs stuff, other than for food in YAPC hackathons 19:41
cj: see if Novell, or your company, is interested... failing both, I'll see if I can find a sponsor too 19:42
19:43 dakkar joined
cj audreyt: alrighty. I'll run it by folks who might be able to pay for it. 19:44
last time I talked to management about perl6 (in 2004-ish), they were skeptical.
I'll use the unit test page as proof that it's not sucking. 19:45
got a url?
audreyt smoke.pugscode.org/
lambdabot Title: Pugs Smoke Reports
audreyt spec.pugscode.org/
lambdabot Title: Official Perl 6 Documentation
cj thanks
will perl6 be able to run perl5 code on the new vm? In short, will perl6 on mono be backward-compatable with perl5 on perl5? 19:47
audreyt we will port also the bridge
cj expects no
audreyt as activestate has once done
and as already existing actually as another OSS projec
audreyt forgot the name 19:48
anyway, we'll be able to run mono and p5 side by side.
cj cool. probably good enough.
audreyt it is good enough with GHC+Perl5 for my $job :)
so mono should have little problem
cj okay. can you give me names of folks you'd like to see included in the trip via email? [email@hidden.address] 19:49
audreyt well, other people are locals :) 19:50
cj great
audreyt and clkao already got ticket to boston
so essentially just for myself...
cj good deal.
clkao: does miguel know you're coming?
clkao i don't know he knows me 19:51
svnbot6 r14683 | fglock++ | MP6 - parses "method" signature
cj was that "I don't know. He knows me." or "I don't know whether he knows me" ?
fglock audreyt: now we just need a "Block" parser - then bootstrap debugging begins... 19:56
clkao sorry, "i don't think he knows me"
19:57 zgh joined
fglock home & 19:57
19:58 fglock left
audreyt @tell fglock I think the strategy for "block" is force declaration of all vars upfront, as the old C 20:02
lambdabot Consider it noted.
20:05 bonsai joined
cj clkao: join #mono on irc.gnome.org and I'll introduce you 20:06
20:09 anatolyv joined 20:16 weinig|bbl is now known as weinig
cj hurm. 99 tests failing on the release on linux? 20:19
20:19 rodi joined
cj maxine, shorten m19s28.vlinux.de/iblech/stuff/pugs-...33299.html 20:20
lambdabot Title: TAP Matrix - Sat Oct 28 16:14:17 2006 GMT, tinyurl.com/y2gq6t
cj oh, wrong channel :)
clkao audreyt: huh? yapc hackathon food was sponsored by us 20:21
audreyt clkao: yapc::vancouver 20:22
clkao ah
audreyt also ::israel
smash_ good evening 20:26
audreyt greetings smash 20:27
cj: it may also help to say that perl6 is effectively defined by that smoke test suite; when the spec is changed ot augmented the test suite reflects that very quickly 20:28
cj audreyt: alrighty. I've been terse in order to get a feeling whether they're interested in hearing more. 20:29
20:29 TimToady joined
audreyt k 20:32
20:36 weinig is now known as weinig|bbl 20:50 duend is now known as duend`school 20:54 amv_ joined 21:00 zgh joined 21:08 nwc10 joined
nwc10 not sure if this is totally helpful, but since when did that ever stop me? :-) 21:08
21:09 root__ joined
nwc10 you may already know, but Damain is hunting for a word beginning with B that is semantically something akin to Bold 21:09
much like I<> "obviously" means Important 21:10
gaal brash?
nwc10 well, more in the "strong" or "important" sense than the Brave sense 21:12
gaal blunt? :) 21:13
latant?
s/^/b/ 21:14
kolibrie audreyt: cj: I'm not exactly local to Boston (Washington, DC), but would love to come up and hack
I have relatives in New Hampshire, so I get up there often enough
PerlJam nwc10: biscuits! 21:16
21:16 root____ joined
nwc10 in the plural? Not that I object to more than one, of course :-) 21:17
gaal brazen... 21:19
boomy
okay, i made that one up 21:20
PerlJam brash
bare
bra
gaal barcrawling
PerlJam binkidoodle
belligerent
But I like biscuits best
[particle] backwards 21:21
gaal badass
bumpkin
bandersnatchy
bonkworthy
[particle] better 21:22
PerlJam belle
gaal best
PerlJam best!
gaal beautiful
anatolyv binding
[particle] bigot
ooh, bilous
gaal boffin'
anatolyv bloomin'
PerlJam bum
[particle] blue 21:23
gaal buttered bread
anatolyv butch
nwc10 I'm sure acme would object to "blue"
gaal bulldozer
[particle] bond
PerlJam boast
[particle] B<james>
anatolyv bestial
gaal bulbous 21:24
anatolyv boldified!
gaal bwahaha
[particle] brave
gaal byzantine
nwc10 bodacious
anatolyv ooh
PerlJam Is someone recording the good B-words to give to Damian? ;) 21:25
[particle] bulky
gaal buxom
[particle] bullock
anatolyv barmaidy
21:25 weinig|bbl is now known as weinig
gaal ballsy 21:25
ballistic
[particle] business
gaal beatific 21:26
big
21:26 root_ joined
anatolyv boingboing 21:26
[particle] bungled
gaal booty
bling
backtrackfree
cjeris braxty 21:27
anatolyv bastard
PerlJam busty
[particle] bare
anatolyv bylined
gaal bipartisan
PerlJam beany
[particle] bullsh*t
anatolyv boojum 21:28
21:28 DiegoAltheman joined
gaal x3 21:28
21:28 DiegoAltheman left
[particle] balk 21:28
gaal boding 21:29
anatolyv borked
gaal b0rk.b0rk.b0rk
but the chocolate bon the boose
[particle] baboonish
gaal bonoboish
anatolyv bathetic 21:30
[particle] backfromthefuture
gaal bombastic
[particle] babyback
bebolden
anatolyv boorish
gaal besmirched 21:31
anatolyv bedridden
[particle] bewildered
gaal bespoke
bewitched
bothered
[particle] beebopalula
anatolyv bearlike
beerlike, too
[particle] beerrun 21:32
gaal beelike whilewe're at it
PerlJam bastion
[particle] baconian
anatolyv bust
gaal beerrun? isn't that the first word of a joyce novel?
[particle] all you need is a ten and a fiver, a key and a car, and a sober driver! 21:33
gaal barrage
anatolyv beanified (compiled to jvm)
[particle] baptized
anatolyv bostonish
gaal bostonian (while we're doing novels) 21:34
anatolyv baroque
gaal like, beep beep beep beep beep
anatolyv basta 21:35
gaal base # belong etc.
mr_ank ?!
lambdabot Maybe you meant: . v
anatolyv bandaidish 21:36
gaal backlog, damian. if you can't find something here I think you're busted.
21:37 bpphillips left
mr_ank Why limit yourself to English? There are several other useful languages. Try swahili. 21:37
gaal bas # arabic 21:38
PerlJam boss
gaal (for "enough")
PerlJam bueno
mr_ank I vote for "bueno"
bueno++ # es bueno
gaal just a day for a bananafish 21:39
bye! &
21:50 penk joined 21:52 DiegoAltheman joined, DiegoAltheman left 21:53 DiegoAltheman joined
nothingmuch moosen 21:55
22:06 larsen_ joined
dmq brazen is a synonym of bold. 22:07
22:14 larsen__ joined 22:16 rodi left
follil bluff, bold, sheer -- (very steep; having a prominent and almost vertical front; "a bluff headland"; "where the bold chalk cliffs of England rise"; "a sheer descent of rock") 22:17
22:20 avarab joined 22:25 ashelyb_ joined 22:26 penk joined 22:31 avarab_ joined, gnuvince joined 22:32 badalex joined 22:41 avar joined 22:52 avarab joined 22:53 Limbic_Region joined 22:57 nicodaemos joined 23:03 avarab_ joined 23:04 eggzeck joined 23:12 avar joined 23:19 cjeris left 23:57 mako132_ joined
nothingmuch audreyt: ping 23:59
oh crap, what was I going to say =/

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!