pugs.blogs.com/dist/Perl6-Pugs-6.2.13.tar.gz released! | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com
Set by audreyt on 17 October 2006.
00:06 mako132 joined 00:09 nicodaemos joined 00:11 Limbic_Region joined 00:16 mr_ank joined
Limbic_Region audreyt - I am at home with a fairly beefy box (1GB of ram, dual processor) if you would like to do any testing with various doodads and their affect on smoke times 00:16
00:20 mr_ank joined
audreyt Limbic_Region: my running hypothesis is the if -> $x {} 00:23
chance
er change
but we'll see... a sec
Limbic_Region I can always time different rev levels too
Limbic_Region is happy to do whatever now that the pumpkins have been carved 00:24
though I did forget to eat
audreyt ok. reverting r14489 alone
and rm blib6/lib/*yml, will test the hypothesis
my brain is influenced with ascoma of tuber fungi, so it's not full functioning yet 00:26
a sec :)
Limbic_Region how does one revert using svn ? 00:28
Limbic_Region has only done that once
audreyt svn merge -c -14489 .
something like that
mr_ank i've been having major issues trying to svk sync the whole repository 00:29
clkao what kind of issue? 00:30
mr_ank my newest error: "Incomplete data: Delta source ended unexpectedly"
clkao btw, replay support landing trunk soon
audreyt are you using latest SVN::Mirror?
it's resolved a bit ago
clkao this is ra_log session timeout. just resync ;)
audreyt clkao: ooh good. will you have some cycles this week to setup commitbit on feather and/or newhost?
mr_ank clkao: every time i resync i get that :( i'll try updating some of svk components
audreyt I'm inclining toward feather for now
clkao audreyt: sure 00:31
audreyt: you want to install commitbit ? you got root right?
audreyt cool. say tomorrow? (12~20hr from now)
I got root
I need to hack in nick support
I can do that myself, but pointers will be appreciated
Limbic_Region audreyt - give me circa 30 minutes - want to let the current smoke finish 00:32
00:32 nperez joined
audreyt sure, no rush 00:32
audreyt is catching up on mail etc
and on this hyperhash thing 00:33
ashleyb <newbe-question> I'm trying to use P::C::R with "complemented character classes" (as S05 puts it) like '<-alpha>', does the comment in the P::C::R TODO stating "-add 'negate' node to Regex", mean that this is not yet implemented?</newbe-question>
audreyt ?eval 1 ~~ /<-alpha>/
00:33 evalbot_r14538 is now known as evalbot_r14556
evalbot_r14556 *** Cannot parse regex: <-alpha>ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤) 00:33
audreyt ?eval 1 ~~ /<alpha>/
evalbot_r14556 Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
audreyt not yet implemented it looks like. you want to hack it in? :)
ashleyb yes and no -- It'd take me a few days as I don't know svk/svn or the code yet. 00:34
and I'm not sure I have a few days.
audreyt you got a commit bit? :) 00:35
ashleyb no
audreyt your email addr?
ashleyb ashleyb.moniker aatt gmail.com
audreyt welcome aboard 00:36
svn crash course at search.cpan.org/dist/Perl6-Pugs/doc...c/Hack.pod
lambdabot Title: Pugs::Doc::Hack - How to hack on Pugs - search.cpan.org, tinyurl.com/yjq4uw
audreyt add yourself to AUTHORS and "svn ci AUTHORS" to test commit; to hack P::C::R cd into perl5/Pugs-Compiler/ 00:37
er perl5/Pugs-Compiler-Rule
?eval 1 ~~ /<-[aeiou]>/
evalbot_r14556 Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 1, ā¤ str => "1", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
audreyt it looks like negated charclass is already in 00:38
the desugaring for /<-foo>/ is just /<foo>./
er I mean
/<!foo>./
?eval 1 ~~ /<!alpha>/ 00:39
evalbot_r14556 *** Cannot parse regex: <!alpha>ā¤*** Error: Error: unknown node: $VAR1 = undef;ā¤ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
eviltwin_b wonders if that document overs the evil he's currently comtemplating 00:40
hm, nope, that's the one I already looked at
audreyt so hacking in <!foo> seems to be sufficient. you can also ask fglock for more directions beyond the ones in README 00:41
ashleyb thank you.
audreyt np :) have fun 00:42
Limbic_Region audreyt - do you know if allison was aware of the threads, parrot, and fear for the future thread at PerlMonks? 00:51
audreyt Limbic_Region: I have no idea, though I'm reasonably sure chromatic will keep her informed 00:53
Limbic_Region well, since she solicited comments I replied with the thread
audreyt that's cool
Limbic_Region not sure it was such a great idea though
I would say more than one contributor to that thread hasn't maintained objectivity 00:54
and drawing more attention to it might bring more of that
mr_ank (clkao: btw, congrats on SVK... /me loves it)
audreyt++ # SVN::Mirror upgrade fixed the issue :) 00:55
audreyt mr_ank: well, yeah, basically for that two days the entire #perl6 yelled at clkao :)
clkao++ # responsive maintainer
eviltwin_b so, what is the right way to define a global ($*BASETIME) such that it won't be overridden by the Prelude? 00:57
(I was a little surprised that marking it as a constant --- which we want anyway --- didn't prevent it from being overwritten) 00:58
00:58 thepler joined
eviltwin_b but I was unable to make sense of how the Prelude gets incorporated into the environment. 00:58
I know how to hack the environment directly, but somehow I don't think that's how it's supposed to be done 00:59
audreyt where does Prelude override BASETIME?
you want it for -M ?
(sorry I lost context) 01:00
eviltwin_b if I add $*BASETIME to the initial environment it gets dumped to the Prelude next time it's generated, then that value is loaded on top of it in later runs
yes, for -M/-C/-A
audreyt ok. the answer is adding it to Pugs.AST._reserved 01:01
eviltwin_b I was getting inconsistent behavior from the command line vs. the test script until I realized that because I was working from an uninstalled pugs the command line didn't load the prelude, but the test scripts do -Iblib6/lib so they do get it
ah.
thanks. audreyt++
audreyt np :) add a haddock comment to filterUserDefinedPad and _reserved while you are at it?
eviltwin_b ok 01:02
audreyt eviltwin_b++
Limbic_Region audreyt - I think I messed up 01:03
I was about to revert everything back 01:04
but apparently you just want that one change reverted?
if so - should I svn up back to current and then do that svn merge thingy?
01:05 OppaiSan joined
audreyt just want that one change 01:05
yes.
01:05 OppaiSan left
eviltwin_b notes that he'd thought the inevitably-fatal "how hard can it be?" when he noticed the -M/-C-/A tests were aded but not planned so t/operators/filetest.t was failing after all tests completed :) 01:06
gnuvince What's a blib? 01:07
01:08 integral joined
eviltwin_b perl modules have "always" staged into blib/lib (blib6/lib for pugs) during builds, so they can be easily included for testing and eventually installed from the blib tree 01:09
not sure what it stands for, my WAG is "build library"
Limbic_Region ok - svn merge requires to rev levels so I don't think that is what I need 01:10
audreyt -c only want one
gnuvince eviltwin_b: thank you.
audreyt but maybe that's 1.4ism
mr_ank Limbic_Region: svk revert ?
or have you already checked in? 01:11
audreyt I'll need to bbl
Limbic_Region mr_ank - I am trying to be current but then apply a single change from a past revision
audreyt instead of staying up and negating jetlag adjustment I've done so far :)
*wave* &
Limbic_Region audreyt - I will figure it out and /msg you
audreyt 'k
01:11 nipotaway is now known as nipotan
audreyt Limbic_Region++ 01:11
gnuvince Is there a method/function to have a quick description of a function/method? Like docstrings in some languages 01:12
mr_ank Limbic_Region: what i do when that happens is just svk checkout -r <old revision> <file> ... i think
i haven't done it in a while 01:13
i just copy the file somewhere else and grab an older version
Limbic_Region mr_ank - that won't work
I need everything current except that one rev
mr_ank you mean except for one file?
Limbic_Region no
for one rev
or to say it a different way 01:14
I need to be current and then revert just one change
mr_ank ah
like "go to current -1"? 01:15
Limbic_Region except it isn't -1
mr_ank i don't quite follow :/
Limbic_Region current is 14556
the change I need to revert is 14489 01:16
eviltwin_b he wants an old revision to be removed from the current tree, keeping everything after it
Limbic_Region I only want to revert change 14489 - not everything in between
eviltwin_b I don't think any of rcs/cvs/svn handle that very well, and in fact it seems to me like a Hard Problem in the case where later revs modify the same code affected by the obsoleted rev
Limbic_Region I would do it by hand but 4 files are involved
audreyt seems to think it won't be an issue with this specific problem 01:17
mr_ank but why work in change 14489?
does people checkout old revisions like that?
Limbic_Region mr_ank - there is a performance problem we are tracking down
and that's where she suspect it lies
mr_ank Ah. I get it 01:18
eviltwin_b would think the easy way is to svn diff that one revision (-r14489:14490) and then apply it to HEAD with "patch -R" 01:19
er.
-r14488:144809
gah
-r14488:14489
Limbic_Region eviltwin_b - do what exactly? 01:20
step 1?
step 2?
etc
Limbic_Region is about to give up
eviltwin_b svn diff -r14488:14489 | patch -R 01:21
use svn diff to extract that one revision, then patch -R applies the diff in reverse to effectively remove those changes from the current tree
if the "patch -R" fails then you have a conflict caused by a later rev modifying the same code 01:22
01:22 dmq joined
eviltwin_b depending on how svn was built you might also want -c or -u in the svn diff arguments (look to see if the default output of svn diff includes context --- I hope that'd only be an isue if it was built on e.g. a Solaris machine without GNU diff installed) 01:23
Limbic_Region I think it worked 01:24
patch -R couldn't find the files
but I think that was a delimiter issue
eviltwin_b hrm.
Limbic_Region it prompted me
and it said success after entering it using winblows \
eviltwin_b ah
hm, might want to take that up with the svn developers --- arguably it should use the host pathnames 01:26
(unless you specify a repossitory path) 01:27
Limbic_Region patch is a mingw thing
not svn's fault
at least I think
eviltwin_b sure, but svn should output diffs against a local tree using the local filename conventions
SamB ah, no! 01:28
portable patches are best
eviltwin_b hm, ok, that also is arguable
SamB after all, one of the most common things to do with them is email them
eviltwin_b so either patch needs t deal with unix pathname separators or svn diff might want an argument to output with ocal conventions
Limbic_Region grr - I don't know 01:29
eviltwin_b I guess the former could be argued as preferable for portability reasons
Limbic_Region I am going to give up for tonight
SamB or maybe svn should have an "apply" command
Limbic_Region is having other issues and doesn't want to screw things up
eviltwin_b you might spot-check one of the files to make sure it actually did the right thing
eviltwin_b knows that one, was thinking much the same thing before he realized why $*BASETIME was messed up
svn apply would seem redundant given patch 01:30
SamB but, we have svn diff and diff... 01:31
Juerd gnuvince: Well, Perl 6 will have POD too.
SamB also, svn apply would be the simplest way to deal with a number of issues
eviltwin_b sure, but unlike patch, svn diff needs access to the repo
SamB such as metadata changes
actually no 01:32
svn diff runs totally out of the working dir ;-)
eviltwin_b metadata's a good point, but I don't think svn diff even tries to deal with that
SamB new and improved over CVS ;-)
eviltwin_b the working dir + the cached stuff that diff doesn't know about
(and shouldn't)
SamB yes, true 01:33
svn apply could just use patch(1) on the relevent files, in theory
but the SVN people would never do it that way
Limbic_Region @tell audreyt couldn't figure out how to revert just that change - svn diff 14488:14489 | patch -R suggested by eviltwin_b didn't seem to work correctly 01:34
lambdabot Consider it noted.
eviltwin_b thinks he won't be trying to make any changes tonight either; badly timeshifted 01:35
might play with another checkout to see what happened with the patch thing, in theory it should work
01:36 AzureStone joined
eviltwin_b hrm, I bet -p0 01:36
01:36 AzureStone left 01:38 swo joined
eviltwin_b yep, patch -p0 -R works here 01:38
forgot patch deletes all leading path components unless you use -pN to limit it 01:39
01:40 mjk joined
eviltwin_b @tell audreyt the correct invocation is "svn diff -r14488:14489 | patch -p0 -R"; I'd forgotten the -p0 01:41
lambdabot Consider it noted.
01:42 swo joined 01:49 lambdabot joined 01:51 Limbic_Region joined
Limbic_Region eviltwin_b - the -p0 with patch seemed to do the trick 01:52
so I am running the smoke now
01:57 gunya joined
eviltwin_b yay 01:58
Limbic_Region spot checking indicates everything worked 01:59
@tell audreyt eviltwin_b++ ammended the suggestion to | patch -p0 -R which did appear to work, smoking now 02:00
lambdabot Consider it noted.
cmarcelo nothingmuch: thanks. are you still there? 02:12
nothingmuch cmarcelo: surprisingly yes 02:13
but i am half in bed
nothingmuch is obsessing over email post shower
cmarcelo hmm.. know when will you be online tomorrow? 02:15
nothingmuch IDT 11-17 or so
then gaal et al and I are going hiking 02:16
overnight + 1 day
cmarcelo hmm.. how many hours from now?
nothingmuch it's 4:!6
you should email me questions if there's anything you'd like to ask
cmarcelo ok
nothingmuch that way you don't have to hunt me down 02:17
(sometimes i ignore IRC if i'm concentrating)
also, if I answer in a cleaner way we can wikify this into some doc 02:18
cmarcelo sure :)
nothingmuch good night 02:25
cmarcelo night
02:36 Limbic_Region joined 02:58 swo joined
Limbic_Region @tell audreyt typo detected far too late svn diff 14488:14480 instead of 14488:14489 - oops, will try again tomorrow night if you haven't tracked it down yourself by then - & sleep 03:08
lambdabot Consider it noted.
03:25 esstone joined 03:59 stevan joined 04:00 justatheory joined 04:27 penk joined 04:30 hexmode joined 04:36 weinig_ joined 04:55 gunya joined 04:56 mjk joined
TimToady will try it here. 05:11
05:29 awwaiid_ is now known as awwaiid 05:38 SubStack joined
TimToady removing if x -> $y {...} seems to have made no difference at all. 06:02
06:14 ruz_ joined
TimToady @tell audreyt removing if x -> $y {...} seems to have made no difference at all. 06:38
lambdabot Consider it noted.
06:50 marmic joined
meppl guten morgen 07:17
07:19 gunya joined 07:32 Daveman joined 07:44 eggzeck joined, kanru joined 07:48 nusgnaf joined 08:15 mr_ank joined 08:22 bernhard joined 08:23 kensanata joined 09:06 penk joined 09:12 ruz joined, drrho joined, elmex joined 09:27 kane-xs joined 09:37 buetow joined 09:42 pjmm joined 09:48 pjmm left 09:50 ruz_ joined 10:08 Yappo joined 10:13 andara joined 10:18 dduncan left 10:33 ruoso joined 10:45 nipotan is now known as nipotaway 11:15 xinming_ joined 11:20 fglock joined
cmarcelo fglock: ola 11:21
fglock cmarcelo: oies 11:24
cmarcelo: will we meet tomorrow? 11:25
cmarcelo fglock: probably Thursday morning I'll show up at holiday inn. and please IM your cell phone (if any) to make contact easier. btw, there's a good chance that a friend will lend me a notebook :) 11:27
fglock cmarcelo: cool - just ask for us at the hotel desk 11:30
cmarcelo fglock: ok. what time will be a reasonable one to show up? [nota bene: I won't be staying at the hotel] 11:31
fglock cmarcelo: I get up early :) 11:32
you could have breakfast at the hotel with us 11:33
cmarcelo cool :) around 8h30, then?
fglock ok :) 11:34
11:46 nusgnaf joined 11:48 thepler joined 11:51 andara left 12:09 Psyche^ joined 12:12 chris2 joined
svnbot6 r14557 | fglock++ | Pugs-Compiler-Rule - fixed <!...> syntax tree; added test for token{<!alpha>} 12:12
12:13 norageek joined 12:14 lisppaste3 joined, FOAD_ joined, nothingmuch joined, statico joined, rafl joined, prism joined, GeJ joined, qmole joined, wilx joined, nothingmuch joined, wilx joined, qmole joined 12:15 prism joined 12:25 Psyche^ is now known as Patterner
svnbot6 r14558 | fglock++ | Pugs-Compiler-Rule - implemented regex{<!...>} - added tests 12:27
cmarcelo @tell nothingmuch re MO: shouldn't MO::Run::Abstract::ResponderInterface role require a 'dispatch' method? 12:32
lambdabot Consider it noted.
nothingmuch yes, it should
lambdabot nothingmuch: You have 1 new message. '/msg lambdabot @messages' to read it.
nothingmuch @moosages
lambdabot cmarcelo said 26s ago: re MO: shouldn't MO::Run::Abstract::ResponderInterface role require a 'dispatch' method?
cmarcelo tks 12:35
12:36 andara joined
fglock inside rules, does char classes operations have a precedence? like <+foo+bar-alpha> - is '-alpha' removing alphas from (foo+bar) or from the input string? 12:39
ok, it's removing alphas from (foo+bar) 12:42
12:42 mako132 joined 12:47 anatoly_ joined, anatoly_ is now known as anatolyv
nothingmuch hola anatolyv 12:47
anatolyv hola 12:48
nothingmuch i mean
yo
anatolyv right!
ok, possibly a stupid question 12:50
is if (as a control statement) properly an operator?
in p6, that is. 12:51
audreyt it's properly a function
lambdabot audreyt: You have 5 new messages. '/msg lambdabot @messages' to read them.
audreyt &statement_control:<if>
but it's actually a macro function 12:52
though if you apply it at runtime you still get what you expect.
not yet implemented in Pugs.
anatolyv audreyt: so eventually i should be able to $a = if <whatever>, but currently unimplemented in pugs, right? 12:53
audreyt: hi, btw!
fglock it's not an operator at the syntax level - in the sense that you can't apply a function on it, like 1+if... 12:54
audreyt anatolyv: you can't. you need to say
$a = &statement_control:<if>(...)
or, better (and implemented in pugs)
$a = do if ...
as "do" transforms a statement into an expression
?eval my $a = do if $a { 100 } else { 1000 } 12:55
anatolyv audreyt: that's my real question. what are statements that are not expressions? if etc. for control? anything else?
12:55 evalbot_r14556 is now known as evalbot_r14558
evalbot_r14558 \1000 12:55
audreyt anatolyv: things put into statement_control are statements
currently: for/loop/repeat/if/while/when/maybe/default 12:56
also package/has/is/does/use/no/trusts
all of them can be expressionified with do 12:57
?eval do use v6; 12:58
anatolyv audreyt: ah, that makes sense, thanks.
evalbot_r14558 undef
audreyt np :)
fglock please check: is <+foo+bar-alpha> the same as [ [ <before <foo>> | <before <bar>> ] & <!alpha> ] .
audreyt looks correct to me. 12:59
cmarcelo nothingmuch: there is a reason to separate MO::C::Class and MO::C::Abstract::Class roles? 13:00
s/there is/is there/
nothingmuch Abstract::* provides no functionality
audreyt fglock: I'm room 816 btw. feel free to knock tomorrow morning
nothingmuch in theory
fglock ok!
nothingmuch i kinda screwed that up a bit
MO::C::Class is getting slowly refactored
it's *WAY* too big right now 13:01
anatolyv audreyt: what's 'do'? a prefix macro op?
cmarcelo I saw it :) I'm working just w/ a subset of it, the enough to make si.t and mi.t (handwaving attribute / roles / etc).. 13:02
audreyt anatolyv: a prefix with a "is parsed" equiv to <stmt>
13:05 HiTech69 joined, orafu joined
cmarcelo audreyt: ola. I "reverted" your patch since I could get moh compiling and keeping the functionality.. 13:05
s/could get/got/ 13:06
audreyt cool
cmarcelo: so you check in the day after tomorrow? 13:11
where are you staying here?
13:12 orafu joined 13:15 Limbic_Region joined
audreyt Limbic_Region: it's not the if-> change; I benched and Timtoady also did 13:18
cmarcelo audreyt: yes, thursday. at my parents' house (a little bit far from Anhembi). but I'm going to bug you and fglock at the hotel :o) 13:19
audreyt good good :)
Limbic_Region audreyt - ok, well what I can do tonight is a binary search to isolate the rev that is the culprit
oooh, not tonight - tonight is halloween
audreyt I check out at 6th so we have plenty of hackathon time
even more if you'd like to come to Rio with me for the next 4 days after that... :D 13:20
fglock: oh that reminds me. should I book a air ticket online for Rio? is there a website you recommend?
13:20 orafu joined
audreyt or is there fast train (or something else)? 13:20
ingy hi audreyt
audreyt ingy: greetings! unfortunately not making it to Seattle Hackathon 13:21
ingy audreyt: it is canceled
audreyt oh, k.
Limbic_Region audreyt - ok, will definately play when I get a chance. I would do it here at work but this is such an awful machine it would take forever
ingy is in Dublin
svnbot6 r14559 | fglock++ | Pugs-Compiler-Rule - parses '... & ...'
fglock audreyt: voegol.com.br/
lambdabot Title: .....:GOL:..:LINHAS A&Eacute;REAS INTELIGENTES:......
audreyt Limbic_Region: k
fglock audreyt: they usually have fair prices 13:22
audreyt ingy: your quite hospitable room mdate invited me to stay at your residence sometime in the future, though, so definitely some day :)
ingy audreyt: you are always wanted and welcome. I hpe you know that by now 13:23
13:23 orafu joined
ingy wtf: > sudo port -vvv install ghc 13:23
---> Building ghc with target all
make: *** No rule to make target `all'. Stop.
Error: Target com.apple.build returned: shell command "DYLD_FALLBACK_LIBRARY_PATH=/opt/local/lib make all" returned error 2
Command output: make: *** No rule to make target `all'. Stop.
Warning: the following items did not execute (for ghc): com.apple.activate com.apple.build com.apple.destroot com.apple.install 13:24
Error: Status 1 encountered during processing.
cmarcelo audreyt: let's see what happen :) will there be a perl event on Rio or sth like?
audreyt fglock: will there? :) 13:25
ingy: you want the bindist I made...
haskell.org/ghc/download_ghc_66.html
lambdabot Title: GHC: Download version 6.6
ingy audreyt: you are correct 13:26
fglock audreyt: you will meet them at CONISLI (mk and MDA)
so you can make arrangements
audreyt mainly need to sort out accomodation 13:27
cool
what's the diff between Ida / Volta tickets? 13:28
fglock departure/return 13:29
audreyt aha 13:30
fglock there is a flag in the upper right corner, it should have an english option (it doesn't work here)
13:30 stevan joined
fglock audreyt: I'll be in Sao Paulo until nov.6 at night 13:31
audreyt english doesn't work here either
cmarcelo audreyt: voegol.com.br/INT/ 13:32
lambdabot Title: .....:GOL:..:LINHAS A&Eacute;REAS INTELIGENTES:......
audreyt cool but I'm well into ordering now
babelfish++
k done 13:39
13:40 thepler joined
fglock I wonder how captures+conjunctions interact 13:47
svnbot6 r14560 | fglock++ | Pugs-Compiler-Rule - token{ ... & ... } works 13:48
fglock it doesn't need '&' for implementing char classes, actually - since <before> and <!...> don't increment .pos 13:50
audreyt so you can just stack them 13:51
fglock yes 13:52
dmq ive got a working implementation of a COMMIT verb for perl5 regexes now. 13:53
Assuming somebody here is interested.
audreyt YES!
fglock :) 13:54
dmq oh, and regex recursion are now optimised properly so you can do embedded rules and still be efficient.
audreyt oh wow.
that's like music to my ears :)
dmq im working on a CUT operator too.
fglock dmq: cool!
dmq which is a lot harder unfortunately. im still trying to get it all sussed out. 13:55
fglock dmq: I added a negate() sub to Pugs::Runtime::Regex a while ago, please merge 13:56
dmq huh?
fglock dmq: are you working with Pugs::Compiler::Rule ?
oh, you mean actual perl5 :) 13:57
not perl6->perl5
dmq yes indeed.
I mean real perl5.
audreyt but perl6->perl5 can benefit from a much better embedding
since the backtrack logic can be done in C now for subrules 13:58
dmq PAT_A (?COMMIT) PAT_B
make the entire pattern fail for the current startpoint if the commit has been entered and PAT_B fails. 13:59
will that help? 14:00
audreyt yes. 14:01
does it interact with subrules well?
I mean can I have a boundary so COMMIT only works within it?
dmq ah, no, not currently.
but ive been thinking of ways to do so. 14:02
for instance i was think / PAT_A (?MARK) PAT_B (?ROLLBACK) /
but maybe if you gave me a better example of what you want and I can see. 14:04
svnbot6 r14561 | fglock++ | Pugs-Compiler-Rule - updated TODO, ChangeLog
ingy audreyt: how do you install that osx ghc binary thing? 14:05
miyagawa_ ingy: hi
ingy there's no pkg or dmg...
miyagawa_: hi
14:06 fglock left
ingy I keep trying you on aim 14:06
nothingmuch ingy: that .tar?
it has a make install, i think
miyagawa_ ingy: yes
audreyt ingy: sh configure make install etc 14:07
14:08 hexmode joined
ingy ok 14:08
audreyt lunch, bbl 14:09
ingy audreyt: now I'm stuck at > sudo port install darcs 14:11
---> Building ghc with target all
Error: Target com.apple.build returned: shell command "DYLD_FALLBACK_LIBRARY_PATH=/opt/local/lib make all" returned error 2
Command output: make: *** No rule to make target `all'. Stop. 14:12
Error: The following dependencies failed to build: ghc
Error: Status 1 encountered during processing.
:(
14:15 cjeris joined
cmarcelo ingy: looks like ports wants to build ghc itself. is there a way to tell him that you already have a non-ports version? 14:19
or a --nodeps thing
14:22 frederico joined 14:37 lanny joined
lanny Limbic_Region and/or audreyt: if you give me a rev range I can smoke em all in background at work today. 14:38
Limbic_Region lanny - that could be a hundred or so revs
lanny Takes me about 20 minutes to build and smoke. Maybe 25. 14:39
Limbic_Region and how many can you do concurrently?
lanny So can squeeze in close to 3 an hour. And I'd use a binary search as well
Limbic_Region ahh - ok, that makes sense 14:40
lanny Never tried concurrently. My new game box (which sadly is far better than any un*x box I can lay my hands on)
I can't seem to get GHC to use multiprocessors btw. mingw but -GHCRTS carps at me 14:41
Limbic_Region oh, I was thinking that you had multiple machines
lanny Nope. But it would be easy to script.
To do the search and compare times
So any suggestion on an earliest rev to look at? 14:42
I can search the logs and find the junction stuff if it's in that timeframe
Limbic_Region I was just looking at that
or rather, I was doing the math to see how large a range you could cover in a reasonable amount of time with the bin search 14:43
lanny Hmm. Not enough coffee to do the exp() vals in my head to tell you. :)
Limbic_Region anything less than 256 should be ok right - 2 ^ 8 = 256 so 8 builds/smokes 14:44
lanny I'm sure I could crank out 20 or so in the next 10 hours. 14:45
Limbic_Region min = 14414, max = 14556
lanny So should easily bracket wherever the problem is
Got it. Gotta get girls in costumes for school. Cheers.
Limbic_Region cheers
@tell lanny if your smoke hangs on any tests then make sure your script removes thoses tests as part of the smoke process 14:46
lambdabot Consider it noted.
14:48 lichtkind joined 14:54 elmex joined 15:00 rlb3 joined 15:04 bonsai joined 15:15 c6rbon joined 15:38 jferrero joined, ruoso joined 15:39 vel joined 15:40 kanru joined
cmarcelo @tell audreyt new moh patches in repo.. i'm going this direction for classes, avoiding IORef et al. sounds sane? 15:46
lambdabot Consider it noted.
15:50 elmex joined 15:53 Pomin joined 15:56 justatheory joined 16:11 penk joined
rgs ?eval "a" ~~ /(.)/ 16:43
16:43 evalbot_r14558 is now known as evalbot_r14561
evalbot_r14561 Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 1, ā¤ str => "a", ā¤ sub_pos =>ā¤ (Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 1, ā¤ str => "a", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤ ),), ā¤ sub_named => {}ā¤) 16:43
16:51 nothingmuch joined 17:00 araujo joined 17:06 fglock joined 17:15 mako132 joined 17:25 iblechbot joined 17:37 GnuVince joined 17:47 explorer__ joined 17:48 ruoso joined 18:01 b_jonas joined 18:06 penk joined 18:12 DHGE joined 18:32 ofer1 joined 18:46 eggzeck joined 18:49 TSa joined 18:53 rindolf joined 19:01 weinig_ is now known as weinig 19:23 hexmode joined
audreyt fglock: when will the opening talk be? 19:48
lambdabot audreyt: You have 6 new messages. '/msg lambdabot @messages' to read them.
audreyt is it Nov3 morning?
fglock audreyt: yes 19:49
audreyt is it for YAPC::SA only or for CONISLI?
audreyt is 0-slides-must-JIT-now mode
TimToady: do you have existing material for "A filosofia da linguagem Perl"?
fglock both - CONISLI hosts YAPC::SA, the opening talk is for the general audience 19:50
audreyt notes fglock's face, esp. hair, resembles a certain Schwern
got it
fglock audreyt: we just figured out some name for TimToady's talk :)
audreyt and how well should I expect the audience's English tokenizing ability? 19:51
fglock s/figured out/made up/
audreyt and how many minutes must I remain on stage? :)
19:51 mr_ank joined
fglock there will be translation, just speak reasonably slowly 19:51
audreyt oh!
I'll have an interpreter
nice
fglock www.conisli.org/grade/grade.html
audreyt interpreters are good, if slow
lambdabot Title: ...:Grade PALESTRANTES:...
audreyt so 1hr 19:52
will you be my interpreter?
fglock about that, some time for questions
they usually hire a professional interpreter
audreyt also maybe I can ask you or cmarcelo to translate all my slides to Portuguese...
buubot correct syntax is: translate to/from language string
audreyt (in Tokyo my talk was all in Japanese slides)
fglock sure, no problem translating it, I think 19:53
audreyt cool. also is it too late to sign up for lightning talk?
fglock btw, I'll take a tablet pc
no that would be nice - we have a few open slots 19:54
please talk to Lorn about that
audreyt I'd like to present ppencode on another continent (to make takesako-san happy)
TimToady audreyt: no, but feel free to steal from anything I've said before, or from anything you think I might say in the future. :)
fglock TimToady: :)
audreyt so they'll be continuously perfect... 19:55
or perfectly continuous...
fglock audreyt: I don't have my slides yet, either
audreyt I went to this interesting molecular-gastronomic restraunt for lunch today. their slogan is "Dominus, Optimus, Maximus" which translates in my mind to "power to the -Ofun" 19:57
and indeed the lunch was a lot of fun and filled my mind with strange talk ideas...
19:58 buetow joined
TimToady "Optimizing for Maximal-Order Perl" --Dominus 19:58
audreyt TimToady++
fglock cool - it would also be nice to include food photos, since you'll be talking just before lunch :)
audreyt fglock: so, about the professional interpreter. can I assume some CompSci training/vocab?
TimToady generally most such terms are transliterated anyway 19:59
audreyt gastronomy is one theme TimToady's talk havn't yet explored, I believe...
fglock audreyt: let me ask the organizers, I think so
audreyt k, that's good then
TimToady: yeah, I had horror experience in .tw conferences where they didn't transliterate them, though... 20:00
though pretranslating the slides to Portuguese can likely prevent that 20:01
TimToady really depends on the culture and on the field of endeavor. in Japanese, you'd be in trouble in chemistry but not in comp sci, for the most part.
fglock yes, it would help the interpreter - or distract her :)
TimToady simultaneous translation, I presume 20:02
fglock yes - that's pretty difficult 20:03
TimToady easy things easy and hard things possible is very much like a food pyramid... 20:06
'course, the problem in Brazil is that they think that the base of the pyramid should be meat... 20:07
audreyt fglock: are people in love with McDonalds here such that if I compare fast food chains to proprietary software they get upset?
20:07 buubot joined
fglock audreyt: no, no problem :) 20:07
Limbic_Region wonders how lanny made out 20:10
audreyt - have you heard?
audreyt no?
with an simultaneous-translation slow-speaking setting, TimToady's usual SotO style seems far more suited than my usual structure-the-talk-as-many-lighting-talks style 20:11
Limbic_Region I noticed you had @messages and I have been AFK most of the day is all
fglock SotO ? 20:12
audreyt State of the Onion
basically, live performance on free association.
fglock try just optimizing for fun 20:13
just-in-time
audreyt which is a style I've never practiced
which means it's fun :)
fglock once I was called to replace a speaker who would give a talk on Bioinformatics - I had a few hours to read a book on it :) 20:14
audreyt nice :)
once I had to finish the second half of my talk in the 10min break after the first half... 20:15
Limbic_Region this one time...at band camp.....er ah $conference
fglock a person in the audience told me that the talk was too difficult - I blame I didn't have time to simplify it :) 20:16
TimToady lunch &
audreyt indeed the shorter the talk is the harder it is to prepare
lightning talk is the most difficult form of all
TSa HaloO 20:20
audreyt TSa: greetings! long time no see 20:21
TSa have we seen us before?
fglock otoh, I've made some quite nice one-slide presentations
20:21 rafl joined
audreyt TSa: you were in #perl6 about a year ago, iirc. 20:21
TSa that's right
svnbot6 r14562 | kudra++ | Placeholder 20:22
TSa I'm actually unsure how I get this TSa account
at login it says that this is not my nick
audreyt you can either ignore that or change to another unregistereized nick 20:23
by ignoring that you can't privately /msg people
20:23 buubot joined
audreyt but that's about the only effect 20:23
TSa but what if the real owner shows up?
which might be me with a lost password 20:24
[particle] TSa: try "/msg nickserv identify TSa"
i think it defaults to your username as password
TSa it doesn't get that password 20:25
what are your local times? audreyt should be early morning, TimToady afternoon 20:27
are these the usual times here?
audreyt TSa: I'm in Brazil 20:28
it's 5:28pm
TSa nothing particularly Perl 6 on the agenda 20:29
I mean here on the channel
kolibrie as I was thinking about it yesterday, north and south America shift time in opposite directions at daylight savings 20:30
just makes a person like time shifting even less 20:31
audreyt TSa: no, but once tomorrow fglock arrives and the day after that cmarcelo arrives there'd be more coding
fglock I'm working on <+alpha+digits> - but I only see bugs so far
audreyt now I'm just wondering how to do a State-of-the-Onion style talk based on Gastronomy
fglock digit
audreyt surprisingly hard
TSa I don't get that. A talk about gastronomy from you, audreyt? 20:32
fglock just make a simple skeleton presentation - the hardest part is that you will not have a clue if the audience is getting it, because it depends on the interpreter
you should use a compiler instead of interpreter 20:33
audreyt fglock: which is why you'll be my first-pass preprocessor 20:35
fglock ok :)
audreyt TSa: you familiar with TimToady's style of free association Perl and an utterly unrelated discipline?
cmarcelo fglock: will the interpreter sound be over her sound? 20:36
audreyt in the previous years he did music, linguistics, chemistry, screensavers, postmodernism, and scientific american
fglock cmarcelo: they provide headsets
Debolaz galleri.andersberle.com/kryp/aaa <- Newest family member.
lambdabot Title: Galleriet :: Smōæ½xE5kryp :: aaa
TSa audreyt: I know his onion speaches from transcripts on the web
audreyt this year it's Family
so I'm trying to pick a subject TimToady haven't exercised
and Gastronomy sounds good
TSa so, something like 'Pugs is like spagetti, ... entangled but good' :) 20:38
20:39 vel joined
audreyt heh :) 20:39
quantums also entangles...
meppl _ pugs -e 'my $input; while ($input = =<>) {}'
if i just put "return" the loop ends :/ 20:40
eviltwin_b haskell coding as making Hollandaise?
audreyt meppl: for =<> -> $input { ... }
new idiom
meppl audreyt, thank you
audreyt =<> autochomps so can't use "while"
np :)
meppl ah, okay
then i can delete my line with the chomp 20:41
audreyt yup
cmarcelo audreyt: don't know what you saw but in brazil rice+beans ("arroz+feijĆ£o") are fundamental in lunch (and dinner) for many people, don't know if this happens in other cultures.. 20:47
audreyt cmarcelo: no, that's nice to know...
cmarcelo: do you know of atala/flavia's GNT culinary show? 20:48
something called Mesa para Dois
cmarcelo audreyt: i heard of but never saw it.. 20:49
audreyt ok...
fglock <-- amateur cook 20:51
svnbot6 r14563 | fglock++ | Pugs-Compiler-Rule - proof-of-concept implementation of <+alpha+digit>
audreyt fglock: oh! have you seen that show then? do they share recipe there, or do they do something else entirely? 20:52
fglock these shows usually have web sites - looking
globosat.globo.com/gnt/programas/pr...sp?gid=349 - click on "receitas" 20:53
lambdabot tinyurl.com/yxbgh3
audreyt excellent. thanks
fglock oh, you have to subscribe - but just look "receitas" on google 20:54
audreyt brings out the very old software-as-service sharing-recipe-as-sharing-source analogy
k
fglock audreyt: I'll bring my recipe archives with me - I've got about 100k recipes :) 20:55
audreyt :D
fglock too bad we can't cook at the hotel
audreyt indeed 20:56
PerlJam why can't you cook at the hotel? 20:57
you don't bring a portable kitchen everywhere you go?
;-)
audreyt "laptop"
20:57 eggzeck joined
[particle] cooking with sony! 20:57
fglock if they don't have sprinklers
20:58 buubot joined, weinig is now known as ghostWeinig
fglock audreyt: there is a partial strike on airports, maybe I'll arrive later than expected 20:59
audreyt ok
11am? 21:00
fglock yes, about that
audreyt roger
fglock home & 21:03
21:04 eggzeck joined, ghostWeinig is now known as weinig|bbl 21:05 buubot joined 21:16 buubot joined 21:21 buubot joined
cmeyer Is this supposed to be valid: 21:25
my $sq = -> $val { $val**2 }; for ^3 $sq;
audreyt no; you probably want 21:27
map $sq, ^3
cmeyer I don't need to capture the result ... just wondering if I can use 'for' to loop over an anonymous sub.
audreyt ah. then no, as for is syntactically only recognizing the block 21:30
eventually you can write
&statement_control:<for>(^3, $sq) 21:31
actually can you check if there's a test for that and if not, write one?
meppl gute nacht 21:33
TSa Is it fashionable to say good-bye? netiquette and such I mean
cmeyer Which synopsis describes statement_control, 6? 21:34
or 4?
seems like four ... reading. 21:35
TSa bye 21:36
[particle] cya tsa
TSa I'll be back
what is actually a good time to be around? 21:37
21:37 trym joined 21:38 TSa left
audreyt cmarcelo: S04 it is 21:38
[particle] hehe, wrong cm<Tab>
audreyt cmeyer: S04 it is :) 21:41
cmeyer :) 21:43
should I make a new .t file (in t/blocks ?) 21:44
audreyt sure 21:46
also while you are at it, make a new t/subs/ directory and move all t/blocks/sub_*.t into t/subs/*.t? :) 21:47
[particle] i thought all the t/subs files were copied to t/blocks some time ago 21:48
audreyt yeah but subs were really not blocks... :) 21:49
maybe t/routines/ will also do
cognominal in S06, %pairs = {:when<now> :what<any>}; # comma missing?
audreyt it's unintentional I think 21:50
fixed
cognominal ok, thx 21:51
audreyt thank you :) 21:52
cmeyer audreyt: I tried copying your syntax, but got a pugs error: *** No compatible subroutine found: "&statement_control:for" 21:56
audreyt cmeyer: yes, that's why we need a test 21:57
as it's not yet implemente
cmeyer what's the appropriate way to write tests that barf (it now says "# Looks like you planned 1 tests, but ran 0")? 21:58
eviltwin_b :todo<something>
audreyt also eval ''
cmeyer thanks 21:59
audreyt :) see t/README for more info 22:00
22:05 anatoly joined, anatoly is now known as anatolyv
anatolyv ok, another silly spec question :) the builtin operators, macros etc. - are they lexicals visible in all scopes, or dynamic in the global package? 22:07
audreyt see S02:1030 22:10
or S11:176
short answer: they are defined in ::GLOBAL but imported lexically
and you can say
no GLOBAL '&statement_control:<if>'; 22:12
lexically, presumably.
(or via some other use-but-exclude syntax)
anatolyv awesome.
TimToady lexical scopes are very important for any language that aspires to be multiple languages. 22:13
anatolyv so when i say our $x, that imports the package '$x' into the lexical scope. But when is '$x' actually entered into the package? At runtime during the 'our'? 22:14
TimToady it's entered into the package as soon as the compiler knows about it. 22:15
22:15 justatheory joined
anatolyv TimToady: so it's OK, for instance, to refer to $x in a qualified way, via the package, lexically before "our $x" is encountered, correct? 22:16
TimToady yes
anatolyv TimToady, audreyt: OK, thanks a lot, again :)
audreyt TimToady: there's no exclusion-import syntax in the spec yet? 22:17
TimToady not explicitly, but presumably 'no' could be made to work. 22:18
audreyt like it better than use foo <!bar> 22:19
TimToady :my<foo>, :no<bar>
audreyt that's a thought too...
TimToady import to nowhere...
audreyt too bad can't quite say :!my
or :!my<foo>
TimToady does Mammal[:no<placenta>] 22:21
audreyt tries to put "cuisine livre" and "joy of cooking" into a coherent thought train 22:23
this style is harder than it looks like from the audience's seat...
TimToady :O
audreyt :O<fun>
22:24 trym_ joined 22:26 Caelum joined 22:28 lanny joined
lanny Howdy. Limbic_Region (et al) I chased a red herring for a bit and still working but problem is between 14485 (duration 1076) and either 11490 (dur 3578 but had a hang and was over at my folks) or 14494 (dur 1142) 22:31
lambdabot lanny: You have 1 new message. '/msg lambdabot @messages' to read it.
lanny @mess
lambdabot Maybe you meant: messages messages? msg
lanny heh
@messages
lambdabot Limbic_Region said 7h 45m 4s ago: if your smoke hangs on any tests then make sure your script removes thoses tests as part of the smoke process
lanny At a guess looking at the logs I'd say it's -> if's even though those have been ruled out. :) 22:32
rev 14489
22:34 mugwump joined
audreyt lanny: hm. write a small loop and microbenchmark it with and without r14489? 22:37
ashleyb ?eval "attr='value'" ~~ /<-['"]>+ = (<["']>)<any-$0><$0> / 22:39
22:39 evalbot_r14561 is now known as evalbot_r14563
evalbot_r14563 *** Cannot parse regex: <-['"]>+ = (<["']>)<any-$0><$0> ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤) 22:39
ashleyb ?eval "attr='value'" ~~ /<-[=]>+ = (<["']>)<any-$0><$0> / 22:40
evalbot_r14563 *** Cannot parse regex: <-[=]>+ = (<["']>)<any-$0><$0> ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<any-$0><$0>/ 22:42
evalbot_r14563 *** Cannot parse regex: (<["]>)<any-$0><$0>ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<any><$0>/
evalbot_r14563 *** Cannot parse regex: (<["]>)<any><$0>ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<any>/
evalbot_r14563 Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 2, ā¤ str => "\"x", ā¤ sub_pos =>ā¤ (Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 1, ā¤ str => "\"", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤ ),), ā¤ sub_named =>ā¤ { "any" =>ā¤ Match.new(ā¤ ok => Bool::True, ā¤ from => 1, ā¤ to => 2, ā¤ str => "x", ā¤ sub_pos => (
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<any>+<$0>/ 22:43
evalbot_r14563 *** Cannot parse regex: (<["]>)<any>+<$0>ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<any>+$0/
evalbot_r14563 Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 7, ā¤ str => "\"xxxxx\"", ā¤ sub_pos =>ā¤ (Match.new(ā¤ ok => Bool::True, ā¤ from => 0, ā¤ to => 1, ā¤ str => "\"", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤ ),), ā¤ sub_named => {"any" => ()}ā¤)
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<any-$0>+$0/ 22:45
evalbot_r14563 *** Cannot parse regex: (<["]>)<any-$0>+$0ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤)
ashleyb ?eval '"xxxxx"' ~~ /(<["]>)<-$0>+$0/ 22:46
evalbot_r14563 *** Cannot parse regex: (<["]>)<-$0>+$0ā¤*** Error: ā¤Match.new(ā¤ ok => Bool::False, ā¤ from => 0, ā¤ to => 0, ā¤ str => "", ā¤ sub_pos => (), ā¤ sub_named => {}ā¤) 22:47
22:49 cjeris left
[particle] ashleyb: perhaps run.pugscode.org would be a better place? 22:49
ashleyb forgot about that -- thanks. 22:50
23:08 eggzeck joined
audreyt cmarcelo, fglock: is "alegria" a better translation than "divertimento" for Fun? 23:09
23:14 jrockway joined
jrockway anyone have success building pugs on openbsd? 23:15
lanny audreyt: just gonna smoke em both.
audreyt jrockway: havn't heard of a built lately. you ran into any problems? 23:16
jrockway yeah, linking pugs.new fails because p5embed.o has an __init symbol (which is also in crtbegin.o) 23:17
audreyt is this trunk or releaes? 23:18
jrockway release, 6.2.13... want me to try trunk instead?
audreyt ghc 6.4 or 6.6? 23:19
jrockway 6.6
audreyt ok. try trunk, but 90% chance not going to help
jrockway ok, i'll let you know how it goes 23:20
23:20 nicodaemos joined
jrockway probably just need to change some flags around when building p5embed 23:21
i'll look in to it
audreyt jrockway++
lanny So 14887 was duration 1075 which is in line with the smokes before it. After I see roughly 6.5% increase in smoke times which I think are located at 14488 which is smoking now. I'll post full data when done with taking my girls out for Halloween trick-or-treeting.
audreyt jrockway: also potentially reverting r14370 alone may be relevant. but changing flags in Pugs.cabal is more hopeful 23:22
lanny: so much thanks
clkao audreyt: did you say you want commitbit today
audreyt clkao: I did say but obra said they had some problems in setting up the host
and it's preferred to feather etc
how's that being sorted? 23:23
obra audreyt: ENOROBRT 23:27
audreyt: I am adding nickname support right htis moment, thought
though
audreyt nod. I took a look
not sure if it's just as simple as
htpasswd -nmb @{[$user->nick]} @{[$user->__value('password')] 23:28
in addition (or as alternative to? a config?) to ->email
(and in two other places in M::Repo too)
I guess for Pugs we always want nicks
never emails
and people can change their nicks. I think.
(the renick part is already part of commitbit I gather) 23:29
23:30 mako132_ joined
obra it's not 23:32
users can cahnge their nicks
er. sorry
I thought that renicking should be a metacommitter operation
too easy for a user to renick and lose history
audreyt fair enough. so metacommitter can renick herself or renick a normal committer? 23:33
obra yes 23:34
audreyt cool 23:35
obra it means that one person can have different nicks on different projects as well
audreyt that sounds about right
nick becomes a handle by which email links to project
clkao: what's the magic svk command to run to give a rundown of active committers? 23:36
eviltwin_b hm, might be useful 23:37
only reason I'm eviltwin_b here is that I'm mainly on this IRC network for #openafs
getting mistaken for Russ gets old
audreyt indeed
obra turns out we need it for bps now, audreyt. svn 1.4 changed auth subtly 23:38
audreyt "it"?
= nick instead of email as tokens? 23:39
obra yeah 23:43
it's now the case that svnserve when run in "requires auth" mode requires auth + valid_user. 23:44
audreyt aha.
audreyt writes some more slides extolling the virtue of dopamine 23:47
obra heh 23:48
23:55 weinig|bbl is now known as weinig 23:59 weinig is now known as weinig|bbl