pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
TimToady and your : after 'Bool::' is doing nothing 00:00
dduncan and a token bool would be a 'Bool:' <enum_bool>
TimToady that looks good
can there be space after Bool: ? 00:01
if so, that might be a rule
dduncan generally, there are no spaces between parts of simple tokens
00:01 pugs_svnbot joined
dduncan the only spaces are between collection items ... 00:01
TimToady can you say "Bool: true" 00:02
avar TimToady: Isn't it preventing backtracking?
TimToady there's nothing quantified before it to backtrack
it only controls one atom
dduncan no, just "Bool:true"
TimToady in that case token is fine
dduncan only collection types like tuple and relation would have whitespace
eg, "Relation:{ Bool:true, Int:9:42 }" 00:03
TimToady the basic rule of thumb is to consider whether the beginning and end want whitespace. tokens sometimes want whitespace in the middle but generally because they call into a subrule that would already handle whitespace
dduncan er, sort of like that
in that case, some whitespace is mandatory, but it can be any amount greater than none
in my case, the only whitespace is BETWEEN things 00:04
TimToady so '(' <EXPR> ')' is a token, despite EXPR being a rule
(in p6)
dduncan not before/after
eg, whitespace only appears after a {, before a }, and between list items
TimToady well, you can specify everything as tokens then; nothin's sez you have to use rules
dduncan okay
TimToady as avar++ points out, it's all just sugar 00:05
dduncan but where whitespace is used, its a variable amount, akin to \s+ of Perl 5 ... is that still a token?
diakopter tastes only the aspartame 00:06
TimToady token is whatever you want it to be
dduncan okay
right then, my grammar is very KISS 00:07
TimToady we're not dealing with an arbitrary distinction between a lexer and a grammar here
avar TimToady++ # including token {} which makes grammars a lot easier to optimize
dduncan yes ... actually, what I'm calling a grammer here is more just a tokenizer
TimToady no, not if you are parsing things like relations
dduncan er yes, the result is a tree 00:08
TimToady you are writing a grammar, and the lexer is just appearing magically out of it without you really having to spec it
that what all the longest-token mumbo jumbo is about
dduncan thank you for your help
TimToady np 00:09
dduncan now, about Blob literals in Perl 6 ... I recommend something in between number literals and character string literals ... 00:10
TimToady maybe b is for blob in 0b011010001100010101101010100 00:11
dduncan isn't that spelling already used for base-2 numbers?
but I was thinking something similar
TimToady a sequence of bits is how we defined blob 00:12
dduncan yes,
TimToady how is that different from a binary number? :)
dduncan but if someone wrote what you said, would it turn into a Blob or an Int?
diakopter TimToady: token integer contains the patterns of octint, decint, etc. Does it not include them by reference for optimization?
TimToady haven't really thought about optimization much with STD 00:13
in particular a non-dfa approach is pretty much guaranteed to run really slow right now.
anyway, if something is inlined, it's merely because it hasn't been abstracted out for some reason 00:14
dduncan ?eval 0b011010001100010101101010100.WHAT 00:15
pugsbot_r17185 ::Int
dduncan which is also what synopsis 2 says
so the Blob literal can't be exactly that
TimToady well, the only difference is that a blob knows exactly how many bits it has 00:16
dduncan I see a logical difference since an Int can be represented using a variety of bit patterns behind the scenes, while a bit string is exactly what it is
TimToady whereas an Int has an infinite number of bits, most of which are 0
dduncan I consider an Int itself to be radix agnostic 00:17
as a concept
and afaik Int in Perl 6 is also radix agnostic
the 0x or whatever is just a map between particular char strings and ints
TimToady so is blob, really
dduncan but part of my point is that the types Blob and Int don't have conceptual 1:1 mappings 00:18
just as Int and Str that look like Int don't 00:19
but the point here ...
is how to spell a Blob literal in Perl 6 so that .WHAT would say 'Blob'
I'm hoping the synopsis would specify that soon 00:20
TimToady well, maybe it's just an overloaded literal, like most numbers
gotta run to the store & 00:21
dduncan mind you, I do recognize that Perl 6 already treats some other types like abstractions over Blobs, or at least some operators do ... eg I think ?&, +&, ~& 00:22
though I myself consider eg Bool|Int|Blob|Str values to be all disjoint
and the above ops are just mappings
er, just do mappings
diakopter but a literal is a an immutable, and immutables are independent of representation, and so they always lose their origin representation, right? 00:23
dduncan all values are immutable, and a literal typically is a representation of a value of some arbitrary type as a character string 00:24
a literal is also a value
containers are mutable 00:25
and hold appearances of values
a value could be called a literal in some situations, or perhaps all?
diakopter dduncan: I guess what you're looking for is a container for literals? 00:26
dduncan values are independent of representation
00:26 meiss|na joined
dduncan a representation is essentially a mapping of one value to another value 00:26
eg, an Int represented as a Str
with some specified base 00:27
a Perl 6 grammar basically takes a Str value and maps it to a tree composed of values of various types
anyway, back to grammar-writing ... 00:28
or rather, a Perl 6 grammar defines a mapping between Str values that are Perl 6 code and tree values composed of values of various types 00:30
I mean, the Perl 6 grammar for Perl 6 was ... Perl 6 grammars in general aren't for Perl 6 code 00:31
no matter
00:42 Eidolos_ joined 00:44 meissner left 00:52 Eidolos left 00:53 Limbic_Region joined 00:57 Eidolos_ is now known as Eidolos 00:59 hkb_aw is now known as hakobe 01:07 flokuehn_ joined 01:11 devogon left 01:19 flokuehn left 01:20 jisom_ joined 01:38 jisom left 01:44 pro21ms4 joined 01:59 flokuehn joined 02:14 flokuehn_ left 02:22 Limbic_Region left 02:40 bonesss left 03:28 araujo joined 03:51 pro21ms4 left 03:59 demerphq joined 04:17 demq left 04:39 flokuehn left 04:43 iblechbot joined 06:01 Aankhen`` joined
Aankhen`` Suggestion for S06: "not because we're mean" would read better as "not *only* because we're mean". :-D 06:09
dduncan further suggestion for S06: "because we're mean" would read better as "because we're having fun" 06:22
for bonus points, you could further append ", and you're not"
Aankhen`` snickers. 06:27
06:34 meiss|na left 06:36 devogon joined 06:40 zard1989 joined 06:47 elmex joined 06:48 iblechbot left, pmurias joined
pmurias good morning 06:59
07:13 wilx left 07:17 weinig left 07:19 hakobe is now known as hkb_aw 07:21 wilx joined 07:24 zard1989 left 07:34 weinig joined 07:35 Aankhen`` left 07:37 dduncan left 07:41 Aankhen`` joined 07:45 xinming joined 07:46 franck___ joined 08:00 melissa left
agentzh morning :) 08:03
08:10 agentzh left, Aankhen`` left 08:12 agentzh joined 08:14 Aankhen`` joined 08:16 Aankhen`` left
moritz_ avar: I guess it's not morning for you...? *g* 08:17
08:17 Aankhen`` joined
moritz_ s/avar/agentzh/ # sorry 08:17
08:18 rindolf left 08:22 Aankh|Clone joined 08:32 BinGOs joined
bloonix good morning 08:32
lambdabot bloonix: You have 2 new messages. '/msg lambdabot @messages' to read them.
agentzh moritz_: it's 16:33 here :) 08:33
moritz_: i'm sitting in my $office :) $job will start tomorrow :) 08:34
08:35 daxim_ joined
moritz_ agentzh: what kind of $job? 08:38
moritz_ managed to convince his $boss that some perl coding is absolutely necessary (I'm converting weird file formats ;) 08:39
agentzh moritz_: a $job in yahoo china. 08:40
moritz_: a lot of my team members are perl guys :)
moritz_ agentzh: cool ;) 08:41
agentzh moritz_: i've convinced my $manager to contribute to pugs and perl 6 ;)
moritz_: he is now reading the synopses :) 08:42
moritz_ agentzh: that's very cool ;)
agentzh yes, it is!
08:42 Aankh|Clone left
agentzh we'll try to use more p5/p6 in our products. 08:43
08:45 Aankhen`` left
moritz_ agentzh: on a side note I just discovered that the charset in the irclog mysql table is "latin1" :( 08:45
agentzh: I changed it in the dump manually to utf8 and imported it locally, and it still works 08:46
08:46 Aankh|Clone joined
moritz_ agentzh: if I can modify it on the server without breaking it, maybe your fancy detection works as well... 08:46
08:47 weinig left
agentzh moritz_: nice 08:48
pmurias latin1--
agentzh moritz_: heh, how about do the decoding *before* importing logs to the db? 08:49
*doing
08:49 daxim left
agentzh not sure if your db can save raw bytes. 08:50
bloonix its 10:58 here and Im sitting on my balcony, drinking coffee and enjoin the sun :) 08:57
moritz_ agentzh: I don't want to do coding magic before importing the data because it could mean data loss 08:58
agentzh: and mysql's handling of blobs... I don't know... ;) 08:59
agentzh moritz_: *nod* 09:00
bloonix: wow
agentzh goes to reach his cup.
moritz_: still can't search utf-8 in irclogs...it seems :) 09:02
moritz_: i wonder it'll be a bit difficult to implement that if db saves raw data ;)
s/implement/implement fully/ 09:03
moritz_ agentzh: actually I just don't know :(
agentzh it'll require some kind of reverse-magic-encoding-detection algorithm...wow
using utf8 in db means potential data loss but it makes our lives easier ;) 09:05
how about two versions of db? (it'll require more room, of course) 09:06
or just save the raw data as local files... 09:07
like colabti.de...
moritz_: usually we don't want to search the raw bytes anyway... 09:09
demerphq is confused 09:13
if you store your data encoded then you can search for encoded strings. 09:14
09:14 demerphq is now known as dmq 09:22 Aankhen`` joined 09:24 feb joined 09:34 dmq left
Juerd agentzh: Data loss? 09:34
I thought unicode was a superset of everything else?
moritz_ Juerd: but if your decoding can't handle the data it was given, it is likely to produce garbage 09:35
09:35 Aankh|Clone left
Juerd That's not data loss, it's data-uselessness ;) 09:36
moritz_ no, if it can handle the charset, it really can lose data
s/can/can't/
09:38 dmq joined
agentzh dmq, Juerd: i'm trying to persuade moritz_++ to use UTF-8 for the logs stored in the db ;) 09:40
moritz_: i think you can save the raw bytes some where outside the db (to serve as a backup copy) 09:41
moritz_ agentzh: I think mysql has problems with indexing chinese and japanese characters anyway :( 09:42
agentzh moritz_: index the nick names is sufficient, no?
*indexing
dmq mysql has problems. 09:43
its like a teenager.
moritz_ agentzh: no, not for the fulltext search
agentzh: a fulltext search in ~1G lines without index kills my server
dev.mysql.com/doc/refman/5.0/en/ful...tions.html 09:44
lambdabot Title: MySQL AB :: MySQL 5.0 Reference Manual :: 11.8.4 Full-Text Restrictions
Juerd moritz_: One single encoding that may contain miscodings is better than mixed encodings, because then practically *everything* is miscoded.
moritz_ 5D#
5DIdeographic languages such as Chinese and Japanese do not have word delimiters. Therefore, the FULLTEXT parser cannot determine where words begin and end in these and other such languages. The implications of this and some workarounds for the problem are described in Section 11.8, ā€œFull-Text Search Functionsā€.
#
09:45 crashmatrix left
moritz_ Juerd: yes, maybe. though in practice most text is just ASCII `subset of` Unicode(as utf-8) 09:45
Juerd IRC programs must either be completely ignorant about charsets, or assume things. The former is worse, so assuming things is okay.
dmq encodings that are in a healthy relationship live longer.
Juerd Assuming that sequences aren't UTF-8 are iso-88591-1 is a sane choice
dmq single encodings tend to encounter health problems later in life. 09:46
09:46 crashmatrix joined
Juerd If most of the line is high bit, you can look for alternatives, but here that's not used much 09:46
09:46 Aankh|Clone joined
moritz_ I'll see if I can get a script running that does in place conversion of the database 09:47
dmq apologises for being in a silly mood today
moritz_ and then I'll make ilbot charset aware
Juerd dmq: The silly mood apologises for being around you ;) 09:48
dmq ssh, dont antagonize it, you never know what might happen
an angry silly mood is a firghtening thing 09:49
09:53 Aankh|Clone left 10:03 drupek12 left 10:04 Aankhen`` left, jisom_ left 10:13 drupek12 joined
feb somebody wanted to do searches in irc logs with utf-8 encoding ? 10:18
moritz_ yes
agentzh wants that, and I provide and maintain the logs
10:18 irclog_utf8 joined, irclog_utf8 left
feb colabti.de/irclogger//irclogger_log...xt=checked 10:18
lambdabot colabti.de/irclogger//irclogger_log...xt=checked 10:19
moritz_ maybe an example with matches would be more impressive ;)
feb: and I'm talking about irc.pugscode.org/
feb the url didnt paste well :(
lambdabot Title: IRC log for #perl6, 2007-08-05 10:20
moritz_ feb: which means there are non-printable chars in the URL - not good :(
feb tinyurl.com/3bw2xm 10:21
lambdabot Title: #perl6 irc log search
feb the copy paste is the culprit, not the url 10:22
moritz_ I always thought colabti did just latin1 logs (at least audreyt told me so once)
feb try the tinyurl
it did :)
moritz_ feb: I did, impressing ;)
10:23 irclog_utf8 joined, irclog_utf8 left
feb actually, i'm the operator of colabti 10:23
just changed it to utf-8, but it's hard-handed as you never know which encoding was used over irc
pugs_svnbot r17213 | Aankhen++ | * [irclog] couple more abbrs.
diff: dev.pugscode.org/changeset/17213
lambdabot Title: Changeset 17213 - Pugs - Trac
10:24 irclog_utf8 joined, irclog_utf8 left
feb btw., your logger has a nice display 10:24
10:24 irclog_utf8 joined
moritz_ feb: I know, that's what I'm struggling with (unknown encoding) 10:24
feb: it's very much customized for #perl6, and has seen much tweeking by Aankhen``++ ;) 10:25
# irc log test: é»„ęœ›ē§‹
feb yes, i saw you hardcoded the nick colors 10:26
moritz_ feb: only the ones of the most active nicks
feb: there are still heuristics for a few more colors
10:26 irclog_utf8 left
feb didt read that far 10:26
moritz_ wait... 10:27
the colors are hardcoded in the csss
s/sss/ss/
just the mapping nick -> css class is not fully hardcoded
10:28 irclog_utf8 joined
moritz_ # another irc log test: é»„ęœ›ē§‹ 10:28
feb: btw do you delete spam from old logs? 10:29
feb no, but i didnt see much spam too
and bots like google are not allowed to crawl the logs 10:30
moritz_ we had a bit of a hard time with spammers about two months ago
feb looks in the logs
anculz ? 10:31
moritz_ url? 10:32
feb: never mind... I just wanted to point out that _if_ you delete spam, the line anchors will change 10:33
feb was searching for the spam.
moritz_ feb: that's why I changed to database ID based line anchors 10:34
feb I once planned to add a way to hide/censor/delete unwanted lines
didnt think about the line anchors change before you discover it 10:35
moritz_ my logs currently have both the line numbers and db ids as anchors, and new links are always to the ids
10:36 irclog_utf8 left, irclog_utf8 joined, iblechbot joined
feb i hoped to allow changing the timezone, but it's annoying with urls based on dates 10:37
moritz_ feb: funny how we encounter the same problems ;)
feb it would break line numbers and the date
agentzh fyi, é»„ęœ›ē§‹ is a muderer in China ;)
feb lol
moritz_ agentzh: thanks ;)
agentzh you may want to use my Chinese name for testing: ē« äŗ¦ę˜„ 10:38
;)
moritz_ feb: you could change the URLs to date=2007-08-04;tz=+1 # for MEZ
and tz defaults to 0 making them GMT times 10:39
feb agentzh: are you a murderer too ? 10:40
10:40 irclog_utf8 left, irclog_utf8 joined
moritz_ ?eval 'ē« äŗ¦ę˜„' eq 'é»„ęœ›ē§‹' 10:41
feb moritz_: but it forces a timezone for the person who receives the url
pugsbot_r17185 Bool::False
feb or you would have to redirect according to some preferences
moritz_ feb: of course. Otherwise URLs point to different things to differnt persons, which would be very bad
the problem is that redirects can't contain anchors, because they are not part of HTTP 10:42
feb true 10:43
moritz_ otherwise I'd make the line anchor links point to id=$number, and redirect to the day which, according to the tz cookie, contains that id
there are workarounds, of course (like inserting a link, and perhaps some javascript magic) but that's all evil hacks 10:44
feb yes 10:45
don't you have trouble with bots harvesting all links in the logs 10:46
when you have a link for each line
moritz_ personally I don't mind, because my timezone is just GMT+1/+2 (depending on daylight saving time)
good point
they could be made nofollow or something 10:47
feb it's one of the reasons, why i disallowed them
as they downloaded hundreds of times the same page
moritz_ currently the line links are just #id-Links, I haven't seen any problems with that
feb got gigabytes of traffic because of them
pure #id-links should work 10:48
moritz_ I monitor my logs closely, and if they start exploding I'll edit robots.txt 10:49
feb i did when i saw the surge in traffic :) 10:50
moritz_ which bots?
feb many different crawlers, like googlebot and others from msn, aol 10:51
and many more i didnt know 10:52
moritz_ I'll see how it works out
10:58 irclog_utf8 left 11:11 mj41 left 11:27 wilx left, buetow joined 11:36 JarJarBi1ks joined 11:46 pro21ms4 joined
pasteling "bloonix" at 87.78.217.174 pasted "system("cat /proc/sys/fs/file-" (12 lines, 184B) at sial.org/pbot/26670 11:50
bloonix I am totaly confused :)
my example runs on /proc/uptime and other virtual files 11:51
11:53 chris2 joined 11:57 pro21ms4 left
pugs_svnbot r17214 | moritz++ | [irclog] ilbot.pl now logs UTF-8 in order to allow utf8 fulltext search 12:06
diff: dev.pugscode.org/changeset/17214
lambdabot Title: Changeset 17214 - Pugs - Trac
12:06 irclogbot joined
moritz_ the database is now recoded in place 12:07
bloonix my $all = =$fh; the same thing 12:10
just the first number
12:12 rintaro joined 12:16 zsh joined
pugs_svnbot r17215 | moritz++ | [irclog] search no doesn't include '#' in the channel name 12:24
r17215 | moritz++ | (more robust this way)
diff: dev.pugscode.org/changeset/17215
lambdabot Title: Changeset 17215 - Pugs - Trac
moritz_ wonders why the fulltext search works locally with chinese characters, but not on the server 12:25
12:34 zsh left 12:43 cognominal__ left 12:50 cognominal_ joined 12:52 turrepurre left
moritz_ # another irclog test: ē« äŗ¦ę˜„ 13:01
13:03 irclog_utf8 joined
moritz_ charsets are weird beasts, and mysql multiplies my confusion 13:04
# another irclog test: ē« äŗ¦ę˜„ (No 2) 13:05
13:05 irclog_utf8 left 13:06 irclog_utf8 joined, irclog_utf8 left
bloonix ?eval 'foo' ~~ /(foo)/; my %h; %h{foo} = $0; say %h{foo}; 13:14
pugsbot_r17185 Error: No such subroutine: "&foo"
bloonix ?eval 'foo' ~~ /(foo)/; my %h; %h{foo} = $0; say %h<foo>;
pugsbot_r17185 Error: No such subroutine: "&foo"
bloonix ?eval 'foo' ~~ /(foo)/; my %h; %h<foo> = $0; say %h<foo>;
pugsbot_r17185 OUTPUT[fooā¤] Bool::True
13:28 hkb_aw is now known as hakobe 13:34 chris2 left
Blwood does the < > operator work ? (the qw// like) 13:40
moritz_ Blwood: yes 13:43
lumi ?eval <foo bar baz>.perl
pugsbot_r17185 "(\"foo\", \"bar\", \"baz\")"
moritz_ ?eval <a bc>
pugsbot_r17185 ("a", "bc")
lumi Blwood: Works as advertised
moritz_ at least in pugs ;) 13:45
13:50 Aankhen`` joined
Blwood okay thank you :) 13:57
?eval <a b c>.end 14:00
pugsbot_r17185 2 14:01
Blwood oh right so .end gives the last index ? I tought it gave the last element 14:02
14:03 wilx joined
moritz_ ?eval <a b c>.last 14:03
pugsbot_r17185 Error: Cannot use this control structure outside a 'loop' structure
moritz_ ?eval <a b c>[-1]
pugsbot_r17185 \"c"
moritz_ I think in the new syntax that's [*-1], but I'm not quite sure about that 14:04
Blwood I read on the dev doc "$#foo is gone. If you want the final subscript of an array, and [-1] isn't good enough, use @foo.end instead."
Oh I understand now :) 14:05
the best doc for perl6 is dev.perl.org/perl6/doc/apocalypse.html ?
lambdabot Title: Perl6 Apocalypses - perl6
moritz_ no, you should read the synopsis 14:06
perlcabal.org/syn/
lambdabot Title: Official Perl 6 Documentation
Blwood woaw :) thank you very much
moritz_ you're welcome 14:07
14:11 cognominal_ left
feb moritz_: you dont log nick changes ? 14:17
14:18 jettero left
feb oh, you do 14:18
14:18 jettero joined 14:21 jettero left 14:22 kane__ joined 14:25 jettero joined 14:30 cognominal_ joined 14:33 cognominal__ joined 14:34 cognominal_ left
feb moritz_: is it ok if I borrow some parts of your CSS ? 14:38
14:39 DarkWolf84 joined, jiing joined 14:46 buetow left, buetow joined 15:02 jettero left, jettero joined 15:03 stevan_ joined 15:05 cognominal_ joined
moritz_ feb: I do, normally (looking...) 15:06
feb: of course, it's (per default) under the artistic license 2 or something
feb something :)
15:06 cognominal__ left
Blwood can we use Socket module in perl6 ? 15:06
15:07 flokuehn joined
moritz_ Blwood: you can try to use perl5:Socket; 15:07
Blwood I see on : perlcabal.org/syn/S16.html that we can use some functions available on Socket; like connect ... 15:08
moritz_ moritz.faui2k3.org/irclog/out.pl?ch...05#i_75012 <-- nick change
lambdabot Title: S16
Title: IRC log for #perl6, 2007-08-05, tinyurl.com/28xea4
Blwood but Socket for perl6 is not adapted yet ? 15:09
moritz_ Blwood: then you can in principle. I just don't know if it's implemented yet...
Blwood: but you could look into the pugs repository in ext/, there are networking examples
Blwood okay I'll have a look, thank you :)
moritz_ like Net::IRC
feb moritz_: yes, you use another working in the logs for nick change
pmurias use Socket:from<perl5> is the kp6 way
moritz_ and there's a HTTP-Server-Simple as well 15:10
feb: it was the first "quick and dirty" hack I could think of ;)
feb s/working/wording 15:11
15:15 flokuehn left 15:18 [particle1 is now known as [particle] 15:21 Blwood left 15:30 Blwood joined 15:35 elmex left 15:40 kanru joined, jettero left, jettero joined 15:43 jettero left 15:47 jettero joined, jettero left 15:54 penk joined 16:15 cognominal_ left 16:16 cognominal_ joined 16:17 jettero joined 16:22 kanru left 16:33 jettero left, jettero joined 16:42 penk left 16:49 Azure-BOT joined 16:50 Khisanth joined, jisom joined 17:00 iblechbot left 17:21 chris2 joined 17:29 Aankh|Clone joined 17:35 buetow left 17:36 toly joined 17:38 toly left 17:41 stevan_ left 17:42 marmic left, fglock joined 17:43 Psyche^ joined 17:44 marmic joined 17:45 Aankhen`` left 17:47 toly joined 17:55 cognominal__ joined 18:01 Patterner left, Psyche^ is now known as Patterner
moritz_ is it right that '&' is the sigil for code blocksƟ 18:01
s/Ɵ/?/
18:05 cognominal_ left 18:09 rintaro_ joined
Juerd moritz_: The & is the sigil for closures, iirc 18:24
spinclad S02:1110: & code/rule/token/regex 18:25
so yes
18:25 rintaro left
spinclad see also S02:1158 18:29
18:38 nperez joined 18:42 cognominal__ left 18:44 Mmmbop joined
moritz_ thanks Juerd, spinclad ;) 18:46
18:48 Mmmbop left 18:50 weinig joined, Aankh|Clone left 18:52 Mmmbop joined 18:53 jettero left 18:54 jettero joined 18:58 cognominal_ joined 19:12 Mmmbop left 19:17 franck___ left
pmurias fglock: hi 19:35
fglock pmurias: hi!
I'm trying to implement matching Captures to Signatures, but I keep getting distracted :P 19:36
19:36 maskd joined
pmurias :) wanted to do the same thing this morning but got distracted to :) 19:37
i think $var->{_dispatch}($var,...) needs to be replaced in emitted code by something which is not as evil 19:39
there is a potential problem with f()->{_dispatch}(f(),'method',...) if f() has side effects 19:40
the DispatchSugar has it's own problem with requiring sugar() calls all over the place 19:42
19:42 araujo left
pmurias using dispatch($invocant,'name',...) is my proposed solution 19:43
or using autobox.pm
fglock hmm - looks good
i mean - dispatch($invocant,'name',...)
it doesn't require a temporary variable 19:44
and it is portable to other backends 19:45
pmurias where should the function go? 19:48
MOP? 19:49
19:49 araujo joined
fglock either MOP.pm, or in GLOBAL (lib/KindaPerl6/Runtime/Perl5/Runtime.pm) 19:51
i think MOP.pm is better
i've got to find something to eat - back in ~1h 19:53
pmurias should it be invoked as a fully qualified sub? not sure where to export it from 19:54
pugs_svnbot r17216 | moritz++ | [irclog] clear non-printable characters from the logs before any 20:08
r17216 | moritz++ | other output processing to prevent bogus \b matches
diff: dev.pugscode.org/changeset/17216
lambdabot Title: Changeset 17216 - Pugs - Trac
20:10 mj41 joined 20:14 fglock left 20:18 nperez left 20:26 rintaro joined 20:28 apeiron joined 20:32 Azure-BOT left 20:33 sunnavy joined 20:37 cognominal_ left
pugs_svnbot r17217 | lwall++ | statement semicolon not eaten correctly as noted by pmichaud++ 20:37
diff: dev.pugscode.org/changeset/17217
lambdabot Title: Changeset 17217 - Pugs - Trac
20:37 cognominal_ joined 20:41 rintaro_ left 20:46 Azure-BOT joined
pugs_svnbot r17218 | pmurias++ | kp6: dispatch() is exported from KindaPerl6/Runtime/Perl5/Runtime.pm 20:55
diff: dev.pugscode.org/changeset/17218
lambdabot Title: Changeset 17218 - Pugs - Trac
20:57 rintaro_ joined
Blwood say() function adds \r\n ? 20:58
I thought it was just \n
szbalint I'd guess it depends on your operating system
Blwood I see some examples in pugs/examples/networking were irc requests are made using say 20:59
szbalint and line ending settings
Blwood and irc request need \r\n
irc requests*
szbalint, ah it would be very nice :)
but I still can't understand the use of say in a http/irc request :s
21:01 cognominal_ left, kane__ left
pmurias sleep& 21:01
21:01 pmurias left 21:02 kane__ joined 21:04 kane__ left
szbalint I don't see \r\n anywhere in pugs/examples/network, where are you seeing such endings Blwood? 21:06
21:12 rintaro__ joined, rintaro left 21:13 cognominal_ joined 21:21 rintaro joined 21:23 rintar___ joined 21:26 rinta____ joined 21:28 rintaro_ left
Gothmog_ szbalint: He just said that lines in an HTTP or IRC request /should/ end with \r\n, which is correct (or at least, if \r\n means CRLF). 21:33
21:33 diotalevi_ left
Blwood szbalint, yeah I don't see \r\n there's just for example : say("GET / HTTP/1.1\n"); normally it should be GET HTTP/1.1\r\n 21:36
but the request works, though :o 21:37
21:37 rintaro__ left
Blwood also, does someone know how to call a method inside the class ? 21:38
I mean for example a self.drink inside the Class dringer { }
drinker*
21:39 rintaro left, rintaro joined, fglock joined 21:43 rintar___ left 21:44 diotalevi_ joined 21:45 Daenyth` joined 21:47 rintaro_ joined 21:48 Blwood is now known as t4pz, t4pz is now known as Blwood 21:56 rinta____ left 21:59 rintaro left
szbalint ah 22:04
22:05 diotalevi left 22:07 buetow joined 22:12 Azure-BOT left 22:15 buetow left
Gothmog_ Blwood: The request probably works because most servers are very tolerant. 22:16
diakopter moritz: pong
moritz: oh yeah, gotta commit that evalbot change. doing so now. 22:20
Blwood Gothmog_, yeah maybe, but for irc it is not so tolerant I think.
Gothmog_ I'm not sure. Fire up telnet and try. :-) 22:21
Blwood yeah good idea
Gothmog_ Or replace say "..."; with print "...\n";
Blwood yes I use print, but I was a bit surprised for the say 22:22
Gothmog_ But I suppose it works. Nevertheless I would always send correct data.
22:24 diotalevi_ left 22:25 cognominal_ left, cognominal__ joined
pugs_svnbot r17219 | diakopter++ | moritz: here are evalbot's temp workarounds 22:41
diff: dev.pugscode.org/changeset/17219
lambdabot Title: Changeset 17219 - Pugs - Trac
22:41 ozo left 22:45 chris2 left 22:46 Limbic_Region joined 22:58 ozo joined 23:17 [particle1 joined 23:32 bonesss joined 23:33 [particle] left, Blwood left 23:38 ozo left 23:45 ezraburgh joined 23:51 fridim joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!