|
Parrot 4.2.0 "Ornithopter" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC Set by moderator on 21 March 2012. |
|||
| whiteknight | these timing numbers don't make any sense to me | 00:14 | |
| at one iteration, the times are the same. At 10-1000 iterations the old algorithm is 3-5 times faster (not growing in a monotonic way) Over 10000 the new algorithm is a hundred times faster or more | 00:17 | ||
| actually, that last number is a fluke. It's only about 8 times faster | 00:19 | ||
|
00:29
particle joined
00:31
kurahaupo joined
|
|||
| dukeleto | bugs bugs bugs bugs BUGS! | 00:32 | |
| Coke | /away | 00:36 | |
| dalek | rrot: 486331d | Whiteknight++ | src/ (2 files): Implement StringBuilder.push_integer vtable, so we can assemble strings by pushing raw codepoints instead of needing to first create a 1-character string (chr_s_i) and then pushing that string to the StringBuilder |
00:44 | |
|
00:44
plobsing joined
|
|||
| bacek_at_work | msg whiteknight your last commit is really really wrong. Converting string in-place is braking our promise about immutable strings. And underlying buffer can be shared between various strings. | 00:51 | |
| aloha | OK. I'll deliver the message. | ||
| benabik | bacek_at_work: Isn't StringBuilder specifically designed for mutable strings? | 00:52 | |
| bacek_at_work | benabik, yes it is. | 00:53 | |
| Ah. We are cloning string in get_string VTABLE. | 00:54 | ||
| msg whiteknight Ignore it. benabik++ pointed out my mistake. | |||
| aloha | OK. I'll deliver the message. | ||
|
01:07
travis-ci joined
|
|||
| travis-ci | [travis-ci] parrot/parrot#242 (master - 486331d : Whiteknight): The build was fixed. | 01:07 | |
| [travis-ci] Change view : github.com/parrot/parrot/compare/2......486331d | |||
| [travis-ci] Build details : travis-ci.org/parrot/parrot/builds/1053488 | |||
|
01:07
travis-ci left
|
|||
| dalek | rrot: 17a886a | jkeenan++ | src/pmc/stringbuilder.pmc: [codingstd] Make src/pmc/stringbuilder.pmc pass c_arg_assert.t. But it's failing pmc_docs.t because no one added documentation for a new function! |
02:46 | |
|
03:00
davidfetter joined
04:30
brambles joined
05:09
birdwindupbird joined
05:16
alvis` joined
05:17
alvis` left
05:24
alvis joined
05:34
preflex_ joined
07:36
lucian joined
07:42
contingencyplan joined
08:26
lucian joined
09:58
brrt joined
10:01
contingencyplan joined
11:05
brrt left
11:46
bluescreen joined
11:57
benabik joined
12:05
whiteknight joined
12:08
brrt joined
|
|||
| tadzik | benabik++++ | 12:30 | |
| (proposals) | |||
| masak | url? | 12:41 | |
| tadzik | whiteknight.github.com/2012/04/09/r...osals.html | ||
|
12:44
ilbot2 joined
|
|||
| moderator | Parrot 4.2.0 "Ornithopter" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC | ||
|
12:45
bluescreen joined
12:54
JimmyZ joined
13:06
whiteknight joined
|
|||
| whiteknight | good morning, #parrot | 13:06 | |
| masak | morgningz, whiteknight | 13:07 | |
| whiteknight | hello masak | ||
| tadzik | 'ello whiteknight | 13:08 | |
| whiteknight | hello tadzik | ||
| Here's an extremely interesting (a little funny, a little sad) article about PHP: me.veekun.com/blog/2012/04/09/php-a...ad-design/ | |||
| tadzik | :) | ||
| whiteknight | I've never seen another rant about PHP that was so complete | 13:09 | |
| it is funny just how many of those problems he mention have to do with functions and their naming/arguments/behaviors/bugs | 13:13 | ||
| It really seems to me like a language with PHP syntax and a sane standard lib would be a major improvement | 13:14 | ||
| Or PHP-similar syntax. We can fix stupid things like array() and empty() not being functions, or fix references, or make functions into first-class objects | 13:15 | ||
| or make namespace syntax non-terrible | |||
| Of course if you make all those changes you start getting towards a language which is almost Perl, minus most sigils, default variables and magical variables | 13:17 | ||
| ...which is not a terrible thing to converge upon, methinks | |||
| JimmyZ | that looks like "The UNIX- HATERS Handbook" | ||
| brrt | or python, if you converge the other way | ||
| whiteknight | JimmyZ: I've never seen that one | ||
| brrt: Yeah, PHP really did borrow so much syntax from other places | |||
| Perl was the really big influence, early on. C and Java too in varying amounts over time | 13:18 | ||
| brrt | all languages, a bit | 13:19 | |
| i kind of agree with the author | |||
| i just try to not think about it :-) | |||
|
13:19
hercynium joined
|
|||
| whiteknight | The problem is trying to separate out the different entities. The PHP interpreter, the apache integration, the standard library, the language syntax, etc | 13:20 | |
| I think there are definitely good things in there, and obviously new users pick up the language in droves for a variety of reasons | |||
| JimmyZ | what I don't like is that PHP is more like java now. | 13:21 | |
| whiteknight | JimmyZ: I think what happens is that people see the impressive performance gains that Java has had over the years, and think that they can steal some of those ideas without understanding how they fit together | 13:22 | |
| brrt | php is really easy to set up | ||
| there, you have it | |||
| whiteknight | brrt: yes, that's a big deal | ||
| brrt | and really really easy to connect to mysql | ||
| which explains the popularity of that db, too :-) | |||
| whiteknight | brrt: I would love it one day if a mod_parrot was just as easy to set up. A few apt-get calls, set up some paths, and upload a few scripts | ||
| brrt too | |||
| especially as i am in fact pretty lazy as far as setting up servers goes | 13:23 | ||
| whiteknight | a default config file with some get-started-fast settings already prepared for a debian distro (for example) would go a long way | ||
| whitelist /var/www/parrot/ by default, dump some scripts there, get started fast | 13:24 | ||
| brrt | yeah, we should do that | ||
| JimmyZ hopes someone will reveiw his M0 code | 13:25 | ||
| benabik | o/ #parrot | 13:27 | |
| whiteknight | hello benabik | 13:29 | |
| brrt | hello | ||
| whiteknight | brrt: I had very high hopes for a rewritten PHP6, but it seems like that is stalled | ||
| Coke | whiteknight: on parrot, or standalone? | 13:30 | |
| brrt | yes, i believe so too | ||
| people printed books about php6 | |||
| benabik | "There is no threading support whatsoever." That's not really in PHP's problem domain. Trying to do threading on top of your web server's threading is probably not a good idea. | 13:32 | |
| brrt | benabik: good point actually, i was wondering about w/ regards to mod_parrot] | ||
| although i do not entirely agree | 13:33 | ||
| nine | My webserver doesn't do threading. It uses processes for dynamic content. Being able to use quick threads for speeding up single requests would be awesome. | ||
| brrt | threading would be very useful sometimes | ||
| as in, 'lets start a thread to resize this image' | |||
| benabik | brrt: Better: send an IMG tag to the user and let that request resize it. | 13:34 | |
| brrt | benabik: no, wrong :-p | ||
| either you use caching of resized files | 13:35 | ||
| and a malicious user can blow up your server by requesting a billion different formats | |||
| or, you do not use caching | |||
|
13:35
bluescreen joined
|
|||
| brrt | and a malicious user can blow up your server by requesting only a million different formats simultaneously | 13:35 | |
| benabik | If you expose completely free-form image converting, yes. | 13:36 | |
| whiteknight | Coke: I've never heard any of the PHP core devs talk about using Parrot as an option | ||
| Coke: Obviously I would love to see a PHP 6 on Parrot | |||
| benabik | brrt: But really, a thread inside your web server is not the place to do that IMHO, esp if the resize might take longer than the original request. Use backend workers or something. | 13:37 | |
| brrt | benabik: that is really a pros-and-cons thing | 13:38 | |
| Coke | whiteknight: it was a parrot pipe dream at some point. | ||
| much like my coldfusion version. | |||
| brrt | especially since php doesn't do fork, just exec() | ||
| Coke | but without mod_parrot and easy DBI, I think both PHP and CF suffer from the same problem. | 13:39 | |
| brrt | i kind of agree that having a billion threads w/o clear rules on when they should be killed is a nightmare for a webserver | 13:40 | |
|
13:47
jashwanth joined
13:59
PacoAir joined
|
|||
| dalek | kudo/nom: b8fdbea | moritz++ | src/core/Temporal.pm: Date.new should not die with LTA error |
14:06 | |
| PerlJam | whiteknight: The link to the GSOC website at the end of your post is broken btw. It goes to whiteknight.github.com/2012/04/09/w...elange.com | 14:26 | |
| whiteknight | blah | 14:27 | |
| detection of relative urls is poor | |||
|
14:30
bluescreen joined
14:50
dmalcolm joined
14:54
brrt left
14:58
brrt joined
|
|||
| dukeleto | ~~ | 15:11 | |
| this looks interesting: github.com/GaloisInc/HaLVM | 15:12 | ||
| msg whiteknight: this might be useful for PLA: github.com/fredrik-johansson/flint2 | 15:13 | ||
| aloha | OK. I'll deliver the message. | ||
| whiteknight | I'll take a look in a bit | ||
| dukeleto | whiteknight: it is number theory though | 15:14 | |
| whiteknight | OH NOES SHIELD ME FROM IT | ||
| dukeleto | whiteknight: it is definitely useful for matrixy | ||
| whiteknight: my first loan fell through and now I am scrambling to make another happen. Fun! | |||
| dukeleto preps for $work | 15:15 | ||
| whiteknight | dukeleto: Damn! | ||
| benabik | Github has enabled syntax highlighting for the DCPU-16 from 0x10c | 15:16 | |
| atrodo | github.com/blog/1098-take-over-the...ith-github This has taken of surprisingly quickly | ||
| benabik> Haha | |||
| benabik | atrodo: I guess you also follow their twitter, hm? | ||
| atrodo | benabik> You would be correct | 15:17 | |
| dalek | kudo/nom: 32c674e | moritz++ | src/core/Backtrace.pm: Backtrace.nice can now produce single lines |
15:24 | |
| kudo/nom: edfd0f2 | moritz++ | src/Perl6/World.pm: [World] avoid a lookup for which we already have a variable |
16:12 | ||
| kudo/nom: 8144a04 | moritz++ | src/ (3 files): implement Routine.package |
|||
|
16:14
plobsing joined
|
|||
| dalek | kudo/nom: 72e8fe8 | moritz++ | src/Perl6/World.pm: [World] further cleanup by spider-mario++ |
16:21 | |
|
17:05
PacoAir joined
17:11
lucian joined
17:58
brrt joined
18:19
lucian joined
|
|||
| dukeleto | ~~ | 18:52 | |
| whiteknight | hello dukeleot | ||
| or dukeleto | |||
| dukeleto | whiteknight: howdy | ||
| whiteknight | sucks about your loan. the bank 'tis a fickle mistress | 18:56 | |
| if you need a burning paper bag of human waste left on somebody's doorstepp, prvmsg me | |||
| I suspect USPS will hassle me about the shipping, however | 18:57 | ||
|
18:57
Hunger joined
19:13
plobsing joined
19:45
brrt left
|
|||
| dukeleto | whiteknight: :) | 20:22 | |
| dalek | rrot: 63c15b7 | NotFound++ | src/platform/generic/dl.c: POD for static helper functions |
20:34 | |
| rrot: ee4c8c9 | NotFound++ | src/pmc/stringbuilder.pmc: add function POD and argument decorators to calculate_capacity |
|||
| kudo/nom: 4bacf1a | moritz++ | src/core/Exception.pm: add line numbers for warnings does not yet do the right things for warnings thrown from inside the setting, but warn() calls should be fine |
21:09 | ||
| kudo/nom: fa2d53c | moritz++ | / (2 files): first shot at bringing Str.indent back |
21:25 | ||
|
22:13
plobsing joined
22:19
alester joined
22:21
alester joined
|
|||
| alester | Is there a list of GSOC proposals somehwere? | 22:21 | |
| I've seen them whiz by and didn't think anything of it | |||
| dukeleto | alester: i think you need to log into the melange website to view them | 22:22 | |
| alester | but then I thought "Oh wait, maybe there's cage stuff that people want to do and I could help" | ||
| benabik | alester: whiteknight.github.com/2012/04/09/r...osals.html | ||
| alester | I don't know what the melange website is. | ||
| benabik | www.google-melange.com | ||
| You'd need to sign up to be a mentor to view the full proposals. Although that doesn't commit you to anything. :-) | |||
| alester | ok, so everything there is outside of my ballpark. | 22:25 | |
| Because I don't understand Parrot in the slightest. :-) | |||
| Well, I do. I just don't know enough to do anything useful and forward-thinking. | 22:27 | ||
|
22:55
whiteknight joined
|
|||
| dalek | rrot: 5f6ac7a | petdance++ | frontend/pbc_dump/packdump.c: Treat printed_flag_p like the boolean it is, rather than pretending it's a counter |
22:57 | |
| whiteknight | good afternoon, #parrot | 23:07 | |
| benabik | o/ whiteknight | ||
| aloha | (parrot/parrot) Issues opened : 755 (PackFile_pack_size returns the wrong type) by petdance : github.com/parrot/parrot/issues/755 | 23:08 | |
| whiteknight | hello benabik | 23:09 | |
| benabik | oops, gtg | 23:11 | |
| whiteknight | I don't know if anybody else has ever heard of "Mari0", the "mario with portals" game | 23:16 | |
| but I think all my free time for the next few days has officially dried up | |||
| alester | Is that part of the m0 branch? | ||
| whiteknight | no, it's a game. Not parrot related | ||
| alester | I know, I kid. | ||
| Besides, maybe if I joke about it I'll be kept from being tempted to go look and maybe download it and play it and OH MY GOD iT'S 3:15 IN THE MORNING WHY AM I STILL UP PLAYING THIS DAMN THING | 23:18 | ||
| (Not that that's happened in the past) | |||
| (week or two) | |||
| whiteknight | what we need is an easy game-creation framework for Parrot | 23:21 | |
| dukeleto | whiteknight: go on... | 23:29 | |
| whiteknight: the SDLx:: namespace on CPAN is pretty interesting | |||
| whiteknight | I don't have any concrete plans, but I | 23:33 | |
| 'm playng this game and I wish it were on parrot | |||
| dalek | rrot: f17bd2d | petdance++ | src/packfile/api.c: Remove unnecessary cast |
23:41 | |
| rrot: 81d3c08 | petdance++ | / (3 files): adding more function annotations |
|||
| nbrown | hello parrot | 23:51 | |
| whiteknight | HELLO NBROWN | 23:57 | |
| sorry for yelling | |||
| nbrown | no problem, it happens from time to time | ||
| whiteknight | my kid hit the capslock | 23:58 | |
| nbrown | ha ha | ||
| I was hoping to find some m0 folks to talk to, but it seems pretty dead in here right now | 23:59 | ||
| whiteknight | dukeleto was here | ||
| nbrown | looks like I just missed him | ||