Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2010-10-21

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:04 xabbu42 left #perl6
00:05 jedai_ left #perl6
00:06 jedai joined #perl6
00:10 f00li5h left #perl6
00:10 f00li5h joined #perl6
00:17 felipe left #perl6
00:18 ggoebel joined #perl6
00:24 ggoebel left #perl6
00:28 f00li5h left #perl6
00:33 rvrgs left #perl6
00:33 wamba left #perl6
00:40 stkowski left #perl6
00:40 QinGW joined #perl6
00:44 QinGW left #perl6
00:48 QinGW joined #perl6
00:57 masonkramer joined #perl6
01:01 ggoebel joined #perl6
01:05 f00li5h joined #perl6
01:06 f00li5h left #perl6
01:06 f00li5h joined #perl6
01:07 SpiceMan joined #perl6
01:09 SpiceMan I'm early adopting rakudo star. compiling right now, never really cared to read about perl 6. is there some perldoc perlsyn like thin?
01:09 SpiceMan *thing
01:09 SpiceMan (or perldoc, actually :)
01:21 jaldhar joined #perl6
01:25 Entonian left #perl6
01:38 meppl joined #perl6
01:40 masonkramer nothing terribly quick SpiceMan
01:42 masonkramer I know you're a perl 5 guy - synopses 2, 3, 12 were the most important ones for me to get started
01:42 masonkramer I'm sure the perl 6 authors contributors will have more to say
01:42 florz left #perl6
01:42 masonkramer there is also a book, but it's still under heavy development
01:43 masonkramer also this http://perlcabal.org/syn/Differences.html
01:45 masonkramer I've also learned a lot from moritz http://perlgeek.de/en/
01:47 zostay left #perl6
01:48 zostay joined #perl6
01:48 florz joined #perl6
01:56 Sherif joined #perl6
01:56 Sherif hello
01:59 Sherif left #perl6
02:15 ggoebel left #perl6
02:21 ggoebel joined #perl6
02:24 SpiceMan masonkramer: thanks, sorry for the late reply, stupid marketing called :p
02:26 f00li5h left #perl6
02:26 SpiceMan hey! use v5; is all I needed :p
02:26 f00li5h joined #perl6
02:29 nadim_ left #perl6
02:33 ggoebel left #perl6
02:38 agentzh joined #perl6
02:39 skids joined #perl6
02:40 masonkramer left #perl6
02:42 cdarroch left #perl6
02:45 f00li5h left #perl6
02:45 f00li5h joined #perl6
02:57 zostay left #perl6
02:58 zostay joined #perl6
03:10 _twitch joined #perl6
03:23 meppel joined #perl6
03:24 meppl left #perl6
03:29 _twitch left #perl6
03:45 meppel left #perl6
03:46 meppl joined #perl6
03:55 SpiceMan hmm, does compiling parrot take a lot of time? it's been in "parrot-nqp --target=pir --output=src/gen/perl6-grammar.pir" since for about 90 mins. (it's a 384MB RAM VM though)
03:55 SpiceMan s/since//
04:00 florz there was some talk about building taking forever recently
04:01 SpiceMan ah, ok. just wanted to know if I had to kill the thing and look for causes or let it be
04:02 florz maybe use one of the * releases?
04:02 florz (assuming you intend to use rakudo ...)
04:03 plobsing SpiceMan: are you using 2.9.0? we (parrot) had to make a 2.9.1 release because of slowness of rakudo compiles.
04:03 SpiceMan I cloned the git repo
04:03 SpiceMan of... rakudo. which I believe fetches parrot with svn
04:04 SpiceMan "parrot_install/include/2.9.1-devel"
04:06 plobsing the gc eats memory in 256M increments. compiling the setting (huge single unit compilation) might require more than 256M, which would exceed your VM's 384M memory and cause crazy swapping.
04:08 SpiceMan right, free mem: 5mb. well, I'm in no hurry
04:10 agentzh left #perl6
04:10 Chillance left #perl6
04:11 skids left #perl6
04:11 agentzh joined #perl6
04:18 Schwern joined #perl6
04:24 dalek 6model: 588770d | mberends++ | java/runtime/ (6 files):
04:24 dalek 6model: [java/runtime] add P6mapping and sync all recent dotnet changes
04:24 dalek 6model: review: http://github.com/jnthn/6model/commit/5​88770ddbedcfedf4c2b644d35111bacaa4cf926
04:39 cognominal left #perl6
04:40 kjeldahl joined #perl6
04:40 colomon joined #perl6
04:50 araujo left #perl6
04:56 colomon left #perl6
04:59 [particle] left #perl6
04:59 [particle] joined #perl6
05:18 pythonian4000afk is now known as pythonian4000
05:21 pythonian4000 is now known as pythonian4000afk
05:21 Schwern left #perl6
05:43 Schwern joined #perl6
05:57 _kaare joined #perl6
05:58 dual left #perl6
05:59 envi joined #perl6
06:05 dual joined #perl6
06:11 sorear SpiceMan: you really want 768 to build Rakudo now.  384 hasn't been supported since Apr-May
06:11 sorear it fluctuates a great deal - in Mar and Jun-Jul you needed 1.5-2 G
06:12 SpiceMan heh
06:12 * sorear is stuck on a 384 MB physical box.  One of a handful of reasons I'm not contributing to the Rakudo project
06:18 spinclad left #perl6
06:26 xabbu42 joined #perl6
06:29 dju_ left #perl6
06:43 justatheory left #perl6
06:54 [particle]1 joined #perl6
06:56 [particle] left #perl6
06:57 pnu_ left #perl6
07:34 muixirt joined #perl6
07:36 * muixirt wonders why so many devs have low memory boxes
07:39 muixirt sorear: what are the mem requirements building or running niecza?
07:40 sorear muixirt: I think you mean so few.  If Rakudo's memory usage was causing problems for the developers it would have been fixed by now
07:41 sorear ~256.  Which is not a problem for me, so improving it is not a priority.
07:42 aloha left #perl6
07:44 muixirt sorear: well I thought of parrot devs as well
07:45 muixirt ok, the mem problems building parrot and rakudo are somewhat related
07:45 sorear only kid51 has ever complained about memory problems among Parrot core devs
07:46 sorear and he has 256MB ram
07:46 muixirt and someone with 128MB complained on the mailing list
07:47 muixirt for example I don't understand the move from ops2c.pl to ops2c.nqp
07:48 muixirt this is cracking nuts with an even bigger slegehammer
07:51 Schwern left #perl6
07:52 sorear it needed a rewrite
07:52 sorear b/c Parrot is moving away from C
07:53 * muixirt snickers
07:53 jfried left #perl6
07:53 rgrau left #perl6
07:53 LionMadeOfLions left #perl6
07:53 IllvilJa left #perl6
07:53 sECuRE left #perl6
07:53 shabble left #perl6
07:53 \xF0 left #perl6
07:53 robinsmidsrod left #perl6
07:53 Maddingue left #perl6
07:53 nsh left #perl6
07:53 nperez left #perl6
07:53 frodwith left #perl6
07:53 xiaolongxia left #perl6
07:53 dual left #perl6
07:53 _kaare left #perl6
07:53 jjore left #perl6
07:53 huf_ left #perl6
07:53 krakan left #perl6
07:53 lue left #perl6
07:53 pochi_ left #perl6
07:53 broquaint left #perl6
07:53 gfldex left #perl6
07:53 peters_mops left #perl6
07:53 sirhc left #perl6
07:53 DarthGandalf left #perl6
07:54 moritz_ good morning
07:54 muixirt hi moritz_
07:54 sorear hi moritz_
07:55 dual joined #perl6
07:55 _kaare joined #perl6
07:55 jjore joined #perl6
07:55 huf_ joined #perl6
07:55 krakan joined #perl6
07:55 lue joined #perl6
07:55 pochi_ joined #perl6
07:55 broquaint joined #perl6
07:55 gfldex joined #perl6
07:55 peters_mops joined #perl6
07:55 sirhc joined #perl6
07:55 DarthGandalf joined #perl6
07:56 kjeldahl left #perl6
07:56 jfried joined #perl6
07:56 rgrau joined #perl6
07:56 IllvilJa joined #perl6
07:56 sECuRE joined #perl6
07:56 shabble joined #perl6
07:56 \xF0 joined #perl6
07:56 robinsmidsrod joined #perl6
07:56 Maddingue joined #perl6
07:56 nsh joined #perl6
07:56 nperez joined #perl6
07:56 frodwith joined #perl6
07:56 xiaolongxia joined #perl6
07:56 kjeldahl joined #perl6
08:00 LionMadeOfLions joined #perl6
08:02 pythonian4000afk is now known as pythonian4000
08:08 krunen joined #perl6
08:09 kjeldahl left #perl6
08:12 dalek niecza/master: 5c30728 | sorear++ | / (6 files):
08:12 dalek niecza/master: Implement <O> and regex args
08:12 dalek niecza/master: review: http://github.com/sorear/niecza/commit/5​c30728db98dd6963c3de5c71a7509940ccce19e
08:19 aloha joined #perl6
08:21 jnthn morning, #perl6
08:21 mce_ro joined #perl6
08:22 dju joined #perl6
08:23 moritz_ mornin' jnthn
08:23 dakkar joined #perl6
08:23 mce_ro left #perl6
08:23 * moritz_ just finished the cake for wive's birthday
08:24 mce_ro joined #perl6
08:24 jnthn moritz_: Nice :-)
08:25 jhuni joined #perl6
08:26 sorear Is wive a typo or a linguistic slip?
08:26 sorear morning, jhuni
08:26 dju left #perl6
08:26 sorear jnthn
08:27 dju joined #perl6
08:27 jnthn v and f are just a voicing apart :-)
08:27 muixirt moritz_ is a muslim and has wives :-)
08:28 sorear that would be wives' then
08:28 jferrero joined #perl6
08:28 muixirt well ok, would be still a typo
08:29 jhuni sorear: thank you, but it is like 10 at night here lol
08:30 jhuni anyways, recently I have been attracted to the idea of visual programming languages like labview
08:30 sorear jhuni: I meant to reply to jnthn> morning, #perl6
08:31 jhuni k
08:31 sorear but irssi and I do not agree on j<tab> :p
08:31 nadim_ joined #perl6
08:31 jnthn Yeah, it sucks they didn't implement mind-reading yet.
08:31 jhuni sorear: do you ever get the feeling text programming is cumbersome?
08:31 frettled And typo-fixing.  Which fixes unintentional typos only.
08:32 sorear I don't think the text per se is an issue
08:32 sorear I do get annoyed by the one-dimensionality of it from time to time
08:33 frettled I get annoyed by having to spend time typing trivial code.
08:33 sorear it makes layout ... hard
08:33 sorear VQ issues
08:33 frettled I probably should start using an IDE more actively.
08:33 sorear s/layout/organizing stuff/
08:34 sorear I found the Lego Mindstorms IDE quite infuriating when I played with it six years ago.  Has the state of the art in visual programming languages advanced much since then?
08:34 jnthn I quite like having an IDE at some level, but only if it doesn't get in my way too much.
08:35 jnthn Vanilla Visual Studio I can cope with. Start adding tools like Resharper and the annoyance factor is higher than the usefulness.
08:35 sorear I do enjoy a decent IDE, but most of them are far too totalitarian for my tastes
08:35 jnthn Yes. I really hate excessive "I must fix your code layout"-ness.
08:35 jnthn I lay things out the way I do for a reason.
08:36 jnthn I don't really need an IDE thinking it knows better.
08:36 jnthn But, such things can normally be enabled/disabled at will.
08:36 sorear I haven't met an IDE that insists on fixing code layout yet.  If I did there would be significantly fewer intact windows in the city.
08:37 jhuni In general it is cumbersome to have a unique identifiers for things, like generally people use google to search for stuff rather then typing out a url.... similarily it might be nicer to be able to just search for a function rather then knowing a unique identifier, this especially applies to php functions lol
08:38 jhuni a decent IDE or a visual environment provides that with stuff like autocomplete
08:39 sorear Every nontrivial work in a written language creates its own language, augmenting the language with local context serving as definitions of vague or meaningless terms
08:39 sorear A program is no different
08:40 sorear If your program has more functions than words in your setting vocabulary, your program is far too large
08:40 * sorear wonders if there's anything like local $::paren = 0 unless $self->passcap in Perl 6
08:42 moritz_ uhm, s/v/f/ of course :-)
08:46 jjore left #perl6
08:54 am0c joined #perl6
08:55 zby_home_ joined #perl6
08:57 wamba joined #perl6
09:12 * sorear out
09:16 dalek niecza/master: 2a653c8 | sorear++ | / (5 files):
09:16 dalek niecza/master: Implement $<sym>
09:16 dalek niecza/master:
09:16 dalek niecza/master: Also restructures and simplifies the regex CHECK process.
09:16 dalek niecza/master: review: http://github.com/sorear/niecza/commit/2​a653c82015f04fc16770b61da299c930cafe3d2
09:19 robinsmidsrod left #perl6
09:19 robins joined #perl6
09:26 simcop2387 left #perl6
09:26 simcop2387 joined #perl6
09:31 muixirt left #perl6
09:36 simcop2387 left #perl6
09:37 simcop2387 joined #perl6
09:47 pnu joined #perl6
09:59 moritz_ http://nopaste.snit.ch/24655 std build failure (for sorear and TimToady
10:06 QinGW left #perl6
10:07 ggoebel joined #perl6
10:09 moritz_ somehow I can't manage to bisect anything lately
10:10 moritz_ just tried to bisect that "too few positional arguments: 2 passed, 3 (or more) expected" bug when a variable is not predeclared
10:11 moritz_ getting quite a different result than before
10:11 mce_ro left #perl6
10:15 ggoebel left #perl6
10:19 jjore joined #perl6
10:20 agentzh left #perl6
10:22 masak joined #perl6
10:22 masak oh hai, #perl6
10:24 moritz_ lolitsmasak!
10:24 masak moritz_: re TimToady saying that the X** in '[+] @nums X** 2' can be parallelized: he seems to be talking about 'these hyperops' in general, so it kinda holds even though X itself is lazy.
10:24 moritz_ right, X isn't hyper
10:24 masak also, interestingly a "clever optimizer" could notice that the X** is inside a slurpy [+], and turn it into a <<*<<
10:24 masak s/ly/ly,/
10:25 moritz_ <<**>> in genral
10:25 masak troo
10:33 masak TimToady: Java has ++, and has had it since day one, AFAIK. the foreach came in Java 5, but there are no ranges, so it doesn't really help in the example shown.
10:37 moritz_ could anybody else please try to bisect it?
10:41 baux joined #perl6
10:45 ggoebel joined #perl6
10:46 colomon joined #perl6
10:46 masak ah, so *that*'s where the "Rakudo is fast as a turtle" meme comes from.
10:46 masak TimToady running his turtle on Rakudo and jokingly pointing out when Parrot does a GC run.
10:47 moritz_ yes, I saw it on youtube too
10:50 masak leaving all unknowable aspects of "which end of the Rakudo/Parrot spectrum is causing the slowness?" question aside, surely there's objective factual evidence that Parrot's GC is... heavily suboptimal?
10:50 ggoebel left #perl6
10:51 flussence left #perl6
10:51 jhuni left #perl6
10:52 colomon masak: is there?  I know I used to blame all the slow on parrot, but I'm not so sure anymore.
10:52 masak I'm not blaming "all the slow" on Parrot.
10:52 moritz_ masak: do you remember pmichaud's mail about GC pressure to parrot-dev?
10:52 masak I'm saying that those almost second-long pauses in TimToady's turtle program are caused by the GC.
10:53 masak moritz_: not immediately, no.
10:53 moritz_ masak: loading perl6.pbc slowed down an othewise fast program by a facto of >10
10:53 moritz_ excluding startup time
10:53 masak I'm also saying that I don't recall the JVM having such long and such frequent GC pauses.
10:53 masak moritz_: ah. it's coming back now.
10:53 moritz_ just because the objects were considered during each GC run
10:54 colomon moritz_: woah.
10:56 moritz_ "A short example of Parrot GC pain"
10:56 flussence joined #perl6
10:56 masak oh, that's why? I never grokked that part.
10:56 masak making a VM is hard. let's go shopping.
10:57 moritz_ shop a VM?
10:57 masak I'm not at the point yet where I'm dismissing Parrot altogether. it might just be a great project with a very long buildup.
10:59 moritz_ well, parrot does have some important features that other VMs lack
10:59 jnthn masak: Me either, but I'm also very keen to break having an absolute dependency on it.
10:59 moritz_ like native support for multi dispatch
10:59 jnthn moritz_: lol
11:00 jnthn moritz_: You do know that we don't use Parrot's multi-dispatch at all in Rakudo?
11:00 colomon jnthn++
11:00 moritz_ jnthn: yes
11:00 masak "native support for multi dispatch" sounds like something a VM *designed with Perl 6 in mind* should have... :/
11:00 jnthn :S
11:00 moritz_ jnthn: but it means that other languages have an easy time using the multi dispatcher from Perl 6
11:00 masak in Parrot's defense, what Perl 6 is keeps changing slightly.
11:01 jnthn moritz_:
11:01 jnthn What really matters here is that there's an API for invocation.
11:02 moritz_ right
11:03 * moritz_ wonders if there's an API for innovation too :-)
11:07 gfldex well, the patent office got a public interface
11:07 gfldex that's a start :)
11:09 moritz_ well, in programming patents are pretty much hinder innovation
11:09 masak this shows pretty clearly the effects of the GC: http://gist.github.com/638298
11:09 masak nom &
11:10 ggoebel joined #perl6
11:11 mdxi left #perl6
11:12 kjeldahl joined #perl6
11:14 ggoebel left #perl6
11:18 nsh_ joined #perl6
11:18 \xF0_ joined #perl6
11:19 nsh left #perl6
11:19 jfried left #perl6
11:19 \xF0 left #perl6
11:19 rgrau left #perl6
11:21 jfried joined #perl6
11:25 Trashlord joined #perl6
11:27 felipe joined #perl6
11:27 mdxi joined #perl6
11:29 colomon masak: just tried that code out here, UGH!
11:34 allbery_b left #perl6
11:38 xinming joined #perl6
11:38 nymacro joined #perl6
11:39 colomon I mean, I'm seeing approx a one second pause for GC every three seconds.  That means GC is adding 33% to the execution time.
11:40 cognominal joined #perl6
11:40 colomon I wonder how many objects are created per step?
11:44 pnu left #perl6
11:46 masak that's what I'm talking about. ugh.
11:46 * masak writes to the parrot-dev thread about this
11:49 colomon it's also using up 366 MB of memory basically to create a linked list which now has 16,000 elements in it.
11:49 colomon on the plus side, memory usage doesn't seem to be growing very quickly at all, so the GC is at least accomplishing something.
11:50 Trashlord left #perl6
11:50 takadonet morning all
11:50 jnthn This is the kinda thing that bacek++'s much-needed current work on generational GC is aimed at improving.
11:53 mj41 left #perl6
11:59 allbery_b joined #perl6
11:59 Trashlord joined #perl6
12:04 aloha left #perl6
12:05 colomon hmmm, wasn't watching closely enough to see when it happened, but we're now at 378 MB memory usage in the run, and 26,000 runs through the loop.
12:06 masak rakudo: say (1 < 3 > 2)
12:06 p6eval rakudo d35769: OUTPUT«Bool::True␤»
12:06 colomon that means the last 10,000 elements in the linked list cost us 10 MB of memory.  somewhere on the order of 1K memory per element.
12:10 jnthn What are the elements?
12:10 Guest84593 joined #perl6
12:10 masak *all* of them? :)
12:10 jnthn (what data type)
12:10 colomon jnthn: just the internal iterator created by  1..Inf
12:11 jnthn On, Int then
12:11 jnthn *OK
12:11 jnthn That's...pricey.
12:12 colomon it may be a bad assumption -- could be the GC isn't doing a good job of deleting dead variables
12:12 jnthn OTOH, today one Rakudo Int is 3 PMCs and that's before considering that we may have a scalar container wrapper in the array too iirc.
12:13 jnthn Which if it has an rw property marked on it is...2 more PMCs (one for container, one for prop hash)
12:13 jnthn If you add that lot up you might get closer to 1K. :/
12:14 colomon still, in any sane world that would be under 100 bytes.
12:15 aloha joined #perl6
12:15 colomon and one would hope a carefully built implementation was more like 24 bytes
12:17 jnthn I'm carefully building. :-)
12:17 colomon ;)
12:17 Juerd Jenga
12:17 jnthn This is one place where 6model should give us a nice win.
12:17 jnthn heh..."Jenga Perl 6"
12:18 Juerd A building project in Rotterdam collapsed today. 5 injured.
12:18 MayDaniel joined #perl6
12:19 jnthn :(
12:19 Juerd Some sources say only the scaffolding collapsed, most indicate it's the building they were building that went down.
12:19 colomon :(
12:20 Juerd Maybe I'm just naive but couldn't someone easily tell the difference between the two?
12:20 Juerd Even without knowledge of architecture... :)
12:20 Juerd (concrete knowledge...)
12:21 Schwern joined #perl6
12:24 jnthn Lack of concrete knowledge may have been the problem here ;-)
12:28 aloha left #perl6
12:29 masak Tom Christiansen is on StackOverflow. \o/ http://stackoverflow.com/questions/3973834​/how-do-i-create-a-perl-regex-that-matches​-non-alphanumeric-characters-except-spac
12:35 MayDaniel left #perl6
12:36 bluescreen joined #perl6
12:37 aloha joined #perl6
12:40 leprevost joined #perl6
12:44 masak merlyn++ (Randal L. Scwartz) did a FLOSS Weekly inteview about Rakudo Perl 6.
12:44 masak I wonder why he's not on here. I think he's on freenode.
12:45 ggoebel joined #perl6
12:47 pythonian4000 is now known as pythonian4000afk
12:55 wamba left #perl6
12:57 colomon left #perl6
12:58 takadonet masak: url?
12:59 masak takadonet: http://twitter.com/merlyn/status/27960589524
13:02 ggoebel left #perl6
13:04 bluescreen left #perl6
13:05 bluescreen joined #perl6
13:09 ggoebel joined #perl6
13:14 zby_home_ left #perl6
13:14 Schwern left #perl6
13:15 tzhs joined #perl6
13:16 PerlJam good morning #perl6
13:16 masak PerlJam! \o/
13:18 PerlJam masak: today is going to be a busy day for me it seems.  Code to write. Interviews to conduct.  oh, and a compiler to "release"
13:18 PerlJam :-)
13:18 leprevost left #perl6
13:19 PerlJam I won't be able to get to rakudo until later today local time, so it may be "tomorrow" for some of our denizens
13:20 takadonet NO!!!! hehe
13:20 Patterner left #perl6
13:20 pmichaud good morning, #perl6
13:20 masak PerlJam: the Europeans forgive you, I'm sure.
13:21 masak pmichaud! \o/
13:21 jnthn Morning, pmichaud
13:23 PerlJam (Why do we call them "releases" anyway?  That only seems to fit under the old regime of "only release when we've built up enough features to warrant it".  Back  then we were holding back until "good enough" and then releasing it out into the world.  With clockwork releases it's more like we're kicking it out the door rather than letting it go.)
13:24 aloha left #perl6
13:25 aloha joined #perl6
13:26 flussence interesting question... it seems like there's a trend, everywhere, toward time-based releases instead of good-based releases :/
13:27 flussence (I think they should be called "snapshots" if they're time-based)
13:28 PerlJam I don't know ... I don't think time-based releases are the antithesis of "good-based" releases.  I think it's more that we're recognizing smaller units of "good"
13:28 plainhao joined #perl6
13:28 ruoso joined #perl6
13:29 flussence with rakudo it's not polar opposites, they do get delayed after all :)
13:29 ruoso left #perl6
13:29 ruoso joined #perl6
13:32 Psyche^ joined #perl6
13:32 Psyche^ is now known as Patterner
13:37 nymacro left #perl6
13:44 masak PerlJam: for better or worse, "release" is a magic word. even knowing this, I tend to look at an email subject line saying "ANNOUNCE: Parrot/Inkscape/MediaWiki/Perl release x.y.z", and automatically go "ah! maybe it's time to upgrade, then". even when I decide not to upgrade, a little flag in my brain saying "this project is alive and well" gets set. that's worth a lot.
13:45 moritz_ rakudo: say now.x
13:45 p6eval rakudo d35769: OUTPUT«1287668749.37868␤»
13:45 PerlJam masak: re merlyn, he's on #git sometimes and #squeak or #smalltalk or something too (as RandalSchwartz), but another #perl on another network soured him a little I think, so he doesn't hang out on #perl-ish channels much
13:47 zby_home joined #perl6
13:48 masak PerlJam: right, I almost suspected that. and yet he's writing a book about Perl 6, and giving interviews about Rakudo. as usual, it just feels so strange when people who are involved with Perl 6 in various ways don't have a presence here.
13:48 PerlJam masak: you have proof that he's actually writing a book ?  :)
13:49 moritz_ not only that, but don't have a presence on any perl6- mailing list either
13:49 pmichaud I suspect that may become the norm someday, though.
13:49 Juerd Heh
13:49 masak PerlJam: I think it was from merlyn that I got a long excerpt about Perl 6's history, yes. that seemed to be part of a book.
13:49 pmichaud I mean, how many people develop things for Perl 5 but aren't active in p5p or #perl or .... :-P
13:49 PerlJam He's been talking about "Learning Perl 6" for several years yet, but I don't think he's actually working on it except sporadically.
13:50 masak don't recall anymore where I saw that excerpt.
13:50 pmichaud He and I talked about that yesterday (outside of the interview)
13:50 moritz_ pmichaud: "developing *for*" ne "developing"
13:50 Juerd I fear 5-to-6 translations of existing Perl books.
13:50 _kaare left #perl6
13:50 masak pmichaud: were you in the interview as well?
13:50 PerlJam Juerd: I only fear *bad* translations  :)
13:50 PerlJam heh
13:51 pmichaud he's been (understandably) waiting for an alpha release of Perl 6 before really focusing on Learning Perl 6
13:51 Juerd Like how some Perl 5 books are updated Perl 4 books.
13:51 PerlJam there's that word again ... "release"
13:51 pmichaud masak: merlyn and I recorded an episode of "FLOSS weekly" about the history of Rakudo and Perl 6
13:51 Juerd New features are discussed in separate chapters or not at all, and all the existing stuff is still 4-ish
13:52 Juerd PerlJam: Do you think we should replace that with real ease?
13:52 pmichaud Juerd++  # nice
13:52 masak pmichaud: oh, cool! looking forward to hearing it.
13:52 Juerd "You keep asking about real ease, but we already have that."
13:53 * moritz_ too
13:53 Juerd +the
13:53 pmichaud it really ended up being more about Perl 6 history than anything else; hope I didn't mangle the facts too much.  :)
13:53 moritz_ pmichaud: I don't have big concerns in that respect
13:53 Schwern joined #perl6
13:54 PerlJam Schwern!
13:54 masak Schwern! \o/
13:54 Juerd Is there any other "Vaporware" project that has such a long and lively history, active development and a regular release schedule? :)
13:55 PerlJam Juerd: I doubt it.
13:55 masak Juerd: I think we should put some effort into finding out.
13:55 Juerd And books
13:55 masak Juerd: I keep a list of candidates for exactly that purpose.
13:55 Juerd :)
13:56 masak Juerd: Duke Nukem Forever (recently re-activated), Apache 2, Mozilla, the Hurd, Xanadu, grub2, the whole thing with 5th-gen computers in the 1980's.
13:57 Juerd I use Apache 2 and Grub 2 in production, so that must mean Perl 6 is safe for production use as well.
13:57 masak at some point, I'd like to do a "compare-and-contrast" article-length blog post about all those projects.
13:57 frettled PHP 6!
13:58 masak frettled: thanks; added.
13:58 masak just to explain (to myself, if to no-one else) where Perl 6 fits into it all.
13:58 frettled It was _started_ around when Perl 6 was announced.
13:58 masak Python 3k also added.
13:58 frettled :D
13:58 frettled Ruby 2?
13:58 x3nU i wouldn't use rakudo in production ;f
13:59 masak x3nU: then you're not like me :)
13:59 x3nU it's too heavy imho
13:59 frettled masak: ooh, I have another one: the GUI replacement for X11.
13:59 frettled X12 or whatever :D
13:59 x3nU of course for some uses it's not problem
13:59 masak frettled: you think you could find its actual name?
14:00 PerlJam "too heavy"?
14:00 masak x3nU: that's the key thing -- it depends on who's the one using the word "production".
14:00 Juerd PerlJam: There's only so much weight a 19" server cabinet can carry.
14:00 frettled masak: It's so vaporvare that it doesn't even exist in name, IIRC.
14:01 Juerd One must always be careful weigh and balance new software before putting it in.
14:01 PerlJam Juerd: mine carries thousands and thousands of software packages. :)
14:01 frettled masak: but have a look-see here: http://en.wikipedia.org/wiki/X​_Window_System#Release_history
14:01 Juerd PerlJam: Sure, so does mine if they're feather light packages
14:01 masak frettled: thanks; added.
14:02 Juerd PerlJam: df --weight
14:02 masak could anyone help me find the actual keynote hinted at in http://www.ruby-lang.org/en/news/2010/10/01/​see-matz-in-san-francisco-or-silicon-valley/ ?
14:02 Juerd Include -h if you want kg instead of g
14:02 masak surely it must be online as a video somewhere?
14:02 * PerlJam heads off to interview a potential ISO
14:03 frettled masak: Seven years between X11R5 and X11R6 was pretty bad, people started claiming that X/X11 «is dead», that the concept itself was dated, that a new and perhaps incompatible release would not work out, etc.  People also seem to agree that the X11 model is fundamentally flawed, but at a loss when asked to provide a functional alternative.  :)
14:03 masak recognizable situation.
14:03 moritz_ PerlJam: how is the release going?
14:04 frettled Hmm.  Emacs 19 -> Emacs 20, perhaps?
14:04 masak possibly.
14:05 masak I'll write that one down as a candidate.
14:05 moritz_ plan 10!
14:05 frettled Nah, that wasn't so bad, not too many years.
14:05 masak moritz_: :P
14:05 masak moritz_: to their credit, they did release a 9 1/2...
14:06 frettled heh
14:06 flussence xmms 2?
14:06 flussence (which I've been using for months now...)
14:08 masak appears to be a complete rewrite.
14:08 masak yeah, might qualify :)
14:09 masak I'm glad I brought this up. my list is twice as long now. :)
14:09 flussence ooh, add e17 too
14:10 moritz_ indeed
14:10 masak noted.
14:12 ruoso left #perl6
14:13 ruoso joined #perl6
14:13 ruoso left #perl6
14:13 alester joined #perl6
14:13 timbunce joined #perl6
14:17 ruoso joined #perl6
14:20 * moritz_ started to take a course on Norwegian, and now ponders writing a vocabular training program in Perl 6
14:21 * moritz_ is a bit puzzled that Compose + ° + a doesn't produce å by default
14:23 MayDaniel joined #perl6
14:23 zostay left #perl6
14:23 icwiener joined #perl6
14:23 masak moritz_: I recognize the feeling -- almost always when I learn a language, I want to write software to help me learn it.
14:23 masak I seldom get around to it, however.
14:25 Juerd moritz_: Use a bigger °, like o
14:26 moritz_ Juerd: I used Compose a a now
14:26 Juerd Doesn't work here
14:26 Juerd compose o a does
14:26 moritz_ both work for me
14:26 moritz_ discovered aa in the vim :digraphs table
14:27 zostay joined #perl6
14:32 masak aa makes sense from a historical perspective.
14:33 flussence altGr + {,a for me... all the extra bits are on punctuation keys
14:33 ruoso left #perl6
14:33 ruoso joined #perl6
14:35 * ruoso trying to use irssi instead of xchat... local connection is weird, better using ssh to a machine with a saner connetion
14:40 masonkramer joined #perl6
14:41 masak rakudo: my @t = 1..10; say (my &stddev)(@t) given &stddev = -> @x { sqrt ([+] (@x X- my $mean) X** 2) / (@x - 1) given $mean = ([+] @x) / @x }
14:41 p6eval rakudo d35769: OUTPUT«3.02765035409749␤»
14:42 * masak cackles evilly
14:43 jnthn eww!
14:43 masak :D
14:43 jnthn But hey, it works...
14:43 masak credit goes to TimToady++ for developing that style of writing.
14:43 icwiener left #perl6
14:44 masak it's a bit similar to 'where' in Haskell.
14:44 icwiener joined #perl6
14:45 masak rakudo: my @t = 1..10; say .(@t) given -> @x { sqrt ([+] (@x X- $_) X** 2) / (@x - 1) given ([+] @x) / @x }
14:45 p6eval rakudo d35769: OUTPUT«3.02765035409749␤»
14:46 masak one can think of it as a sort of "named $_" pattern.
14:46 MayDaniel left #perl6
14:47 jnthn yeah
14:47 jnthn On second glance it's not quite so bad.
14:48 masak I like it.
14:48 flussence I can almost understand that!
14:48 masak "declare now, define later"
14:48 flussence (which is good, given I never understood it in school)
14:48 masak oh, you mean the standard deviation.
14:49 flussence yeah, the thing.
14:49 masak it's just a maths formula. chances are the parts you don't understand are just-so magic anyway :)
14:50 masak but basically, if you assume that your samples follow a gaussian distribution, the standard deviation is a measure of the "width" of the bell curve.
14:51 moritz_ \o/ first, primitive version of vocabular trainer is running
14:51 masak moritz_++
14:51 masak moritz_: when will we see the source?
14:52 masak some of us have been waiting for hours already :P
14:52 masak there are murmurs about vapourware...
14:52 moritz_ masak: hours? you mean, about 0.5 hours?
14:53 masak it *feels* like hours. :)
14:54 masonkramer left #perl6
14:54 masonkramer joined #perl6
14:55 moritz_ http://github.com/moritz/perl6-vocabular-coach
14:55 moritz_ currently it's *very* primitive, and stateless
14:56 * masak reviews
14:57 masak I like the use of smilies.
14:57 jnthn I need to make one of these for Russian, but that can test me on declensions. :-)
14:58 justatheory joined #perl6
14:58 jnthn .oO( Trombone (Instrumental): )
14:58 colomon joined #perl6
15:00 jnthn @lang[$_].=trim for ^2;
15:00 jnthn Why now
15:00 jnthn @lang>>.=trim;
15:00 jnthn *not
15:00 moritz_ does that work?
15:00 moritz_ rakudo: my @a = ' a ', ' b'; @a».=trim; say @a.perl
15:00 p6eval rakudo d35769: OUTPUT«["a", "b"]␤»
15:01 masak rakudo: 'my $pair = "OH HAI" => 42; my (Pair($key, $value)) := $pair; .say for $key, $value # jnthn: what am I doing wrong here? I want to "automatically" extract $key and $value
15:01 p6eval rakudo d35769: OUTPUT«===SORRY!===␤Confused at line 22, near "'my $pair "␤»
15:01 orafu joined #perl6
15:01 masak rakudo: my $pair = "OH HAI" => 42; my (Pair($key, $value)) := $pair; .say for $key, $value
15:01 masak er.
15:01 p6eval rakudo d35769: OUTPUT«Not enough positional parameters passed; got 0 but expected 2 in sub-signature␤  in 'infix:<:=>' at line 1␤  in main program body at line 22:/tmp/gnS7v6hB9V␤»
15:01 jnthn Missing space?
15:01 masak rakudo: my $pair = "OH HAI" => 42; my (Pair ($key, $value)) := $pair; .say for $key, $value
15:01 p6eval rakudo d35769: OUTPUT«Not enough positional parameters passed; got 0 but expected 2 in sub-signature␤  in 'infix:<:=>' at line 1␤  in main program body at line 22:/tmp/qqm0K1W6eU␤»
15:02 jnthn Named params?
15:02 jnthn though hm
15:02 masak rakudo: my $pair = "OH HAI" => 42; my (Pair (:$key, :$value)) := $pair; .say for $key, $value
15:02 p6eval rakudo d35769: OUTPUT«Mu()␤Mu()␤»
15:02 masak better, though.
15:02 jnthn rakudo: my $pair = "OH HAI" => 42; say $pair.Capture.perl
15:02 p6eval rakudo d35769: OUTPUT«\()␤»
15:02 moritz_ I think the Capturification of Pair is broken
15:02 moritz_ right
15:02 jnthn Yeah
15:03 * masak submits rakudobug
15:03 jnthn That's the real issue.
15:03 jnthn It woulda worked if that had.
15:03 * PerlJam returns
15:04 tzhs left #perl6
15:06 jnthn Oh, I think I remember us diagnosing this one before too.
15:08 masak I can make a quick search. I've already submitted it.
15:11 masak rakudo: my $code = "\$" ~ 1 x 1024; eval $code
15:11 p6eval rakudo d35769: OUTPUT«error:imcc:syntax error, unexpected IDENTIFIER ('Inf')␤   in file 'EVAL_11' line 68␤»
15:12 jnthn wtf :-)
15:13 masak an old one, reported by bbkr :)
15:13 masak at least it's clever. probably not very common, though.
15:13 masak but I agree that it's a bug.
15:14 jfried left #perl6
15:17 jnthn *yawn* almost home time...
15:17 * jnthn mighta been done for the day if Snaketrafiken weren't so crap at getting him into Malmo on a morning. :|
15:20 timbunce left #perl6
15:21 * moritz_ now added ASCIIficiation to his comparison routines
15:23 MayDaniel joined #perl6
15:26 timbunce joined #perl6
15:28 jnthn decommute
15:28 Ross joined #perl6
15:30 masak Parrot GC runs last for about a second, with a standard deviation of about a quarter of a second: http://gist.github.com/638685
15:31 masak needless to say, this is the kind of "statistics" of the "damnd lies" kind.
15:31 masak s/damnd/damned/
15:31 moritz_ masak++ # hard numbers about GC
15:31 moritz_ fsvo "hard" :-)
15:31 masak right.
15:32 masak biggest measuring error is probably that the logic inside the loop takes time, too.
15:32 risou joined #perl6
15:32 masak probably saving all the timestamps and processing them afterwards might be better.
15:32 masak but then memory consumption might be an issue instead :/
15:33 * masak runs it once without IO, too.
15:36 masak in my humble opinion, the fact that the GC is invoked 28 times in a simple 1..1000 `for` loop on a 4GB computer, is a kind of crazy deserving of its own category of crazy.
15:37 masak similar results without IO: http://gist.github.com/638685
15:41 [particle]1 left #perl6
15:42 masak keeping all timestamps and doing sorting later also gives similar results: http://gist.github.com/638685
15:42 [particle] joined #perl6
15:43 masak after this, I won't hesitate saying "Parrot GC takes one second".
16:00 bluescreen left #perl6
16:15 robins is now known as robinsmidsrod
16:15 takadonet ...
16:16 stkowski joined #perl6
16:16 _kaare joined #perl6
16:16 Amine1285 joined #perl6
16:18 Amine1285 Depending where you are, Good morning or good evening!
16:18 Amine1285 I am new here
16:18 Amine1285 and realy like to do something with perl for perl
16:18 Amine1285 I like perl
16:19 takadonet Amine1285: hey
16:19 masak hi there! we like Perl too :)
16:19 Amine1285 Hi :)
16:19 masak rakudo: say "oh hai Amine1285!"
16:19 p6eval rakudo d35769: OUTPUT«oh hai Amine1285!␤»
16:20 masak Amine1285++
16:20 masak aloha: karma Amine1285
16:20 aloha masak: Amine1285 has karma of 1.
16:20 masak consider that a welcome bonus :)
16:20 Amine1285 :D
16:20 Amine1285 what  does it mean karma of 1?
16:21 masak it means that we think you're awesome.
16:21 masak just wait until you get a karma of 2!
16:21 masak aloha: karma takadonet
16:21 aloha masak: takadonet has karma of 3.
16:21 jnthn aloha: karma jnthn
16:21 aloha jnthn: jnthn has karma of 118.
16:22 jnthn I'm SO awesome!
16:22 jnthn ;-)
16:22 masak jnthn++ # has a lot of karma. oh wait
16:22 jnthn Maybe we should track modesty too ;-)
16:22 masak we do. we just haven't been letting you know :P
16:22 * jnthn feels negative
16:22 jnthn :-)
16:23 takadonet i have 3 karma.... wow
16:24 jnthn masak: Those GC numbers are...eww.
16:24 masak Amine1285: now -- how can we assist you on your quest?
16:24 Amine1285 Ok
16:24 Amine1285 I understand now
16:24 Amine1285 mmm
16:24 masak Amine1285++ # understands now
16:24 masak aloha: karma Amine1285
16:24 aloha masak: Amine1285 has karma of 2.
16:24 Amine1285 I just want to be involved in perl6
16:24 masak :)
16:25 Amine1285 because I like perl
16:25 Amine1285 :P
16:25 masak Amine1285: easiest way to do that is to write Perl 6 code. doesn't have to be advanced in any way.
16:25 masak just start somewhere, and before you know it -- you'll be involved.
16:25 Amine1285 is it different from perl 6?
16:25 Amine1285 I mean 5
16:25 masak yes and no.
16:25 Amine1285 beacue I already write in perl 5
16:26 masak if you go into Perl 6 assuming it's Perl 5, you'll get some surprises.
16:26 Amine1285 but when I want to start perl6 I notice that is still "under construction"
16:26 masak it is.
16:26 masak but people've constructed a great lot already.
16:27 Amine1285 I have a stupid question
16:28 Amine1285 why we move to perl 6?
16:28 masak because there is room for improvement.
16:28 Amine1285 i.e.?
16:28 masak Perl 5 is great, but some aspects of it are limited and limiting.
16:29 Amine1285 like?
16:29 masak many aspects.
16:29 masak OO
16:29 masak parsing
16:29 takadonet grammars!!!!!
16:29 masak signatures
16:30 masak language extensibility
16:30 PerlJam Amine1285: also note that Perl 5 and Perl 6 live comfortably together.  There's no mutual exclusion.
16:31 masak Amine1285: what PerlJam said. if you go into a Perl 5 channel and ask "why we move to perl 6?", they answer "we don't".
16:31 aloha left #perl6
16:31 masak there are many use cases for Perl 5, and many for Perl 6.
16:31 masak some people will stick with one or the other simply because they are more comfortable with that language.
16:32 PerlJam People who are comfortable haven't been pushing the boundaries of their language too much  :)
16:33 Amine1285 I have also one other stupid question :s
16:34 Amine1285 why Perl is much slower than Java
16:34 Amine1285 ?
16:34 Amine1285 perl 6 resolve this?
16:35 PerlJam That's actually an interesting question.
16:36 PerlJam For most of Java's lifetime, it's actually been the reverse.  Perl has been faster than Java.
16:36 masak Amine1285: you'll find that Perl 5 is really really fast. it solves most problems while the JVM is still booting up.
16:37 Amine1285 there is one test on the net, that show that java could achieve task in seconds, while perl hang for ages...
16:37 PerlJam Amine1285: One sample does not a trend make  :)
16:38 Entonian joined #perl6
16:38 PerlJam Amine1285: The current Java compilers may be faster than Perl (I don't have any hard numbers handy so I don't know for sure) because just-in-time compilation.
16:38 masak Amine1285: Perl 5 is really fast. especially after one has gotten used to the speed of Rakudo.
16:39 Amine1285 Rakudo?
16:39 s1n left #perl6
16:40 masak rakudo: say "that's me!"
16:40 p6eval rakudo d35769: OUTPUT«that's me!␤»
16:43 PerlJam Amine1285: rakudo is the front-runner perl 6 compiler (for now :)
16:45 _buno_ joined #perl6
16:46 _buno_ left #perl6
16:47 pmichaud masak++ # very cool gc timings
16:47 masak thanks.
16:49 pmichaud I agree, the fact that GC is invoked 28 times in a 1000-iteration loop is.... bizarre.
16:49 inket joined #perl6
16:50 masak it's tragic.
16:50 inket is there a differenc ebetween scripting languages and general programming languages
16:50 inket or they all come under the same folder
16:50 masak inket: the line is really blurry, I'd say.
16:50 masak a script is just a small program.
16:50 inket right
16:50 inket perl is a scripting language isnt it
16:50 inket not a programming language
16:51 masak it's both.
16:51 inket but in a resume
16:51 inket what do you put it under
16:51 masak :)
16:51 masak depends how you use Perl, I guess.
16:51 masak if you never write larger things with it, put it under "scripting language".
16:51 inket but in a resume do you generally have to disntinguish the two
16:51 masak why?
16:51 inket as in scripting languages and regular joe programming languages?
16:52 inket oh dunno just thought people take a liking to that because you know the difference
16:52 arnsholt My resume just has a list of the most relevant/interesting programming languages I know
16:52 masak inket: I think you read too much into the difference.
16:52 inket but nothing wrong in it is it
16:52 inket just adds more value me thinks
16:53 masak sure. unless you think that scripting languages aren't programming languages.
16:53 inket atleast gives the hiring committee an idea at a glance that i know it's different and not the same
16:53 inket tcl is not a programming language
16:53 * masak sighs
16:53 inket if it is, why is it called a scripting language
16:54 flussence I'd put perl under both
16:54 masak some programming languages are called scripting languages.
16:54 inket thats practical, putting it under both
16:55 masak Perl is all about practicality :)
16:55 justatheory left #perl6
16:56 inket perl wasa text processing language
16:56 inket but over time itbecame an all purpose lanbguage
16:56 masak "was"?
16:56 inket not the same with a scripting language like tcl
16:56 inket or say bash
16:56 timbunce left #perl6
16:56 inket yeah was, its more a general purpose language now
17:00 masak I see the usefulness of distinguishing programming languages from things that aren't programming languages (like HTML, or CSS). you should definitely distinguish those on your resumé.
17:00 masak but "scripting languages" are just a loosely-defined set of programming languages.
17:01 timbunce joined #perl6
17:01 masak "scripting" itself is a loosely defined activity. it's a bit like talking about "bioinformatics languages". which ones are and which ones aren't depends on who's asking.
17:03 ruoso left #perl6
17:03 inket true
17:03 ruoso joined #perl6
17:03 masak evening &
17:03 masak left #perl6
17:06 dakkar left #perl6
17:07 aloha joined #perl6
17:07 inket isnt tcl different to php when it comes to scripting?
17:07 inket both TCL and PHP both have interpreters which take the script, and both provide output based on it, they should be the same then right?
17:07 cdarroch joined #perl6
17:07 cdarroch left #perl6
17:07 cdarroch joined #perl6
17:08 patrickas joined #perl6
17:09 patrickas o/
17:09 flussence inket: they're built for different purposes
17:09 flussence for example, Tcl's logo is a feather, while PHP's is an elephant. I think that explains the difference pretty well
17:10 patrickas TimToady ping
17:10 inket no but they both have interpreters which take the script
17:10 inket php is considered a general purpose language than a scripting while tcl is just scripting, weird
17:10 Schwern left #perl6
17:11 flussence PHP also has the roadsend compiler, and some other one(s)
17:11 inket so is it fair to clal it just a scripting language?
17:12 x3nU what
17:12 x3nU >php is considered a general purpose language
17:12 flussence "scripting language" is just a vaguely defined subset of programming languages
17:12 Amine1285 left #perl6
17:13 inket ok what is the different between tcl and php and perl though
17:13 inket all scrioting but major different being
17:13 x3nU php sucks while perl not
17:13 inket they all serve the same purpose, you give it input, it gives output
17:13 x3nU eot
17:13 x3nU ;)
17:13 inket and they all have bindings for many tasks
17:21 inket java and C are general purpose languages, I mean they're not pure compiled languages.. they run in VM's
17:21 pmichaud wikipedia's article on "scripting languages" isn't too bad
17:21 inket wouldnt they be scripted?
17:21 pmichaud http://en.wikipedia.org/wiki/Scripting_language
17:21 inket right
17:21 inket by definition tey say tcl and php are
17:21 Juerd x3nU: More often than not, PHP is considered a web programming language and not at all general purpose.
17:21 inket but wouldnt java and C be too then?
17:21 inket they';re not exactly general purpose languages are they
17:22 pmichaud I wouldn't consider java and C as scripting languages, no.
17:22 inket java is
17:22 inket pmichaud: they're not compiled languages
17:23 Juerd In my book, scripting is extending existing programs with code that is evaluated (whether compiled or not) during these programs' runtime.
17:23 inket they run ina  VM
17:23 pmichaud running in a VM isn't the defininition of "scripting language"
17:23 inket what is
17:23 Juerd So with Perl, it's definitely scripting if you're building extensions for Irssi, but it's not scripting if you're building a standalone program.
17:23 pmichaud in the case of java, you run a compiler on the source, producing bytecode, this is then independently run on a VM
17:23 diakopter I think the simplest definition is that a scripting language's source files can be "executed" directly from a shell by an "interpreter/compiler" without necessarily requiring a *visible* compilation step
17:23 Juerd Perl happens to be great for both :)
17:23 Juerd diakopter: So with Inline::C, C becomes a scripting language? :)
17:24 timbunce left #perl6
17:24 inket you can not compile Perl
17:24 inket it is still interpreted
17:24 Juerd diakopter: And with explicit compilation, Python's no longer a scripting language?
17:24 patrickas pmichaud o/ how is the cold going ?
17:24 diakopter Juerd: what? I said "*necessarily*"
17:24 inket it still runs as a script
17:24 Juerd inket: Actually, Perl code is compiled before it is executed. It's compiled over and over and over, once for every execution, but it still is compiled.
17:24 pmichaud patrickas: cold is largely gone, but replaced by some other issues :(
17:25 inket so diakopter, pmichaud: a scripting language still requires a 3rd party program to run
17:25 inket Juerd: either perl, something built into irssi that calls perl, what ever
17:25 inket Juerd: so is php and the rest
17:25 huf_ so is c, by that logic. good luck running most things without a libc and friends
17:25 huf_ along with the kernel
17:26 patrickas well I hope the other issues will follow the cold soon!
17:26 Juerd I'm not sure about current versions of PHP, but 3 years ago PHP was, by default, not compiled to an intermediate format before being executed.
17:26 huf_ Juerd: it is now
17:26 pmichaud has been for quite a while, iiuc
17:26 Juerd That's wonderful for those who have to work with PHP
17:26 huf_ it's compiled on every page request by default, but it is.
17:26 inket huf_: If I compile my C app, I can remove all external dependencies from the binary.. C can run "stand alone"
17:26 inket php can not
17:26 inket perl can not
17:26 huf_ inket: depending on what you mean by standalone
17:27 inket java can not
17:27 diakopter inket: my definition was "from source files"
17:27 Juerd I remember installing something called, APC, which was a PHP compiler, and getting an enormous performance boost for Mediawiki
17:27 huf_ inket: as i said, most c code still depends on kernel and libc, so not standalone
17:27 inket Juerd: yes it is
17:27 Juerd (A much needed performance boost, I might add)
17:27 inket The script is compiled and optimized into optcode
17:27 huf_ inket: you can compile c to run truly standalone, but you could make such an environment for perl too
17:27 diakopter inket: I stand by my simple definition of the source files being involved directly in the initial invocation.
17:27 inket huf_: ALL code requires a kernel of some sort
17:27 huf_ inket: nope
17:28 inket ?
17:28 huf_ there's no kernel under the bit of code that boots your pc
17:28 huf_ altho even that's not standalone :)
17:28 huf_ (it uses the BIOS)
17:28 huf_ and whatever else is there, i have no idea
17:28 inket I can compile my C application to not require in any libc libraries but rather include the desired functionality as part of the resulting binary file
17:28 huf_ inket: pp can do the same for perl scripts, iirc
17:28 pmichaud I've certainly written programs that didn't require external libraries of any sort.
17:28 inket pp?
17:29 pmichaud (not recently, and not on modern OS's, but I've certainly done it before :-)
17:29 inket huf_: not for perl
17:29 inket I have seen some hackish perl -> binary compilers
17:29 huf_ inket: what?
17:29 inket and some php -> binary as well honestly
17:29 timbunce joined #perl6
17:30 diakopter inket: hackish?
17:30 inket but "Out of the box" no, you're dead wrong
17:30 huf_ inket: so you did not bother to look up what pp is.
17:30 patrickas does anyone know what is (64,32,16 ...^ Rat) supposed to return? that's RT:78324 ... Is 64,32,16 the right answer ?
17:30 diakopter inket: see Reini Urban's work
17:30 huf_ inket: the distinctions you're trying very hard to make dont seem to make sense... abandon them :()
17:30 inket hackish as in using something for something its not designed for
17:30 huf_ :)
17:30 inket link me diakopter
17:30 flussence C wasn't designed for use outside a PDP-11, for instance.
17:31 flussence *shrug*
17:31 pmichaud patrickas: 64, 32, 16, 8, 4, 2, 1
17:31 diakopter inket: http://search.cpan.org/sear​ch?query=B::C&amp;mode=all
17:31 inket diakopter: got the link/
17:31 inket thx
17:32 pmichaud patrickas: i.e., it should stop when the series produces a Rat, which would be 1/2
17:32 inket pmichaud: yeah you can
17:32 inket you dont need external libs and such as huf_ puts it
17:32 patrickas rakudo: say (16/2).WHAT
17:32 p6eval rakudo d35769: OUTPUT«Rat()␤»
17:32 inket This compiler backend takes Perl source and generates C source code corresponding to the internal structures that perl uses to run your program. When the generated C source is compiled and run, it cuts out the time which perl would have taken to load and parse your program into its internal semi-compiled form. That means that compiling with this backend will not help improve the runtime execution
17:32 inket speed of your program but may improve the
17:32 inket start-up time. Depending on the environment in which your program runs this may be either a help or a hindrance.
17:32 inket look at that
17:32 inket C compiles
17:32 inket perl doesnt
17:32 inket it transforms perl into C
17:32 tadzik joined #perl6
17:32 inket and generally as a rule, I've never seen a signal one of these type of applications do anything worth a dick
17:32 diakopter that's compilation.
17:32 tadzik ¡Hola!
17:33 inket any sort of complex coding breaks things
17:33 pmichaud patrickas: ah, that's a good point.  I suppose one can argue that   the correct answer should be 64, 32, 16 then
17:33 inket nd you lose ALL your language nuances
17:33 Ross left #perl6
17:33 pmichaud since 8 is the first Rat produced in the sequence
17:33 inket turning perl source into c source is compilation
17:33 diakopter inket: your OT-ness and rudeness is about to get you kb'd
17:33 tadzik what transforms Perl into C?
17:33 patrickas pmichaud: that's what I got when working on the bug, but I wanted to get a second opinion about
17:33 inket this particular thing diakopter showed me
17:33 diakopter tadzik: reini's B::C
17:33 x3nU diakopter: c'mon cursing isn't anything inacceptable imo
17:33 inket its not compilation
17:34 huf_ inket: so what kind of arcane magic *is* compilation then?
17:34 pmichaud patrickas: I haven't seen anything that says that   Int / Int   should return Int if there's no loss of precision
17:34 patrickas pmichaud: it does seem surprising at first! But I am not sure it is the sequence's Job to figure out the real Rats from the "Fake" ones
17:34 diakopter pmichaud++ patrickas++ staying on topic
17:34 pmichaud I'm sure it's not the sequence's Job
17:35 pmichaud I suspect the real question is whether    16 * (0.5)   produces an Int or a Rat
17:35 inket huf_: compilation is the process of taking source code (C, what ever) and turning it into binary instructions designed to talk to the Operating system and interact with the user
17:35 pmichaud and I think I'm likely to vote for Rat.
17:35 inket After which a new file is created that can be ran
17:35 huf_ inket: in that case, gcc isnt a compiler.
17:35 patrickas pmichaud: ok so I'll fix it to return 32,16,8 and call it THAT bug resolved .. then we can think about 16/2 :-)
17:35 inket without requiring a compiler again
17:35 pmichaud patrickas: note that it's not really 16 / 2, though.
17:35 inket huf_: you're right, gcc is a frontend
17:36 pmichaud the factor used in   64, 32, 16   is  0.5
17:36 pmichaud (a Rat)
17:36 wamba joined #perl6
17:36 tadzik inket: what's your point anyway?
17:36 huf_ inket: so the only compiler in the entire gcc heap is the assembler?
17:36 patrickas pmichaud: right
17:36 risou left #perl6
17:36 pmichaud and we get the next element by multiplying by a Rat, not by dividing by an Int
17:37 inket huf_: i said compilation is the *process* of turning source into object code
17:37 inket however many programs that takes.
17:37 * diakopter facepalm
17:37 huf_ inket: so how is an interpreter not a compiler?
17:37 huf_ inket: it's just another process for turning source into machine code, however many programs that takes
17:37 inket and diakopter your B::C application is just taking one source code, and translating it into another source code
17:38 tadzik hey hey! I'm not here for long, and although I read the backlog, I still don't now: what's the point?
17:38 patrickas pmichaud: thanks
17:38 inket an interpreter doesn't leave you with object code
17:38 inket it interprets it and reacts to the code as its processed
17:38 huf_ inket: btw, machine code is just another form of source code too, nothing special
17:38 flussence ew
17:38 flussence line-by-line execution is nasty.
17:38 flussence not even php does that any more.
17:38 huf_ inket: almost no interpreter does that
17:38 inket an interpreter is object code that reads an interpreted language and executes its own object code to produce the resultant behavior
17:39 flussence so?
17:39 inket your code is never directly translated to object code
17:39 flussence and?
17:39 diakopter what is "your"
17:39 huf_ inket: yes, so a compiler is an iterpreter and an interpreter is a compiler. your definitions make no sense.
17:40 inket an interpreter performs instructions, a compiler transforms source code into object code.
17:40 flussence I'm with tadzik here, I don't see the point of language-lawyering over the definitions of scripting/programming like this.
17:40 huf_ flussence: precisely.
17:40 tadzik flussence++
17:40 * diakopter goes to look at irc.perl6.org to find inket's original beef/question
17:40 inket diakopter: we've come a long way since then
17:40 tadzik I sense a little "Scripting is for stupid and compiled is for pr0", but it might be something with my nose today
17:40 inket tryig to define compilation here
17:41 diakopter inket: this isn't the channel for that.
17:41 huf_ inket: compilation is the act of transforming source code into another language (be that machine code or VM code or anything)
17:44 inket tats what wikipedia says
17:45 tadzik So Perl is compiled! At last, let's drink, or do something productive
17:45 huf_ i'm hacking php atm, this flame is by definition more productive than that ;)
17:45 inket huf_: compiling is source -> object
17:45 inket thats about it
17:45 wamba left #perl6
17:45 huf_ inket: you're confusing things again, stop with the rainclouds!
17:45 diakopter inket: I should clarify. this isn't the channel for "defining terms", *per se*. let's talk about your goals/ideas here.
17:45 huf_ paint smiley faces.
17:46 inket diakopter: just a regular discussion man
17:46 inket gees
17:46 inket huf_: what part am i confusing here
17:47 huf_ *sigh* nothing, nothing, i'll keep quiet
17:47 diakopter inket: my point is, why do you need to define these things?
17:47 flussence everything that runs on an x86 cpu from the last decade is compiled into micro-ops on the processor anyway, so that answers that :)
17:47 huf_ flussence: \o/
17:47 inket because i like to  know the right stuff
17:47 diakopter why?
17:47 inket learning?
17:47 diakopter I can play sophistry too
17:48 diakopter huf_++ hacking php atm :)
17:49 tadzik yuk
17:49 cotto left #perl6
17:49 huf_ diakopter: i dont think i deserve a ++ for that... :(
17:49 huf_ huf_--
17:49 diakopter well, at least for mentioning it; heh
17:49 huf_ :)
17:49 inket cant have a discussion here ?
17:49 inket ok
17:50 huf_ this was a discussion?
17:50 inket its about basic definitions in computing
17:50 flussence trolling is generally frowned upon
17:50 inket how is it trolling? becuase i have a difference of opinion?
17:50 diakopter I can't claim to have never been too combative/defensive here, but the combativeness/defensiveness level was too high.
17:51 diakopter just type more slowly and use fewer newlines between others' replies
17:51 diakopter that helps a lot
17:52 justatheory joined #perl6
17:52 tadzik diakopter: you have an extra newline before 'that help a lot' :)
17:52 diakopter :D
17:53 diakopter inket: can we talk about the definition I proposed?
17:53 inket which is
17:53 inket for compilation/
17:53 diakopter you can take a look at http://irc.perl6.org/ for a good view of the irclog
17:53 diakopter no, for scripting language
17:54 inket for scripting language to some extent i agree with what you said
17:54 tadzik script is what you give an actor
17:54 inket but for compilation i dont get how
17:54 * tadzik hides
17:54 inket perl just doesnt compile
17:54 tadzik why not?
17:54 takadonet perl is alive!
17:54 tadzik Compilation is transforming one code into another, isn't it?
17:54 tadzik Either C code to machine code, or Perl 6 code to Parrot/.NET code, nay?
17:54 meppl left #perl6
17:54 huf_ inket: you're hung up on this "object code" thing, as if that were some well-defined state, but it isnt. object code can be further transformed, interpreted, it's just code.
17:54 diakopter inket: explain "i dont get how perl just doesnt compile"
17:55 diakopter which perl?  perl5 or perl6 implementations?
17:56 inket the ewxample you showed me
17:56 inket uber raiki orsomething
17:56 inket cant possibly call that compiling
17:56 tadzik :)
17:57 diakopter inket: write a compiler or two, then you'll understand.
17:57 inket i know its hard
17:58 inket but im saying that part where c transforms it into perl is not compiling
17:58 inket c compiles it
17:58 araujo joined #perl6
17:58 huf_ inket: c is a language, it doesnt compile anything, it doesnt even run
17:59 tadzik I disagree. If you compile Java to JVM bytecode, it later gets compiled from bytecode to machine code, with JIT
17:59 diakopter you need to llllleeeettttt ggggooooo of your magical hangup on the word 'compiler'
17:59 tadzik so JVM compiles it, yet javac compiles it too, nay?
17:59 huf_ but this really is extremely futile, even by my irc flame standards ;)
17:59 huf_ (still better'n php tho :)))
18:00 bluescreen joined #perl6
18:00 diakopter inket: "that part where c transforms it into perl is not compiling" makes zero sense
18:01 diakopter even accounting for every braino/typo I can imagine
18:01 inket diakopter: I said that B::C app you showed me was a perl to c transformers
18:01 inket not compiling
18:01 diakopter yes.  it is.
18:02 huf_ inket: you are using a nonexistent definition of compiler. stop now.
18:02 diakopter and ALL of us are trying to tell you that "tranform" is another word for "compile"
18:02 pyrimidine joined #perl6
18:03 inket diakopter: i like perl and all but except for a few hacks, perl doesn't leave you with a compiled object
18:03 inket which means its not compiled
18:04 inket I have perl source, I run it through perl, it does stuff, I'm left with perl and my perl source.
18:04 inket now with c
18:04 inket I have C source, I run it through gcc, the code in the source file does NOT run, I end up with my source AND an object file
18:04 diakopter inket: You still haven't answered my question about which version of Perl you're asking about.
18:04 PerlJam inket: i can do exactly the same thing (as perl) with C
18:05 huf_ PerlJam: c isnt compiled? shit.
18:05 inket PerlJam: you serious? can you show an example
18:05 inket diakopter: any
18:05 MayDaniel left #perl6
18:05 inket perl5/6
18:05 jferrero left #perl6
18:06 PerlJam inket: of course I'm serious
18:06 inket ok i'd like to see an example if you dont mind
18:06 diakopter inket: just because there is no compilation result left on disk, doesn't mean no internal transformation/compilation took place.
18:07 PerlJam inket: Are *you* serious?  You want me to show you a C interpreter?
18:07 flussence llvm.
18:07 diakopter inket: most JIT compilers emit object code to memory only
18:07 PerlJam as if such a thing could never exist
18:08 diakopter inket: you're being too demanding.  if you want to know whether something exists, research it yourself.
18:08 inket llvm  is a C interpreter?
18:08 inket im not demanding
18:08 inket PerlJam said he could
18:08 inket i thought he'd have an example. thats all
18:09 tadzik tcc -run
18:09 flussence right, and you're excessively pressing that point along with several others. That's called being demanding.
18:09 inket whats the example?
18:09 inket llvm?
18:09 flussence and an hour later, you still haven't answered the most important question
18:09 flussence who cares?
18:10 tadzik tcc -run is an example
18:10 moritz_ (llvm is exactly what the name says -  a low-level virtual machine. It' not a C compiler or interpreter)
18:10 moritz_ oh hai
18:11 jnthn o/ moritz_
18:11 * moritz_ tilbake fra norsk course
18:11 jnthn ...that looks like rong Swedish. :P
18:11 flussence (if I'd said "llc" nobody would know what I meant)
18:11 MayDaniel joined #perl6
18:11 tadzik oh, jnthn
18:11 flussence oh that reminds me
18:11 diakopter (but it's a "VM" in the sense most people think, since it doesn't have a garbage collector).  but the "llvm project" does include the prod-quality "clang" c compiler
18:12 tadzik jnthn: you think it'd be possible to generate Perl 6 function prototypes for Zavolaj by examining the .h files?
18:12 diakopter it's *not* a "VM", I meant
18:12 flussence jnthn: the readme for zavolaj mentions something about parrot and hard-coded tables, does that mean I can't just use it with any random lib?
18:12 moritz_ diakopter: that's the "ll" part :-)
18:13 jnthn tadzik: Sounds like an interesting tool.
18:13 patrickas o/ moritz_ jnthn
18:13 tadzik inket: C script supported : just add '#!/usr/local/bin/tcc -run' at the first line of your C source, and execute it directly from the command line. # http://bellard.org/tcc/
18:13 PerlJam tadzik, jnthn: I thought someone in parrot-land was working on such a tool even.
18:13 patrickas It's been a while! :-)
18:13 moritz_ \o patrickas
18:13 jnthn tadzik: I think somebody wrote a Perl 6 grammar for C source files already.
18:13 tadzik jnthn: I though like a NativeCall submodule, like `importlib "libgtk2.0"`
18:13 jnthn tadzik: So you needn't even re-invent that wheel.
18:14 jnthn tadzik: Could be interesting.
18:14 tadzik I'm quite short of internet access these days
18:14 jnthn flussence: I think it depends on how your Parrot is built. Though it may also depend on if the branch that removes the limitation got merged yet too...
18:15 flussence oh, that sounds promising :)
18:15 tadzik but but, how to export subroutines in runtime? I mean, creating them first, some sort of eval-ution?
18:15 jnthn Just checking on #parrot
18:16 MayDaniel left #perl6
18:16 inket tadzik: yeah tried it
18:16 jnthn 20:27 <@dukeleto> jnthn: it is one failing test away from being mergable
18:16 flussence yay
18:16 jnthn flussence: ^^ That's status of the branch in question. :-)
18:17 jnthn tadzik: Ah, I thought you were thinking more of a tool that generates the .pm file from a .h file at first. Hm
18:17 tadzik jnthn: well, that's the easy way, but the better way would be better I think
18:17 inket anyhow thanks for the discussion
18:18 inket i didnt mean to cause a flame war or anything
18:18 jnthn tadzik: I think you may - if you can do it early enough - be able to put stuff into the Module::Name::EXPORT::DEFAULT namespace...
18:19 jnthn I'm not quite sure how to tie it all together off ahnd without thinking about it a bit more.
18:19 standz left #perl6
18:20 standz joined #perl6
18:21 wormwood joined #perl6
18:22 am0c left #perl6
18:23 Schwern joined #perl6
18:26 mberends joined #perl6
18:26 dalek roast: 1b6274b | patrickas++ | S03-sequence/basic.t:
18:26 dalek roast: added test for RT #78324
18:26 dalek roast: review: http://github.com/perl6/roast/commit/1b​6274b5f513474b6e67aa0b00b0b9a383f97c4e
18:28 moritz_ patrickas: ah, that's the bug I hoped to get you working on :-)
18:29 patrickas moritz_: I got some free time today so I fixed it :-)
18:29 pmichaud rakudo:  (32,16,8 ...^ Rat).perl.say
18:30 p6eval rakudo d35769: OUTPUT«(timeout)»
18:30 pmichaud huh
18:30 moritz_ patrickas: then push -- you have commit access
18:30 pmichaud I'm surprised rakudo doesn't have that right already
18:30 patrickas pmichaud: it was a stupid thing I had done in the new code , was really trivial to fix
18:30 pmichaud rakudo:  (32,16,8 ...^ 0).perl.say
18:30 p6eval rakudo d35769: OUTPUT«(32, 16, 8, 4/1, 2/1, 1/1, 1/2, 1/4, 1/8, 1/16, 1/32, 1/64, 1/128, 1/256, 1/512, 1/1024, 1/2048, 1/4096, 1/8192, 1/16384, 1/32768, 1/65536, 1/131072, 1/262144, 1/524288, 1/1048576, 1/2097152, 1/4194304, 1/8388608, 1/16777216, 1/33554432, 1/67108864, 1/134217728, 1/268435456,
18:30 p6eval ..1/5368…
18:31 pmichaud rakudo:  (32,16,8 ...^ 1).perl.say
18:31 p6eval rakudo d35769: OUTPUT«(32, 16, 8, 4/1, 2/1)␤»
18:31 flussence wha
18:31 moritz_ note that it produces rats immediately
18:31 patrickas when the limit is not .defined the sequence would be infite
18:31 pmichaud ....?
18:31 flussence shouldn't it omit the /1 ?
18:31 pmichaud patrickas: that statement doesn't make sense to me
18:32 patrickas in the code I wrote i had a line that says return infite-series unless $limit.defined :-o
18:33 pmichaud oh, that would be wrong
18:33 pmichaud so, I understand now.
18:33 patrickas which I changed to return infite-series if $rhs ~~ Whatever
18:34 timbunce left #perl6
18:35 aloha left #perl6
18:36 patrickas the whole line is not really necessary but it seemed like a good idea to avoid the check if the sequence is finished when we know it is infinite.
18:36 pmichaud it'll also be useful when we can start to tag infinite lists
18:36 pmichaud afk for a while
18:38 patrickas ok tests done ... I'll push to the rakudo repo directly for the first time ever, if all hell breaks loose you know who to blame
18:39 patrickas actually, how do I fake push to master, just to make sure only the right patch will be pushed before I do it for real ?
18:40 PerlJam patrickas: why not just: git diff origin/master master
18:40 timbunce joined #perl6
18:40 moritz_ patrickas: there's no fake push... one way is to write down the patch SHA1, then   git reset --hard origin/master; git cherry-pick $sha1; git push
18:41 Guest84593 left #perl6
18:41 patrickas Thanks moritz_ and PerlJam ... indeed there were a bagillion different things between the branch I am on and origin/master
18:42 * PerlJam works on the rakudo release
18:43 patrickas PerlJam: maybe it is wise for me to wait till after the release to push
18:43 PerlJam patrickas: what are you pushing?
18:44 patrickas a minor patch for sequence but I am a total git noob and this is the first time I am pushing without the training wheel (ie moritz_ applying my patches)
18:45 flussence patrickas: check out a clean branch from master, then git cherry-pick that one patch
18:45 aloha joined #perl6
18:46 patrickas flussence that's what I am doing (in theory) but that would be my first cherry-pick ever and I am bound to screw up somehow ... :-)
18:46 PerlJam patrickas: just do it.  There's enough git-fu around that we can clean up if you make a mess.
18:48 Guest82379 joined #perl6
18:49 PerlJam you know ... the most onerous thing about rakudo releases for me is that I have to wait for the spectests   :)
18:49 PerlJam (I'm not patient)
18:50 saaki left #perl6
18:51 mavrc joined #perl6
18:52 inket left #perl6
18:53 patrickas Ok I pushed, thanks for being patient with me :-)
18:53 dalek rakudo: 7a2990d | patrickas++ | src/core/operators.pm:
18:53 dalek rakudo: Patch for sequence not terminating when the rhs is not .defined
18:53 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/7​a2990d9a84774eea84c2b22825d128ce3c51376
18:54 moritz_ patrickas++ # looks good to me
18:55 patrickas First thing I did is check the history in github to make sure only one line was changed :-)
18:56 patrickas btw is it better to push to rakudo then add the tests when I have a patch ready to avoid dalek spam? or should I always add fudged tests, push patch, unfudge the test ?
18:57 saaki joined #perl6
18:58 moritz_ patrickas: whatever you like best - for us dalek doesn't spam, but inform us :-)
18:58 PerlJam patrickas: "always" is a bit too restrictive.  If you have the tests first, go ahead and add them (no need to wait)
18:58 dalek roast: ef896c7 | pmichaud++ | S03-sequence/basic.t:
18:58 dalek roast: Update test message in basic.t to be more accurate.
18:58 dalek roast: review: http://github.com/perl6/roast/commit/ef​896c7fb7cc361d5e5d79b2c1faab32ff4a4fed
18:58 bluescreen left #perl6
18:58 PerlJam patrickas: but if the tests aren't quite there but the patch is, go ahead and add it too :)
18:59 PerlJam patrickas: share early, share often.
18:59 patrickas :-)
19:00 moritz_ aye; if you get sidetracked and don't get around to fixing, at least everybody has the tests
19:05 justatheory left #perl6
19:05 Entonian left #perl6
19:08 dalek roast: ff0784a | patrickas++ | S03-sequence/basic.t:
19:08 dalek roast: unfudge test
19:08 dalek roast: review: http://github.com/perl6/roast/commit/ff​0784ad6ad066bf376c7657e582f36e3c703cb3
19:08 dalek roast: c1b2547 | patrickas++ | S03-sequence/basic.t:
19:08 dalek roast: unfudge test
19:08 dalek roast: review: http://github.com/perl6/roast/commit/c1​b2547a5c30c3f8795dff4e234066e4469b75ca
19:09 patrickas actually patrickas-- for the first one :-)
19:10 PerlJam Draft release announcement:  https://gist.github.com/907a574db2b58509c51b
19:11 PerlJam Anyone see anything missing or otherwise wonky?
19:11 PerlJam patrickas: before pushing, I probably would have combined those two commits.
19:12 patrickas PerlJam how can I do that?
19:13 bluescreen joined #perl6
19:13 PerlJam patrickas: git rebase -i <SHA1>   # then do what it says in the comments.
19:13 PerlJam patrickas: the <SHA1> should be one prior to the ones you want to combine
19:14 moritz_ PerlJam: looks fine to me.. though I don't know off-hand what the "numerous documentation improvements" are
19:14 PerlJam patrickas: and be sure to not rebase any commits that have already been pushed as that changes history
19:15 PerlJam moritz_: looking through the log inse the last release I see lost of "fix typo" and such
19:15 PerlJam s/lost/lots/
19:15 PerlJam s/inse/since/  # dunno /what/ happened there!
19:16 flussence .oO( fixing typos )
19:16 moritz_ PerlJam: ok
19:17 patrickas PerlJam ... thanks, I'll try that next time. Now that you mention it I am sure moritz_ arealy tried to squeeze that through my thick skull once but oviously failed :-)
19:17 PerlJam although now that I look at it again, it doesn't seem like "lots" this time.
19:17 moritz_ PerlJam: maybe mention improved error messages instead
19:17 moritz_ rakudo: (1, 2) >>+<< (1, 2, 3)
19:17 p6eval rakudo d35769: OUTPUT«Sorry, sides are of uneven length and not dwimmy.␤  in 'hyper' at line 184:CORE.setting␤  in main program body at line 1␤»
19:17 PerlJam and warnings
19:18 PerlJam I'll do that.
19:18 moritz_ huh, that's still the old one
19:18 takadonet rakudo: (1, 2,3) >>+=<< (1, 2, 3)
19:18 moritz_ s/error/diagnostic/ :-)
19:18 p6eval rakudo d35769: OUTPUT«Cannot modify readonly value␤  in '&infix:<=>' at line 1␤  in <anon> at line 211:CORE.setting␤  in 'hyper' at line 1␤  in main program body at line 1␤»
19:18 takadonet rakudo: 2 [+=] (1,2,3)
19:18 p6eval rakudo d35769: OUTPUT«Cannot modify readonly value␤  in '&infix:<=>' at line 1␤  in main program body at line 22:/tmp/Hn7Wf217sX␤»
19:18 takadonet rakudo:  [+] (1,2,3)
19:18 p6eval rakudo d35769:  ( no output )
19:19 takadonet rakudo: say [+] (1,2,3)
19:19 p6eval rakudo d35769: OUTPUT«6␤»
19:26 dalek rakudo: 33bdcee | perlpilot++ | docs/ (2 files):
19:26 dalek rakudo: release preparation
19:26 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/3​3bdceeea4b00539654ca525e23b88079ae20bef
19:27 orafu left #perl6
19:28 orafu joined #perl6
19:36 _buno_ joined #perl6
19:40 moritz_ huh. My ssh agent just died.
19:40 _buno_ left #perl6
19:41 sorear good * #perl6
19:41 moritz_ good * sorear
19:42 moritz_ can I configure git somehow to use separate push and fetch URLs?
19:42 DJAB joined #perl6
19:42 flussence yes
19:42 moritz_ how?
19:43 DJAB whats up
19:43 tadzik #34 :)
19:43 flussence moritz_: one sec, I know it's possible, just can't find where I've done it...
19:44 moritz_ ah, separate pushurl in .git/config
19:44 moritz_ git remote set-url --push origin git@github.com:perl6/roast.git
19:45 flussence yeah, figures it'd be something obvious yet I don't remember what :)
19:45 DJAB left #perl6
19:47 xabbu42 left #perl6
19:49 moritz_ very handy for t/spec/
19:51 envi left #perl6
19:51 dukeleto moritz++ # i didn't know about that
19:51 patrickas moritz_ how do you use it with t/spec/ ?
19:53 moritz_ [remote "origin"] fetch = +refs/heads/*:refs/remotes/origin/* url = git://github.com/perl6/roast.git pushurl = git@github.com:perl6/roast.git
19:53 patrickas I mean why is it very handy for t/spec ?
19:53 wamba joined #perl6
19:53 flussence fetch over git, push over ssh maybe
19:53 moritz_ patrickas: because 'make spectest' automatically pulls. If the ssh key is not available, 'make spectest' hangs
19:54 flussence the git one will be slightly faster due to not encrypting stuff, too
19:54 patrickas moritz_ oh ok
19:54 sorear moritz_: I am starting to wonder if and when I should do a release announcment to p6c
19:54 moritz_ and it turns out ssh-agent sometimes dies if it runs out of disc space
19:54 moritz_ sorear: as soon as you do a release
19:57 tadzik niecza release?
19:57 justatheory joined #perl6
19:57 wamba left #perl6
19:58 moritz_ sorear: I'm sure it would be a good thing, raising awareness, giving people something to link to apart from just the repo
19:58 sorear I'm also wondering if an when there should be a release. :)
20:01 PerlJam sorear: Do you want other people playing with niecza?
20:05 patrickas evening &
20:05 patrickas left #perl6
20:06 * PerlJam continues to wait for the spec tests to run.
20:06 PerlJam s/run/finish running/
20:07 Ross joined #perl6
20:08 moritz_ rakudo: say { a => 1, b => 2 }.pick.perl
20:08 p6eval rakudo d35769: OUTPUT«"b"␤»
20:11 plainhao left #perl6
20:12 ggoebel left #perl6
20:15 Entonian joined #perl6
20:17 PerlJam finally!
20:17 PerlJam seems it failed only one test #8 in t/spec/S05-transliteration/trans.rakudo
20:24 MayDaniel joined #perl6
20:24 MayDaniel left #perl6
20:24 MayDaniel joined #perl6
20:26 zostay left #perl6
20:26 leprevost joined #perl6
20:27 zostay joined #perl6
20:29 justatheory left #perl6
20:31 justatheory joined #perl6
20:32 wormwood_ joined #perl6
20:34 wormwood left #perl6
20:35 ggoebel joined #perl6
20:36 zostay left #perl6
20:37 zostay joined #perl6
20:39 Guest82379 left #perl6
20:42 bluescreen left #perl6
20:43 sftp left #perl6
20:44 breinbaa1 left #perl6
20:44 cggoebel joined #perl6
20:45 moritz_ yep; that must be fudged out
20:45 ggoebel left #perl6
20:45 breinbaas joined #perl6
20:47 sorear PerlJam: Do I want other people playing with niecza...?  Tough one.
20:47 baux left #perl6
20:48 dalek niecza/master: 876edf3 | sorear++ | src/ (2 files):
20:48 dalek niecza/master: $/ and $! are contextuals
20:48 dalek niecza/master: review: http://github.com/sorear/niecza/commit/8​76edf3a9a7cece4d510183ae59797a2c750a127
20:48 dalek niecza/master: 94651cd | sorear++ | / (2 files):
20:48 dalek niecza/master: Fix try to put error in $!, not inline
20:48 dalek niecza/master: review: http://github.com/sorear/niecza/commit/9​4651cd8ecad302f09e7ce662921a428e452e32b
20:48 dalek niecza/master: 4b358b4 | sorear++ | / (5 files):
20:48 dalek niecza/master: Implement temp $*foo
20:48 dalek niecza/master: review: http://github.com/sorear/niecza/commit/4​b358b4f94abfaf2290b47b39924e9a29e1cfc9c
20:48 dalek niecza/master: 7a7e3f2 | sorear++ | / (2 files):
20:48 dalek niecza/master: Fix <foo> <foo> capturing
20:48 dalek niecza/master: review: http://github.com/sorear/niecza/commit/7​a7e3f2810bad31ade3f0981f785447b1e5803f7
20:48 zby_home left #perl6
20:49 orafu left #perl6
20:50 orafu joined #perl6
20:51 jaldhar left #perl6
20:52 pmurias joined #perl6
20:52 pmurias hi
20:52 sorear hello pmurias
20:52 sjohnson yo dudes
20:56 timbunce_ joined #perl6
20:58 sorear ugh.  FindBin has failed me.
20:58 timbunce left #perl6
20:58 timbunce_ is now known as timbunce
20:58 araujo left #perl6
20:59 bluescreen joined #perl6
20:59 moritz_ are symlinks involved?
20:59 sjohnson FailBin
21:00 jaldhar joined #perl6
21:01 cggoebel left #perl6
21:01 justatheory left #perl6
21:01 kjeldahl left #perl6
21:01 justatheory joined #perl6
21:03 moritz_ rakudo: my %a; %a<b><c><d> = 5; say %a.perl
21:03 p6eval rakudo d35769: OUTPUT«{"b" => {"c" => {"d" => 5}}}␤»
21:04 moritz_ rakudo: my %a; %a<b><c><d>  += 5; say %a.perl
21:04 p6eval rakudo d35769: OUTPUT«{"b" => {"c" => {"d" => 5}}}␤»
21:05 sorear moritz_: could you try building std again?  it no longer looks for an installed STD.pmc while rebuilding
21:06 * moritz_ tries
21:06 sorear my goal is to be able to demonstrate STD.parse("2 + 2", :setting<NULL>) at next #phasers
21:07 moritz_ looks much better (STD build)
21:08 leprevost left #perl6
21:08 moritz_ std: say "success!"
21:08 p6eval std :  ( no output )
21:08 moritz_ uhm, I need "make snap"
21:09 moritz_ std: say "success!"
21:09 p6eval std a194beb: OUTPUT«sh: /opt/perl-5.12.1/bin/perl: not found␤»
21:10 p6eval left #perl6
21:10 p6eval joined #perl6
21:10 moritz_ std: 1
21:10 dalek evalbot: 568324e | moritz++ | evalbot.pl:
21:10 dalek evalbot: [std] use system perl
21:10 dalek evalbot: review: http://github.com/perl6/evalbot/commit/5​68324e9cf7aa23d289fe6ed84b452ee85bf4218
21:11 p6eval std a194beb: OUTPUT«ok 00:01 117m␤»
21:12 timbunce left #perl6
21:13 whiteknight joined #perl6
21:15 Ross left #perl6
21:17 timbunce joined #perl6
21:17 bluescreen left #perl6
21:17 pyrimidine left #perl6
21:33 masak joined #perl6
21:33 masak ahoy, #perl6!
21:33 jnthn yayitsmasak!
21:34 masak mam lagi...
21:34 sorear masak!
21:35 sjohnson hi masak + jnthn + sorear
21:36 orafu left #perl6
21:37 masonkramer left #perl6
21:38 pmurias sorear: how's niecza doing?
21:38 _twitch joined #perl6
21:39 sorear Very well.
21:39 pmurias that's good
21:39 sorear I think I'll be able to demonstrate STD.parse("2 + 2", :setting<NULL>) at #phasers next week
21:39 masak ++sorear
21:39 pmurias nice
21:40 MayDaniel left #perl6
21:40 _kaare left #perl6
21:41 mavrc left #perl6
21:47 spinclad joined #perl6
21:53 Schwern left #perl6
21:55 Schwern joined #perl6
21:57 araujo joined #perl6
21:58 pmurias left #perl6
21:59 xabbu42 joined #perl6
22:02 dalek 6model: e47a5fb | jnthn++ | dotnet/compiler/DNST.pm:
22:02 dalek 6model: [dotnet] Add DNST::TryCatch and DNST::Throw nodes.
22:02 dalek 6model: review: http://github.com/jnthn/6model/commit/e​47a5fbaddd0b1402e4fdee41ed3b606eb29edf5
22:02 dalek 6model: c6bbf47 | jnthn++ | dotnet/compiler/DNST2CSharp.pm:
22:02 dalek 6model: [dotnet] Compile DNST::Throw and DNST::TryCatch.
22:02 dalek 6model: review: http://github.com/jnthn/6model/commit/c​6bbf47b5c4db5c3fb722afb5a5b6ce8f26a5334
22:02 dalek 6model: 670415d | jnthn++ | dotnet/ (3 files):
22:02 dalek 6model: [dotnet] Start to stub in the bits we'll need to implement .leave.
22:02 dalek 6model: review: http://github.com/jnthn/6model/commit/6​70415d2645a8817c43af4fbb3a3e15f219938a4
22:02 dalek 6model: 47c868a | jnthn++ | dotnet/compiler/ (2 files):
22:02 dalek 6model: [dotnet] Min codegen e kass - make things a bit less noisy.
22:02 dalek 6model: review: http://github.com/jnthn/6model/commit/4​7c868ada6acff2fb912be70396aa2e5163e1770
22:03 masak left #perl6
22:04 xabbu42 left #perl6
22:07 dalek 6model: c788c81 | jnthn++ | dotnet/runtime/Runtime/Ops.cs:
22:07 dalek 6model: [dotnet] Add a leave_block op. This is what we'll do when needing to do some stack unwinding.
22:07 dalek 6model: review: http://github.com/jnthn/6model/commit/c​788c8100154111491deb8f9456ae76921892961
22:07 dalek 6model: 3c156f0 | jnthn++ | common/NQP/NQPSetting.pm:
22:07 dalek 6model: [common] Add a NQPCode with a leave method.
22:07 dalek 6model: review: http://github.com/jnthn/6model/commit/3​c156f0c791269bac3dbebfa764be556f5249eb1
22:07 dalek 6model: bd57e32 | jnthn++ | dotnet/runtime/Init.cs:
22:07 dalek 6model: [dotnet] Use NQPCode instead of the hack LLCode.
22:07 dalek 6model: review: http://github.com/jnthn/6model/commit/b​d57e329bc20d7e892d87ba32fd38167290ab327
22:07 dalek 6model: 4281db6 | jnthn++ | dotnet/compiler/PAST2DNSTCompiler.pm:
22:07 dalek 6model: [dotnet] Fix a silly thinko and...we have &building.leave($elvis) working. \o/
22:07 dalek 6model: review: http://github.com/jnthn/6model/commit/4​281db6b9dda04d1a2e9384b9988294be8f7291e
22:08 masak joined #perl6
22:09 _twitch left #perl6
22:10 _twitch joined #perl6
22:11 masonkramer joined #perl6
22:12 timbunce left #perl6
22:15 PerlJam Is there some more direct correspondence between the spec tests repo and the rakudo repo?
22:15 sorear no
22:17 dalek rakudo: acce294 | perlpilot++ | VERSION:
22:17 dalek rakudo: bump VERSION
22:17 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/a​cce29405c996b40f4ca5bc49f48760e083bea98
22:18 sorear PerlJam: there is a tension in my mind between "too many tasks, too little time" and "this is MY sandbox!"
22:26 nadim_ left #perl6
22:29 dalek 6model: 9445e1d | jnthn++ | dotnet/runtime/Runtime/Exceptions​/LeaveStackUnwinderException.cs:
22:29 dalek 6model: [dotnet] I forgot to finish the
22:29 dalek 6model: review: http://github.com/jnthn/6model/commit/9​445e1d0500b8d62c76f67c3158ed9098df73fcd
22:35 Schwern left #perl6
22:36 PerlJam sorear: there's still a chance for niecza to become the defacto Perl 6 compiler  ;)
22:39 dalek roast: f8d9c6f | perlpilot++ | S05-transliteration/trans.t:
22:39 dalek roast: fudge trans test
22:39 dalek roast: review: http://github.com/perl6/roast/commit/f8​d9c6fd26ed5fbbfa6213ee8aca9d8a6f0fcfe3
22:39 _twitch left #perl6
22:43 zostay left #perl6
22:44 masak` joined #perl6
22:45 masak left #perl6
22:46 masak` is now known as masak
22:47 zostay joined #perl6
22:47 _twitch joined #perl6
22:50 icwiener left #perl6
22:51 justatheory left #perl6
22:51 justatheory joined #perl6
22:56 jnthn sleep &
22:57 masak not a half-bad idea, actually.
22:57 colomon sleep sounds heavenly, but I've got to get my boy to bed first.
22:58 colomon and I'll be lucky if he's ready to settle down an hour from now.
23:00 ashleydev left #perl6
23:01 * moritz_ finds that JSON is broken :(
23:01 aloha left #perl6
23:02 masak :/
23:03 moritz_ say from-json('{ "a": 1 }').perl;
23:03 moritz_ "a" => 1
23:03 cotto joined #perl6
23:06 alester left #perl6
23:07 moritz_ rakudo: class A { method TOP($/) { make { a => 1 } } }; grammar B { token TOP { ^. } }; say B.parse('x', :actions(A.new)).ast.perl
23:07 p6eval rakudo d35769: OUTPUT«{"a" => 1}␤»
23:07 * masak thinks that's right
23:08 moritz_ right
23:08 moritz_ it's just that the code in JSON::Tiny::Actions does basically the same, but loses the curlies somewhere along the way
23:08 masak <inket> huf_: compilation is the process of taking source code (C, what ever) and turning it into binary instructions designed to talk to the Operating system and interact with the user
23:09 masak I'm almost glad I didn't stick around for the conclusion of that discussion.
23:10 flussence I'm glad I wasn't paying attention :D
23:10 masak it's funny how some people enter the channel, start with a simple question, and half an hour later, they've only proven themselves to be wrong and untechable. luckily, they're few and far between.
23:10 masak s/tech/teach/
23:11 aloha joined #perl6
23:12 * moritz_ wonders if the disappearing hash wrapper is related to the .item fix in rakudo
23:16 masak looking forward to the bug report with the golfed code.
23:16 masak having backlogged inket's confused ramblings, I do believe that he meant no actual harm. he just hasn't had any good teachers so far. that happens.
23:17 masak I seriously hope I'll never be as hard-headed as he was in the presence of ten or so people trying to educate me.
23:18 PerlJam boy scout meeting with my son and then release!
23:18 * PerlJam bbl
23:18 moritz_ nope, that wasn't it
23:18 s1n joined #perl6
23:19 masak PerlJam: good luck with the release!
23:19 * masak heads bedwards, with the faint hope of sleeping
23:19 masak 'night.
23:19 masak left #perl6
23:21 stkowski left #perl6
23:31 moritz_ rakudo: class A { method TOP($/) { make ( a => 1 ).hash; say $/.perl } }; grammar B { token TOP { ^. } }; say B.parse('x', :actions(A.new)).ast.perl
23:31 p6eval rakudo d35769: OUTPUT«Match.new(␤ ast  => "a" => 1,␤ from => 0,␤ orig => "x",␤ to   => 1,␤)␤"a" => 1␤»
23:31 moritz_ rakudo: say (a => 1).hash.perl
23:31 p6eval rakudo d35769: OUTPUT«{"a" => 1}␤»
23:31 moritz_ rakudo: class A { method TOP($/) { make ( a => 1 ).hash } }; grammar B { token TOP { ^. } }; say B.parse('x', :actions(A.new)).ast.perl
23:31 p6eval rakudo d35769: OUTPUT«"a" => 1␤»
23:32 moritz_ rakudo: class A { method TOP($/) { make ( a => 1, b => 2 ).hash } }; grammar B { token TOP { ^. } }; say B.parse('x', :actions(A.new)).ast.perl
23:32 p6eval rakudo d35769: OUTPUT«("a" => 1, "b" => 2)␤»
23:32 moritz_ rakudo: class A { method TOP($/) { make ( a => 1, b => 2 ).hash.item } }; grammar B { token TOP { ^. } }; say B.parse('x', :actions(A.new)).ast.perl
23:32 p6eval rakudo d35769: OUTPUT«{"a" => 1, "b" => 2}␤»
23:33 stkowski joined #perl6
23:37 ggoebel joined #perl6
23:39 moritz_ rakudo: say ()>>.uc.flat.perl
23:39 p6eval rakudo d35769: OUTPUT«Incompatible versions of `perl6_ops' oplib. Found 2.8.0 but loaded 2.9.1␤»
23:40 ashleydev joined #perl6
23:40 * moritz_ hates it
23:42 cdarroch left #perl6
23:44 orafu joined #perl6
23:49 moritz_ rakudo: class A { method TOP($/) { make () } }; grammar B { token TOP { ^. } }; say B.parse('x', :actions(A.new)).ast.perl
23:49 p6eval rakudo : OUTPUT«Any␤»
23:50 moritz_ that's wrong too
23:50 Entonian left #perl6
23:51 moritz_ rakudo: sub f($x) {·say $x.perl }; f(())
23:51 p6eval rakudo : OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 22␤»
23:52 moritz_ rakudo: sub f($x) { say $x.perl }; f(())
23:52 p6eval rakudo : OUTPUT«()␤»
23:53 Entonian joined #perl6
23:55 zostay left #perl6
23:57 zostay joined #perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo