»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend! | feather will shut down permanently on 2015-03-31
Set by jnthn on 28 February 2015.
00:07 laouji joined 00:09 thou left, laouji_ joined, tardisx joined 00:10 laouji__ joined 00:11 tardisx left, tardisx joined 00:12 laouji left 00:13 laouji_ left 00:25 davido_ joined 00:38 ssqq joined 00:39 yqt left 00:40 ssqq left 00:45 ssqq joined 01:05 colomon left 01:08 colomon joined 01:10 agentzh_ joined 01:11 aborazmeh joined, aborazmeh left, aborazmeh joined 01:16 agentzh_ left 02:10 ilogger2 joined, ChanServ sets mode: +v ilogger2, agentzh_ joined 02:25 rmgk_ joined, rmgk_ is now known as rmgk 02:28 agentzh_ left 02:34 aborazmeh joined, aborazmeh left, aborazmeh joined 02:39 raiph joined 02:52 aborazmeh left 02:53 colomon joined 02:59 kurahaupo joined 03:21 raiph left 03:26 thou joined 03:39 agentzh_ joined 04:04 kaare__ joined 04:12 agentzh_ left 04:38 rvchangue joined 04:39 rvchangue left 04:52 rvchangue joined 04:54 agentzh_ joined 05:07 agentzh_ left 05:25 spider-mario joined 05:31 aborazmeh joined, aborazmeh left, aborazmeh joined 05:33 araujo joined, araujo left, araujo joined 05:45 thou left 05:48 ufobat joined 05:59 domidumont joined 06:02 domidumont left 06:03 domidumont joined 06:07 rba_ joined 06:08 laouji joined
dalek pan style="color: #395be5">perl6-examples: 68f6bb0 | paultcochrane++ | categories/cookbook/04arrays/04-0 (2 files):
[cookbook] indent code consistently
 06:08
pan style="color: #395be5">perl6-examples: b9b413d | paultcochrane++ | t/categories/cookbook/04arrays.t:
[cookbook] add tests for 04arrays examples
06:15 rba_ left 06:20 aborazmeh left 06:26 FROGGS joined 06:29 rba_ joined 06:32 rba_ left 06:34 kurahaupo left 06:39 quester joined 06:43 [Tux] joined 06:44 _mg_ joined 06:49 zakharyas joined 06:51 abraxxa joined 06:52 agentzh_ joined 06:56 RabidGravy joined 06:57 Ven joined
sjn arnsholt, jnthn, pmichaud: We have a room at Teknologihuset for some hacking, from 09:30 today 06:59
07:01 haidarian joined
FROGGS morning #perl6 07:05
ohh, I wish a lot of fun at Teknologihuset 07:06
07:08 telex joined
moritz sounds like a fun place to be 07:08
masak morning, #perl6 07:09
07:10 kurahaupo joined
moritz \o masak, FROGGS, sjn, * 07:11
07:13 poul joined 07:31 larion joined 07:32 lizmat joined 07:34 pdcawley joined 07:35 virtualsue joined, virtualsue left
lizmat so, I'm the only one having this slowdown?? 07:37
lizmat nukes install and starts from scratch
07:39 agentzh_ left
FROGGS lizmat: seems so... I have no slowdown 07:42
sjn lizmat: we have the room at teknologihuset today, btw :) 07:44
07:45 quester left
FROGGS sjn: is tadzik also at your place? 07:47
RabidGravy marnin' 07:49
tadzik FROGGS: I was not
and I
'm now aboard flytoget to OSL
masak flytoget! with free wifi! \o/ 07:51
tadzik yes :)
also the interfaceless ticket buying
1) swipe your card 2) get on train
what a time to be alive 07:52
FROGGS *g*
07:55 larion left
lizmat sjn: on our way now 07:57
07:57 lizmat left
Ven today's masak's autopun. twitter.com/kmett/status/596472641894129664 07:59
07:59 laouji_ joined
Ven also, o/ #perl6 07:59
08:02 eiro joined 08:03 laouji left
sjn pmichaud, jnthn, arnsholt: I'm at teknologihuset now, in the boardroom two stairs up. Espen (the host) will show you the way 08:06
08:09 bjz joined 08:10 lizmat joined, larion joined
lizmat after an install nuke, all back to normal :-) 08:11
08:13 espadrine_ joined 08:14 lizmat left 08:16 darutoko joined
FROGGS phew :o) 08:19
pmichaud good morning, #perl6
FROGGS morning pmichaud
masak morning, pmichaud!
arnsholt o/ 08:20
FROGGS hopes that pmichaud slept more than what FROGGS has heard on other conferences/workshops
pmichaud I did finally start to catch up on sleep last night, yes. 08:22
sjn is afraid he may have kept jnthn up for too long last night. 08:23
the beer was good though
masak this is the first time I see POD (in Perl 5) being written exclusively on lines prefixed '#pod ': metacpan.org/source/DAGOLDEN/Hash-...Ordered.pm 08:27
is that a new thing in Perl 5? or have I simply missed it for a very long time?
DrForr Maybe a release tool gone wrong? 08:28
masak oh! possible.
nope. api.metacpan.org/source/DAGOLDEN/H...Ordered.pm
or, at least not unless I misunderstand your question.
08:29 g4 joined, spider-mario left 08:31 silug joined, silug left
DrForr I was just thinking that he might have an editor that skips '#pod' marked lines and normally removes the #pod markers before publishing. 08:31
Which is admittedly a stretch.
masak well, weak evidence against that is the fact that the module renders properly :) 08:35
dalek kudo/nom: 92bbf5c | arnsholt++ | t/04-nativecall/04-pointers.c:
Fix C issue in NativeCall tests.

One of the pointer tests returned a pointer to astack allocated string outside of the function, which is not very safe. Replace it with a strdup()ed string that's freed on verification.		 08:36
masak metacpan.org/pod/Hash::Ordered
`perldoc perlpod` doesn't seem to mention this at all. 08:37
08:38 lizmat joined 08:46 silug joined 08:48 rindolf joined
FROGGS arnsholt++ 08:49
08:52 lizmat_ joined 08:53 lizmat left
dalek kudo/nom: 16847e6 | lizmat++ | src/core/CompUnit.pm:
Let CompUnit.precomp fail verbosely when needed
 08:54
08:54 lizmat joined 08:57 lizmat_ left 08:58 lizmat_ joined 09:01 virtualsue joined, lizmat left 09:06 lizmat_ is now known as lizmat
raydiak :q 09:07
erm...gnight #perl6
lizmat gnight raydiak 09:08
masak raydiak has quit the vim of daytime. 09:09
FROGGS gnight raydiak
09:10 Ven left
nwc10 good *, * 09:18
lizmat
.oO( Whatever :-)
nwc10 yeah, whatever.
masak m: say "$_." for <what ever> 09:19
camelia rakudo-moar 92bbf5: OUTPUT«what.␤ever.␤»
masak m: say Bool.pick ?? "{uc $_}." !! "$_." for <what ever>
camelia rakudo-moar 16847e: OUTPUT«WHAT.␤EVER.␤»
masak m: say Bool.pick ?? "{uc $_}." !! "$_." for <what ever>
camelia rakudo-moar 16847e: OUTPUT«WHAT.␤EVER.␤»
masak hm.
m: say Bool.pick ?? "{uc $_}." !! "$_." for <what ever>
camelia rakudo-moar 16847e: OUTPUT«what.␤EVER.␤»
masak oh, good. :)
09:24 bjz_ joined 09:27 bjz left 09:34 lizmat_ joined
RabidGravy The S29 says several times "The NameServices role in S16 covers most of these." but no such is mentioned or even eluded to in S16, is it just somewhere I failed to noticed it or just not specced at all? 09:34
masak alluded* :) 09:35
lizmat_ m: say ."{<uc lc>[Bool.pick]}"() for <what ever> # masak
camelia rakudo-moar 16847e: OUTPUT«what␤ever␤»
lizmat_ m: say ."{<uc lc>[Bool.pick]}"() for <what ever> # masak
camelia rakudo-moar 16847e: OUTPUT«what␤ever␤»
masak lizmat_: nice :)
lizmat_ m: say ."{<uc lc>[Bool.pick]}"() for <what ever> # masak
camelia rakudo-moar 16847e: OUTPUT«WHAT␤EVER␤»
masak except the reliance on Bool's numification ;) 09:36
lizmat_ why is that a bad thing ?
masak damn you RandalSwartz, now I'm getting allergic to it too :)
lizmat_: no reason, nvm 09:37
09:37 lizmat left 09:38 lizmat_ is now known as lizmat 09:39 lizmat_ joined 09:40 sjn_phone joined
RabidGravy masak, quite. I think having a La Trappe dubbel before bedtime last night may have broken my brain. 09:40
pmichaud RabidGravy: iiuc, S29 is fairly old and may contain many fossils. It's never really been accepted anyway. 09:41
masak lizmat: he once said in a p6l email (I think it was) that +Bool::True ought to be 42 just to teach people not to rely on it being 1... :)
RabidGravy: that's a good beer.
DrForr I have a 1l bottle of Westmalle trippel left over from FOSDEM.
09:41 Su-Shee joined
Su-Shee good morning 09:41
RabidGravy pmichaud, right.
masak morning, Su-Shee! 09:42
RabidGravy masak, it is indeed. I sensibly eschewed the Quadrupel however.
Su-Shee jnthn: see msg for the book :)
masak: why did I not see you in oslo? ;)
masak RabidGravy: maybe if you had that one you'd have typed "olooded" instead :P 09:43
09:43 lizmat left
masak Su-Shee: because timing was extremely unfortunate with @family stuff. 09:43
09:43 lizmat joined
masak comforting to hear I was missed, though... :> 09:43
RabidGravy masak, I think it would have been more: "like, you know, sorta kinda not thinged ...." ;-) 09:44
Su-Shee masak: oh, I see. it was extremely lovely. very .. family like :) and oslo is really nice, imho.
masak agreed.
I'll really really try to make it to YAPC::EU this year.
beyond that, not much is guaranteed. :/ 09:45
Su-Shee masak: yes, please. I think the spaniards are trying to make it extra-nice.
masak: also, I promised the program commitee to kick all of your perl6 butts for talks.
masak: so you have no choice.
"I am the borg. resistance is futile." <-- my management style. 09:46
FROGGS is it really "I am the borg" in English?
masak understood.
Su-Shee FROGGS: yes, the borg queen says it.
masak ok, #perl6 channel -- what should I talk about this year at YAPC::EU?
FROGGS Su-Shee: ahh, yeah 09:47
09:47 lizmat_ left
pmichaud Should I talk about something? 09:47
masak YES
Su-Shee masak: I would love to emphasis real world perl6 so to speak, so it would be awesome to show off common problems
pmichaud the timing would be very good for a discussion of language spec.
update/reprise of my FOSDEM talk, I suspect.
masak Su-Shee: ok, good angle. I do have a lot of real world Perl 6 code.
lizmat Missing test file: t/spec/S05-modifier/ignorecase-and-ignoremark.t # did someone forget to git add ?
pmichaud jnthn++'s perl 6 for beginners session yesterday went really well. A lot of excitement and enthusiasm among the participants. 09:48
masak \o/
virtualsue i was sorry to miss that
FROGGS ohh dang
lizmat: wait a sec
masak was it filmed?
09:48 agentzh_ joined
pmichaud I'm going to see if it's possible to squeeze in a similar event at yapc::na 09:48
Su-Shee masak: in no special order: things like e.g. streaming, parsing, processing, how to write a perl 6 module (on the plane I thought that maybe whipping up your own json module would illustrate how to make good practical use of grammars), networking up and down would be awesome.. 09:49
pmichaud masak: not filmed, no.
Su-Shee masak: C library bindings.. (I hear it's extra nice and easy)
pmichaud masak: it was more workshop than presentation.
masak pmichaud: aww :) but ok.
Su-Shee: I'm glad I asked. these are all good ideas, but I wouldn't have been able to come up with them myself.
Su-Shee masak: functional perl 6, async perl 6.... 09:50
masak: it's mostly that I ask myself: new language? what do I do with it usally, what are my every day uses?
dalek ast: 948770b | FROGGS++ | S05-modifier/ignorecase-and-ignoremark.t:
add test file for ignorecase+ignoremark
masak I wonder if I could make it along the lines of "Solved in Perl 6", in a near-Q&A style...
FROGGS lizmat: there you go
09:51 lizmat_ joined
Su-Shee masak: I would very much like to skip beginner's perl 6 stuff beyond maybe a day long workshop which would be awesome and make talks be firmly rooted in "now we're really using it for real things" 09:51
FROGGS lizmat: I just pushed the test file
Su-Shee masak: oh now THAT would make an awesome workshop - imagine half a day of "applied problem solving in perl 6 for the dirty, messy world" :) 09:52
nwc10 pmichaud: ++jnthn++'s thing being this 160 minute tutorial? act.osdc.no/osdc2015no/talk/6154
pmichaud nwc10: yes 09:53
masak m: my $jnthn; ++$jnthn++
camelia rakudo-moar 16847e: OUTPUT«Cannot assign to an immutable value␤ in block <unit> at /tmp/Vdues1aIft:1␤␤»
pmichaud it was highly attended
masak read "it was highly attenuated"
Su-Shee masak: and I would very much appreciate if you'd tell me whose butt I should kick as well :)
nwc10 perl6++ # better than ++C++ :-)
RabidGravy speaking of C language bindings is there something that I have missed that does something like metacpan.org/source/JSTOWE/TermRea...nfigure.pm or is it make one myself?
Su-Shee so. MR arnsholt! you are next ;)
DrForr Starting to wish I'd not decided to leave on the 12th :/
masak Su-Shee: jnthn, pmichaud, niner, arnsholt...
Su-Shee: FROGGS 09:54
pmichaud or, I should say, there were far more people present than jnthn (or I) expected.
Su-Shee masak: I just put on my good shoes to kick arnsholt's ;)
FROGGS I'm not coming to Grenada
masak Su-Shee: maybe you could make nwc10 give a p6 talk, I dunno.
FROGGS: :/
Su-Shee FROGGS: that's good because it's in granada ;)
FROGGS ahh :P
09:54 lizmat left
Su-Shee :) 09:54
FROGGS I'll be at the Swiss Perl Workshop though 09:55
btyler makes a note to submit YAPC::EU talk proposal tonight
DrForr Oh, this is ::EU discussion.
RabidGravy I think I may be in a position to start on making a p6 version of Term::ReadKey this week
masak m: say ((&lc, &uc).pick)($_) for <grenada granada what ever>
camelia rakudo-moar 16847e: OUTPUT«grenada␤granada␤what␤ever␤»
masak m: say ((&lc, &uc).pick)($_) for <grenada granada what ever>
camelia rakudo-moar 16847e: OUTPUT«grenada␤GRANADA␤what␤ever␤»
masak lizmat_: that's how to do it, btw. obviously ;)
09:55 bjz_ left
Su-Shee arnsholt: I would very much like to order a yapc::eu talk from you which centers around languages, encoding and such. not too academic. maybe internationalized regex stuff for example and how to deal with mixed language environment - e.g. you process a text containing greek and latin.. 09:55
pmichaud DrForr: I did throw in a ::NA note above :) 09:56
RabidGravy I haven't been to a YAPC::EU (or any perl conference for that matter,) since 2006
09:56 bjz joined, lizmat joined
virtualsue i remember that! you did some of the best talks too 09:56
FROGGS RabidGravy: we have LibraryMake
RabidGravy ;-)
Su-Shee btyler: what are you thinking about as a subject?
masak Su-Shee++ # your kicking butt is amazing to behold! 09:57
btyler Su-Shee: something like "distributed systems for poets and perl hackers"
Su-Shee nwc10: I'm supposed to make you give a talk. what does it take? cookies or violence?
DrForr Violent cookies.
arnsholt Su-Shee: I must confess that I know more about languages than about how to encode them =) 09:58
09:58 flussence joined
Su-Shee btyler: distributed systems is awesome. 09:58
nwc10 Su-Shee: an alpha release.
FROGGS you can throw cookies for sure
Su-Shee arnsholt: IT'S TIME TO LEARN IT THEN? ;)
nwc10 at which point, I might have a plan.
pmichaud for the last couple of days I've been unable to access yapcna.org -- anyone else able to reach it?
masak in some cultures, throwing a cookie is a great insult.
nwc10 pmichaud: I can access it from work
FROGGS pmichaud: I can't
DrForr Problem here too.
nwc10 and from a machine in the UK 09:59
RabidGravy FROGGS, yeah I'm thinking it's probably extend LibraryMake to do anything else that might be needed, the P5 Term::ReadKey actually writes some of its own C code depending on what it finds - I'm looking for something like autoconf but not so shit
arnsholt Su-Shee: If you write my PhD thesis for me, sure O=)
Su-Shee arnsholt: well I still would like a nice talk about language processing, not too academic. maybe we can whip up a couple of real world-ish practical uses of NLP?
09:59 lizmat_ left
FROGGS RabidGravy: "no so shit" :P 09:59
RabidGravy: IMO you can only create another kind of shit when you wanna do something autoconfy 10:00
lizmat FROGGS++
10:00 abraxxa left 10:01 abraxxa joined
RabidGravy yeah it's just the M4yness of autoconf that I don't like, I mean that was fine twenty years ago and all 10:01
Su-Shee pmichaud: after our short discussion I would love anything from a lightning talk to more along the lines of "cool shit of perl6", you know everything ranging from the "no floating point trap" to this "many versions next to each other" and "strings done right" and all the other things nobody knows ;) 10:02
dalek ast: 6085e09 | FROGGS++ | S29-type/declarations.t:
remove outdates Perl 6 types tests

There are better tests in S02-types/declare.t		 10:03
Su-Shee pmichaud: and after you mentioned you need an audience: "24 year old non-academic javascript hipster hacker" ;)
dalek kudo/nom: 2e8f308 | FROGGS++ | t/spectest.data:
remove removed test file
RabidGravy :)
10:04 lizmat_ joined
masak Su-Shee: one surprising thing about Perl 6 in the real world is that I almost never use junctions. or if I do, I don't think much of it. 10:05
pmichaud Su-Shee: I think I might come up with a series of 1-minute summaries of cool perl 6 features. Then my lightning talks can be a sampling of 5 from that series. :)
Su-Shee pmichaud: *hihi* "..and now the 38753 lightning talks by patrick michaud.." ;) 10:06
pmichaud: but that would be most lovely. and my promise to write it up as a more advertising paper holds. 10:07
pmichaud yeah, I'm thinking I could even do a @talks.pick(5) at the beginning of the talk and that determines what we hear about. :)
Su-Shee masak: I don't even remember what junctions are and why I would want them.
10:07 lizmat left
Su-Shee pmichaud: excellent. 10:07
FROGGS I did a lightning talk on Thursday, starting with subroutine signatures, over sub MAIN to multi MAIN where one had a where clause checking for the existence of the pfovided file... 10:08
RabidGravy gwoovy
FROGGS provided*
Su-Shee I mean, it's kind of really necessary to actually show things and talk about things and talk about it massively NOW shortly before the release which is perceived as 2the real one" is upon you..
masak Su-Shee: junctions are `if $value == 1 | 2 | 3` 10:10
FROGGS btw, the atmosphere at the GPW was quite nice P6 wise... I got many ppl asking for details because they tried stuff and of course used P5isms
Su-Shee oh! oh! and I want VERY MUCH a talk about the jvm backend in terms of "and like in clojure you just bind this stuff here and load it there and then suddenly your java things works perfectly.."
masak: ah those :)
masak right.
they look very cool on paper.
FROGGS ahh, and my lightning talk actually stated with: "Now you all go to perl6.org, click the download button and follow these instructions." :o)
10:11 laouji joined
Su-Shee now I have a cat on my ketboard... 10:11
10:15 laouji left, Isp-sec joined
pmichaud I wonder who/how I should contact about the yapcna.org site being inaccessible. 10:18
nwc10 pmichaud: for you, at what level does it fail? Can you traceroute to the machine? 10:19
pmichaud nwc10: I'm unable to telnet to port 80 on the machine.
10:20 xfix joined
nwc10 pmichaud: Oh odd. 10:20
DrForr traceroute fails at cpanel.net.
pmichaud I can try it from my home machine, though. just a sec.
Su-Shee it's www.yapcna.org - ypacna.org doesn't work. 10:21
10:21 lizmat joined
Su-Shee err yapcna of course 10:21
pmichaud ah. Google search results return "yapcna.org"
nwc10 ahaa.a
I guess my browser is "helpfully" adding the www and I'm too, er, $impolite-word to notice that
Su-Shee I just tried both. 10:22
pmichaud I did Google search for "yapc na 2015", then clicked the first result, and no response.
nwc10 and then, fool that I am, I copied the URL from the desktop browser (at work) to the remote machine
pmichaud okay, that helps. Now I at least have a contact address. 10:23
DrForr Because every website must have www. .
Su-Shee I always wonder how non-technical people manage when we already fail so often.
DrForr: you just configure it properly so it doesn't matter...
DrForr goes back to working on docker configs. 10:24
10:24 lizmat_ left
nwc10 pmichaud: I don't know who [email@hidden.address] or [email@hidden.address] are 10:24
(listed in the whois for 208.74.120.55)
lizmat jnthn has entered the building
nwc10 but they are screwed, as the DNS says that:yapcna.org mail is handled by 0 yapcna.org. 10:25
and the host is down
so mail is going to be bounced after 5 days
10:27 diana_olhovik_ joined 10:28 rindolf left
jnthn afternoon, #perl6 10:28
masak jnthn! \o/ 10:29
jnthn: time to submit a talk for YAPC::EU! :D
nwc10 heresy!
jnthn masak: When's the deadline? :)
dalek kudo/nom: 438274c | lizmat++ | src/Perl6/ModuleLoader.nqp:
Make module loading fail list better
masak jnthn: the deadline is Su-Shee is gonna badger you until you do! :D 10:30
Su-Shee I'm not worried about jnthn ;)
masak haha :) 10:31
10:34 lizmat_ joined 10:37 lizmat__ joined, lizmat___ joined
RabidGravy scrolls back a bit 10:38
10:38 lizmat left, lizmat___ is now known as lizmat
lizmat (switch from flaky wifi to hopefully not so flaky 3G) 10:38
RabidGravy DrForr: fancy writing a p6 version of "fig" or "docker-compose" or whatever it's called these days? It's on my TODO list, but keeps getting pre-empted 10:39
10:40 lizmat_ left
DrForr Maybe. I need to get what I've got off my plate... 10:40
RabidGravy I think the only particular yak that may need shaving is the http on unix domain sockets thing 10:41
10:41 lizmat__ left
DrForr RabidGravy: Specifically getting ReadLine documented so we can merge it :) 10:44
RabidGravy :) 10:49
10:49 Sysaxed` joined, agentzh_ left, diana_olhovik_ left 10:54 diana_olhovik_ joined
dalek ast: 7afaea0 | lizmat++ | S10-packages/precompilation.t:
Just try to unlink, don't create a race
 10:54
11:00 RabidGravy left
pmichaud .u 1f37b 11:06
yoleaux U+1F37B CLINKING BEER MUGS [So] (🍻)
pmichaud .u 1f37a
yoleaux U+1F37A BEER MUG [So] (🍺)
pmichaud r: my $beer = "\C[BEER MUG]"; say unicode $beer; 11:07
camelia ( no output ) 11:08
..rakudo-moar 438274: OUTPUT«5===SORRY!5===␤Unrecognized backslash sequence: '\C'␤at /tmp/tmpfile:1␤------> 3my $beer = "\7⏏5C[BEER MUG]"; say unicode $beer;␤Undeclared routine:␤ unicode used at line 1␤␤␤»
lizmat m: say "🍺".succ
camelia rakudo-moar 438274: OUTPUT«🍺␤»
pmichaud r: my $beer = "\c[BEER MUG]"; say unicode $beer;
camelia rakudo-jvm 438274: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Unrecognized character name BEER MUG␤at /tmp/tmpfile:1␤------> 3my $beer = "\c[BEER MUG7⏏5]"; say unicode $beer;␤»
..rakudo-moar 438274: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Undeclared routine:␤ unicode used at line 1␤␤»
pmichaud r: my $beer = "\c[BEER MUG]"; say uniname $beer;
camelia rakudo-moar 438274: OUTPUT«BEER MUG␤»
..rakudo-jvm 438274: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Unrecognized character name BEER MUG␤at /tmp/tmpfile:1␤------> 3my $beer = "\c[BEER MUG7⏏5]"; say uniname $beer;␤»
pmichaud m: my $beer = "\c[BEER MUG]"; say uniname $beer; $beer++; say uniname $beer; 11:09
camelia rakudo-moar 438274: OUTPUT«BEER MUG␤BEER MUG␤»
11:12 laouji joined 11:17 laouji left 11:19 brrt joined
masak like with infinity, incrementing beer mug just results in beer mug. 11:20
brrt \o 11:21
DrForr This is amusing. =begin Documentation .. =end Documentation throws "=begin not terminated by matching '=end pod'." Edit '=end Documentation' to '=end pod', guess what error I get? 11:22
Yep, "=begin not terminated by matching '=end Documentation'."
11:25 rurban joined 11:26 agentzh_ joined
masak DrForr: provide some minimal inputs to demonstrate this, and you have yourself a rakudobug ;) 11:26
++DrForr
DrForr In a moment. it's probably *actually* bad documentation :) 11:27
11:28 sjn_phone left
DrForr mispeling. 11:29
masak well, that saves some paperwork :) 11:30
DrForr Actually it may be a bug still, testing.
'=begin pod\n=begin Documentation\n=en Documentation' exhibits this behavior. I think it's still a bug as it's not picking up the misbalanced =begin..=end pairs. 11:33
masak DrForr: is there a typo ('=en') in what you just wrote? 11:36
or is it deliberate? 11:37
DrForr On channel, yes, I tyoped. Let me put it in a gist.
masak ah, good.
11:39 raiph joined
DrForr gist.github.com/drforr/3e789b78897759d6cc28 11:39
Now of course the real problem is that these aren't properly nested. 11:40
11:40 muraiki joined
DrForr A better error would be "=begin at line 1 has no matching =end"? 11:43
To be fair, if you pick up on the fact that the line in the *second* error isn't actually in the file, you can figure out what's going on, but it's hard to spot. And if you've been doign this for years you'll assume that the compiler simply got the line wrong :) 11:49
11:52 yqt joined, virtualsue_ joined
pmichaud would it be at all difficult to change the error message to read "=begin at line 3 not terminated by matching '=end Documentation'" ? 11:52
(i.e., adding "at line 3" to the message)
FROGGS I believe we have that for another message already 11:53
pmichaud or, perhaps better: '=begin Documentation' at line 3 not terminated by matching '=end Documentation'
FROGGS something about a block starting at an earlier line... TimToady++ might remember
11:53 virtualsue left, virtualsue_ is now known as virtualsue
DrForr pmichaud: That'd work as well. 11:54
11:54 aborazmeh joined, aborazmeh left, aborazmeh joined
DrForr Should I file a bug? 11:57
dalek kudo/nom: f9c9822 | FROGGS++ | t/spectest.data:
run test (where all tests are skipped)
ufobat Another Question, i've got anouther Task::Star test error. I guess its also regarding the change to Nil, like it was before with URI.pm 11:58
==> Testing Bailador 11:59
# Failed test 'route GET /foo exists'
# at t/01-route-existance.t line 13
# expected: [200, [:Content-Type("text/html")]]
# got: [200, [:Content-Type("text/html")], Any]
11:59 silug left
moritz yes, looks like it 12:02
moritz waits for the question 12:03
either open an issue for Bailador, or send a pull request that fixes it :-)
12:04 grondilu joined
dalek ast: 21024b0 | lizmat++ | S10-packages/precompilation.t:
Some more sanity and better layout
 12:07
12:11 domidumont left, Alina-malina joined 12:14 silug joined
dalek ast: a5a4fc2 | lizmat++ | S32-exceptions/misc.t:
Fix too specific test for Exception
 12:14
12:16 Ven joined 12:17 domidumont joined 12:20 eiro left 12:25 eiro joined
Su-Shee tadzik: you did panda, didn't you? 12:32
pmichaud DrForr: Yes, file a bug if you haven't done so already. Thanks. 12:33
DrForr Will do so.
12:33 aborazmeh left
lizmat Su-Shee: tadzik is travelling atm, I think 12:34
I'm looking at panda atm, is there something I can help you with ?
Su-Shee lizmat: no, I was just thinking a talk "how to module in perl6" would maybe be nice.. (creation, deployment, versioning etc etc) 12:35
lizmat: and I somehow saved in my memory that tadzik did a lot of "module"
lizmat he did 12:36
and he knows a lot about it
Su-Shee excellent. will threaten him with invasion if he doesn't comply :)
nwc10 and cake if he does?
12:37 rindolf joined
Su-Shee well I can invade and threaten him with cake, of course :) 12:37
12:43 laouji joined, rurban left 12:46 domidumont1 joined, domidumont left 12:48 laouji left 12:57 diana_olhovik_ left
masak tadzik: I would love a "war stories about making a package manager" talk or something like that. 12:59
12:59 cognominal joined
Su-Shee it's in general not as easy and simple to the uninitiated as people think because it's a rather long chain from opening your editor until upload/release. 12:59
dalek kudo-star-daily: 01c92a1 | coke++ | log/ (2 files):
today (automated commit)
 13:00
rl6-roast-data: 25827d6 | coke++ | / (9 files):
today (automated commit)
Su-Shee also, a talk from the area "deployment/software management/almost devops" might be nice. 13:01
FROGGS $ perl6-m -e 'say("flubber¼½worms" ~~ /<:Name(/:s LATIN SMALL/)>+/)'
｢flubber｣
FROGGS loves regexes
Su-Shee regexes are awesome. it was the very first thing I learned when I started "programming" and "computer" 13:02
masak m: .say for "flubber¼½worms".comb: /<:Name(/:s LATIN SMALL/)>+/ 13:10
camelia ( no output )
masak o.O
m: .say for "flubber¼½worms".comb: /r/ 13:11
camelia rakudo-moar f9c982: OUTPUT«r␤r␤»
13:11 yqt left
masak FROGGS: are you running a version from the future? 13:11
FROGGS masak: fsvo of future, yes 13:14
DrForr "15 minutes into the future." 13:15
ufobat moritz: i just wanted my thought to be confirmed since i have no clue 13:16
zostay what is the :D for in signatures and such?
ufobat but you did, so thanks :)
moritz zostay: it stands for "defined"
arnsholt zostay: :D is defined, :U is undefined, :T is type object (I think)
moritz zostay: that is, Int:D only allows Int instances, not the type object itself 13:17
FROGGS though, :T is not yet implemented and :U does what :T should do
Juerd found :D very ugly in code, but at least it adds a lot of smileys to compensate for the rather sad };
masak Juerd: also, you rarely need }; in Perl 6 code
zostay aha, thanks :D 13:18
FROGGS :D
masak };
krakan m: say 1 ~~ (1, 2, 3)
camelia rakudo-moar f9c982: OUTPUT«False␤»
Juerd masak: I know, but it feels wrong to leave out the ; after non-builtin-ish things...
ingy what's the general object serialization method in p6?
masak krakan: needs some any() 13:19
Juerd ingy: .perl? ;)
masak m: say 1 ~~ any(1, 2, 3)
camelia rakudo-moar f9c982: OUTPUT«True␤»
ingy ta
krakan masak: ah, ok, thanks
Juerd Is there a safe reverse of .perl?
masak Juerd: that's just because you've mostly been working in languages where the distinction between language and userland is constantly visible.
Juerd masak: Yes.
ingy .perlfect!
masak Juerd: I think I heard something about val() once 13:20
Juerd ingy: You perlvert! :D
masak Juerd: don't know what happened to val(), though.
m: say val("5")
camelia rakudo-moar f9c982: OUTPUT«5===SORRY!5=== Error while compiling /tmp/zc5n1qKkjG␤Undeclared routine:␤ val used at line 1␤␤»
ingy Juerd: You talkin to me, perlchance?!?!
Juerd Perlhaps. 13:22
ingy giggles and leaves
masak you two are very perlsistent.
13:22 molaf joined
Su-Shee shall I start singing "perls are the girl's best friend" and hop out of a cake? 13:23
pmichaud looks for more superlatives and fails.
Juerd masak: That's just your perlspective 13:24
13:24 agentzh_ left
Juerd pmichaud: Can I perlsuade you to continue looking? 13:25
pmichaud Juerd: perlhaps.
Su-Shee pmichaud: it would probably be perls before swine anyways ;)
FROGGS :P
Su-Shee++
DrForr The first pun that comes to mind I steadfastly *refuse* to use. 13:26
13:26 ab5tract joined
pmichaud perls before swine: drive.google.com/file/d/1n5Xl2Q6nI...sp=sharing 13:34
13:34 laouji_ left 13:41 agentzh_ joined
[Coke] (star-daily log broken?) not updated in a month because nothing has changed in a month. 13:43
whereas the version log changes every time we build a new version.
13:43 skids joined
Su-Shee pmichaud: hahahaha :) swine before perls! ;) 13:44
pmichaud Su-Shee: I'm reading left-to-right here. :) 13:45
Su-Shee *hihi* I read from the z-axis ;)
pmichaud: shall I ask why you have a SWINE handy? ;)
pmichaud I have no idea. Wendy brought it, I think.
Su-Shee well I have at least a postgres elephant. ;) 13:46
13:47 Isp-sec left
masak .oO( pmichaud has been doing teh perls for so long, he automatically has the swine to put the perls before ) :P 13:56
13:56 thou joined
jnthn
.oO( He uses Perl to bring home the bacon? )
 13:58
13:59 laouji joined
Su-Shee well. ;) loldamn.com/wp-content/uploads/2012...piglet.jpg 13:59
pmichaud Well, for a long time I was using s-pork for my slide presentations. :-P
FROGGS awwwww 14:00
Su-Shee I have cat pictures to illustrate any subject. ;) 14:01
pmichaud Su-Shee: ... a catalog?
14:02 RabidGravy joined
TimToady such backalog! 14:02
Su-Shee pmichaud: it's just moving from collection to catalog.. ;)
pmichaud good *localtime, TimToady
14:03 laouji left 14:04 lolisa joined 14:06 diana_olhovik_ joined 14:09 airdisa joined, domidumont1 left 14:10 domidumont joined
dalek ast: 039d420 | FROGGS++ | S05-metasyntax/unicode-property-pair.t:
add tests for unicode property pairs
 14:14
ast: 9fc4b56 | FROGGS++ | S05-metasyntax/unicode-property-pair.t:
fudge unicode test for jvm
 14:23
kudo/nom: 8741051 | lizmat++ | src/ (3 files):
Make 'use cur' and @?INC completely operational

With some great help from jnthn++		 14:24
p: 547bb85 | FROGGS++ | src/vm/moar/QAST/QASTRegexCompilerMAST.nqp:
implement <:Name(/:s LATIN SMALL LETTER/)> for moar
kudo/nom: 4c1d2d5 | FROGGS++ | t (2 files):
bump nqp and run another unicode test file
 14:25
FROGGS ohh, nice progress is nice 14:26
timotimo sweet
FROGGS and it is monday again *cough* :o) 14:27
timotimo yes, i know :) 14:28
i was waiting for more nice progress to appear from the hackathon, of course ;) 14:29
PerlJam good * all. 14:30
pmichaud PerlJam: /o
FROGGS hehe
pmichaud er, o/
FROGGS hi PerlJam
pmichaud (actually, my right arm was bothering me much of last week, so perhaps /o is correct. :)
moritz sounds like \o\
PerlJam moritz: not if he's facing me. 14:31
:)
timotimo .o( if The Perl Foundation sponsors a Pyra for me, I may end up doing a lot of portability work on MoarVM and Rakudo for ARM systems ... )
moritz: that kind of looks like "get back here, you damn arm!"
moritz
.oO( if TPF sponsors an extended trip to the Bahamas for me, I may end up doing a lot of Bahamas compabitlity hacking on Rakudo and MoarVM )
 14:32
Su-Shee moritz: veeery smooth! 14:33
moritz: "das merken die niiiie!" ;)
timotimo :D
moritz Su-Shee: :-)
b2gills masak: the pod in Hash::Ordered is copied after the __END__ marker by a Dist::Zilla plugin, and then the original pod is commented out so that line numbers match between the repository and the released module ( which is then hidden by default on metacpan )
DrForr Extended trip to the Bahamas? Hello...
masak b2gills: oh, so it's a Dist::Zilla thing? thank you -- TIL. 14:34
moritz dzil -- all the magic you ever wanted, and then some 14:35
14:35 FROGGS left
pmichaud more magic: www.catb.org/jargon/html/magic-story.html 14:36
hoelzro ahoy #perl6 14:38
PerlJam pmichaud: My first thought was "ground loop" and the 1994 paragraph at the end sounds like that's probably what it was. 14:39
pmichaud PerlJam: somehow that story came up in conversation this weekend and so I thought I'd past it in now 14:40
masak I remember finding a print version of the jargon file while still a teenager. made an impression on me. 14:41
PerlJam dzil's magic is more like "you are staring at a box with many switches, some of which are invisible. The box can do magical things ... if only you flip the right switches"
14:42 ribasushi joined 14:43 Isp-sec joined, xfix left
ab5tract hey lizmat 14:44
what is "use cur"
timotimo like "use lib"; cur stands for CompUnitRepo
ab5tract timotimo: ah, makes sense :)
muraiki imports mongrel dogs
ab5tract and resolves my complaint (which was, please no more ambiguous shortening of names!) 14:45
nwc10 er, epic FAIL 14:46
write string requires an object with REPR MVMOSHandle
14:46 cognominal left 14:50 agentzh_ left 14:51 _mg_ left
lizmat ab5tract: it's the new use lib 14:53
if the functionality is all operational, there will be a rm lib/lib.pm6 and s/cur/lib/ in World 14:54
*when
PerlJam Is there some problem building panda with rakudobrew? 14:59
dalek kudo/nom: fce74e1 | jnthn++ | src/core/Inc.pm:
Ensure we don't end up with standard handles in SC

Fixes the regression in 8741051.
jnthn Or at least, the biggy...lizmat++ is working on the regression 54-use-lib.t
14:59 FROGGS joined
ab5tract lizmat: interesting :) 15:00
15:03 isBEKaml joined
TimToady suspects that "cur" is a bit you-think-that's-cute-today-ish 15:04
PerlJam I just cloned a fresh rakudobrew and used it to install rakudo-moar which happened without a hitch. But "rakudobrew build-panda" did this: gist.github.com/perlpilot/26ba5f94a253d60460e9
Is this a known thing?
15:04 Sysaxed` left
TimToady or is 'cur' a temporary expedient? 15:04
jnthn TimToady: Temporary 15:05
TimToady: Will become 'lib' when it works nicely, afaiu
TimToady so it's intentionally cute today :)
jnthn aye
pmichaud gives rakudobrew a try
TimToady we need to figure out what a Dog type would be someday though, to go with the Rats and Cats 15:06
15:06 isBEKaml left
pmichaud .u 1f408 15:07
yoleaux U+1F408 CAT [So] (🐈)
pmichaud .u 1f400
yoleaux U+1F400 RAT [So] (🐀)
pmichaud .u 1f415
yoleaux U+1F415 DOG [So] (🐕)
15:10 larion left
lizmat TimToady: "cur" is a temporary expedient 15:10
15:10 larion joined
TimToady yeah, I figgered that out, after an extra cup of coffee or so 15:10
15:11 molaf_ joined 15:12 kurahaupo left
jnthn train & 15:12
15:14 rarara joined, molaf left 15:18 isBEKaml joined 15:19 cognominal joined 15:21 poul left 15:22 zakharyas left, g4 left 15:23 agentzh_ joined
brrt .tell jnthn could you try building moar+nqp on windows again 15:23
yoleaux brrt: I'll pass your message to jnthn.
brrt in 4 hours time or so
:-)
FROGGS brrt: I can 15:24
brrt oh, please do
i have no access to a VM right now
15:25 larion left, kurahaupo joined 15:27 lolisa left 15:28 telex left 15:30 telex joined 15:31 ab5tract left
brrt afk 15:32
maybe i'll have a vm somewhat later today, i will try to test 15:33
FROGGS brrt: I'm building it right now 15:34
brrt ah ok
great :-)
i'll check for results later
15:34 brrt left
dalek kudo/nom: 7fada0f | lizmat++ | src/Perl6/World.nqp:
Make sure @?INC is a Parcel
 15:35
kudo/nom: 57ae807 | lizmat++ | t/01-sanity/54-use-lib.t:
Make sure we test @?INC at compile time
FROGGS .tell brrt I was able build: perl6 version 2015.04-221-gfce74e1 built on MoarVM version 2015.04-105-gbec36ae 15:38
yoleaux FROGGS: I'll pass your message to brrt.
15:42 Ven left 15:51 itz_ joined 15:52 Ven joined
[Coke] ~~ from the preent. 15:53
*present!
15:53 domidumont left 15:54 agentzh_ left 15:55 Ven left, laouji joined
pmichaud PerlJam: here's what I got when trying "rakudobrew build panda": gist.github.com/pmichaud/701651b03fbfc6125093 15:56
(fail, but for a different reason than yours) 15:57
PerlJam pmichaud: my guess right now is that the latest rakudo+panda is temporarily suffering from the CUR changes 15:59
pmichaud PerlJam: that's been the line going around here as well for the last day or so
16:00 laouji left 16:01 lolisa joined
TimToady so, in *theory*, adding a submethod DESTROY () { self.sink } to Failure shouldn't have any effect... 16:01
in *practice*, not so much
blows up 18 test files
16:01 isBEKaml left
TimToady so we've got some Failure leaks somewhere 16:01
probably faulty sinking 16:02
ah, no, just 11 files, was out-of-sync with github 16:06
maybe Failure.DESTROY should detect that but just issue a warning for now
16:08 espadrine_ left
TimToady or maybe it's just an order-of-destruction issue for failures that were stored in variables, hmm... 16:08
still, probably indicates logic errors somewhere 16:09
not out-of-sync with github, we just have 7 or so failing spectests currently, having to do with modules 16:11
lizmat is surprised and checks 16:12
ingy TimToady: I was thinking… you probably know I do a ton of Bash programming these days, and for CLI stuff the code really can't be beat by Perl or anything in its class. Mostly because IPC stuff is so beautifully integrated.
so I was wondering if Perl6 could be awesome in that realm it might be part of the evasive "killerness" 16:13
16:13 gfldex joined
ingy I mean "beat" in terms of succinctness of the task, and also handling all the devils in IPC stuff (that other langs bolt on) 16:14
lizmat TimToady: confirrmed here, looking at them 16:15
16:16 eli-se joined
ingy anyway, just wanted to float that out there 16:16
eli-se morning 16:19
cognominal ingy got a pipe dream? 16:23
ingy cognominal: cute :) 16:24
lizmat m: use lib "foo"; .say for @*INC; EVAL q/BEGIN .say for @*INC/ # huh?
camelia rakudo-moar 57ae80: OUTPUT«file#foo␤file#/home/camelia/.perl6/2015.04-223-g57ae807/lib␤inst#/home/camelia/.perl6/2015.04-223-g57ae807␤file#/home/camelia/rakudo-inst-2/share/perl6/lib␤file#/home/camelia/rakudo-inst-2/share/perl6/vendor/lib␤file#/home/camelia/rakudo-inst-2/s…»
TimToady cognominal: nah, he's just trying to redirect the conversation
lizmat m: use lib "foo"; say +@*INC; EVAL q/say +@*INC/ # huh? 16:25
camelia rakudo-moar 57ae80: OUTPUT«9␤8␤»
cognominal TimToady, but not to /dev/null 16:26
16:26 kbenson_ joined
ingy TimToady, cognominal: well you punsters may have answered me a bit. Perl 6 can lexically make syntax be (s)hellish :) 16:26
lizmat m: use lib "foo"; say +@*INC; EVAL q/say +@*INC/; say BEGIN +@*INC # huh?
camelia rakudo-moar 57ae80: OUTPUT«9␤8␤9␤»
cognominal ingy, there is a place for a slang that mimics the shell
16:27 airdisa left, airdisa joined
ingy cognominal: ? 16:28
16:29 maddingue left
ingy cognominal: I'm not sure what you mean 16:30
cognominal Perl 6 is about weaving sublanguages (or slangs), so far the "normal" slang, the quote slang, the regex slang. A shell language would be a nice addition.
ingy ah, yes
I think we can mark this one "solved in theory" then :) 16:31
16:31 maddingue joined
cognominal I have thinking about it a long time but I am not much a programmer. 16:31
ingy Seems like something I might be interested in doing at some point
dalek kudo/nom: 8013195 | TimToady++ | src/core/Failure.pm:
warn if Failure.DESTROY finds unhandled one

Hopefully not a user-facing warning (much), but should help us find failure leaks and plug 'em.
cognominal I am trying to get up to speed with nqp.
16:32 mohij joined, airdisa left
ingy cognominal: If I were going to do this I'd look through all my code and find the nicest parts to keep 16:32
cognominal My main complaint so far is that nqp is a different language from Perl 6 making difficult to derive slangs.
ingy I use bash in xt/ tests a lot for CPAN modules now
especially the CLI ones (which are a lot) 16:33
16:33 spider-mario joined
ingy I ported Test::More to bash some time ago 16:33
so I can run all the perl and bash tests with the same `prove` statement
since prove honors the hashbang 16:34
although I think it may not work with perl6
moritz does that work on Windows?
ingy moritz: which part? 16:35
moritz ingy: the tests written in bash, and running them with prove
ingy that's 2 parts :)
moritz ingy: and do any of them work on Windows? 16:36
ingy they are xt/ tests, so the first answer is "not critical"
moritz evasion means "no" :-)
ingy but the second answer it "I care, and getting closer"
moritz still sounds like a "no" to me :-) 16:37
ingy moritz: no
moritz: Is Perl 6 ready for production? 16:38
I mean it's a loaded question
cognominal ingy, in fact, recently, I said here shell for the Perl 6 repl, and apparently eyes glazed. In my mind, that's the same thing, just != syntaxes.
16:38 lizmat left
ingy do you want to flog me or talk about it? :) 16:38
nwc10 ingy: I think a fair answer is "not before the GLR is done"
ingy nwc10: so "no"
nwc10 ingy: indeed. not yet. 16:39
moritz ingy: "do the tests run on windows" doesn't quite sound like a very loaded question to me
ingy well you clearly wanted to point out that I hadn't thought about windows
but I have and I'm working on it 16:40
moritz ingy: I didn't want to point out anything; I simply wanted to know
skids hugme: hug ingy, moritz
aww no hugme?
ingy well I explained it to you and you pushed me into a corner 16:41
moritz I *know* you can run bash scripts on windows if certain conditions are met (some kind of POSIX env installed)
ingy just sayin
moritz ingy: sorry about that
ingy np :)
my point was that I write a lot more (dev/release) tests for CLI system interaction now, because it is easy 16:42
I'd like that ease in perl6 16:43
16:44 lizmat joined
cognominal me too. Lke the Apple motto is "there is an app for that", the Perl 6 will be "there is a slang for that" 16:45
but currently to much an involved operation 16:46
not even sure there is an advent post about that 16:47
perl6advent.wordpress.com/2013/12/...16-slangs/ # I am wrong
but that is a mere appetizer 16:48
16:48 abraxxa left 16:49 airdisa joined
ugexe -I doesn't appear to work anymore 16:50
jnthn .tell brrt I can confirm it's foxed
yoleaux 15:23Z <brrt> jnthn: could you try building moar+nqp on windows again
jnthn: I'll pass your message to brrt.
jnthn .tell brrt I can confirm it's *fixed*
yoleaux jnthn: I'll pass your message to brrt.
lizmat ugexe: can you elaborate? 16:54
ugexe if i do 'perl6 -Ilib t/some-testfile.t' it will tell me it cant find 'Some::Module' and then lists the compunitrepos but /lib is not listed 16:55
16:56 virtualsue left
ugexe same when i try to install panda... it cant find File::Find 16:56
dalek rl6-roast-data: 84b93ac | coke++ | / (9 files):
today (automated commit)
 16:57
tony-o i get file#lib in my output for that, i do have the issue with panda though
16:57 spider-mario left
tony-o Missing or wrong version of dependency 'src/gen/m-CORE.setting' (from 'lib/File/Find.pm') 16:58
lizmat ugexe: confirmed, investigating that as well
ugexe tony-o: you have leftovers from a previous install 17:00
ive made a habit of doing rakudobrew nuke moar && rakudobrew build moar
tony-o i don't know what nuke is going to do but it sounds like fun 17:01
ah ^
zengargoyle a few days ago i did an upgrade of a long dormant perlbrew setup... had lots of breakage and ended up nuking everything and starting from blank
tadzik Su-Shee: yes, I did panda :)
17:01 spider-mario joined
tony-o maybe that's why i also see file#lib with -I 17:01
tadzik masak: hmm, talk idea sounds interesting 17:02
masak: although the war stories would be rather peaceful :)
ugexe tony-o: i am trying to run a test which needs it before compile time (via a use statement). doing -Ilib -e 'say @*INC' is after compile time 17:03
timotimo japhb: it'd be nice if bench quickstart would complain about the several perl5 dependencies missing before starting to do a lot of stuff 17:05
[Coke] still getting a bunch of failures on the daily runs in integration/advent2009-day17 , across backends. 17:06
lizmat error: a cherry-pick or revert is already in progress
how do I fix that ?
Am about to revert my commits of today :-(
timotimo resetting git with --hard perhaps?
that'll nuke all state i suppose, and all changes made so far
lizmat did that already 17:07
[Coke] looks like that's a stress test, not a spectest.
timotimo huh
[Coke] git cherry-pick --abort ? 17:08
(assuming there was some merge conflict that got in the way there on the cherry pick?)
17:09 _mg_ joined
lizmat I had to stash stuff after I tried it for the first time 17:09
[Coke] maybe the stash and the half done cherry pick conflicted? iunno. 17:10
lizmat I cloned again, did the revert there, copied the files, running spectest now 17:12
17:12 pyrimidine joined
tony-o ugexe: i know - here is a gist: gist.github.com/tony-o/8a97b8d965703729d6b7 17:12
ugexe: this is freshly built ^ 17:13
17:13 yqt joined 17:14 _mg_ left
ugexe so it would seem its if its a dependency of something that is 'use'd it doesnt get found before @*INC is populated? 17:15
lizmat we're shutting down in Oslo right now 17:16
ugexe so if your 'use Dumy' was a real module but had a 'use Dummy2', it wouldnt find Dummy2 17:17
dalek kudo/nom: bcc8e2c | lizmat++ | src/ (3 files):
Revert the @?INC work of today, it broke too much

This should unbreak the spectest, but am not sure about panda		 17:18
lizmat away for a few hours at least&
17:18 lizmat left
Su-Shee tadzik: lovely! did you see the plans I have in mind for you? ;) 17:24
tony-o ugexe: ill give it a shot with a real module
vendethiel lizmat++ # experimenting a lot to improve what we have :) 17:25
Su-Shee tadzik: any other talk idea welcome too.
ugexe lizmat: github.com/rakudo/rakudo/blob/8741...d.nqp#L644 + Line 648/649. Its just pushing 1 or 0 onto $INC
oh nm
17:28 koo6 joined 17:29 FROGGS left
tadzik Su-Shee: yeah, something about cake :P 17:29
timotimo is currently running a benchmark session
tadzik Su-Shee: so was this about talks or inv^Wcake?
Su-Shee tadzik: I promised to kick perl 6 butts for yapc::eu to have more perl6 - real world perl6 - talks and more than "jnthn does 5 talks" :) and I didn't want to leave out you ;) 17:31
tadzik :) Alright
17:31 FROGGS joined
tadzik when does cfp end? 17:31
Su-Shee tadzik: so one idea was "how to module in perl6" - something along the line of toolchaining/deployment/ and such. there's more to a module than just making it 17:32
tadzik: no idea, lemme look
17:33 pdcawley_ joined, pdcawley left
Su-Shee tadzik: it doesn't say. ;) 17:33
DrForr Jun 30. 17:34
Su-Shee DrForr: where did you find that?
DrForr act.yapc.eu/ye2015/pricesanddates.html
[Coke] bartolin: Looks like 125146 may also relate to the crash in t/spec/integration/advent2009-day17.t 17:35
DrForr I'm tempted to create a talk on grammars actually.
Su-Shee DrForr: well as long as you make it "real world" and not some rare uber-fancy thing you'll encounter exactly once in your life and nobody elses.. :)) also, ENCODING? ;) 17:36
tadzik: 30th june
DrForr Does parsing JS count? :)
Su-Shee DrForr: for example.. 17:37
PerlJam DrForr: I think if you could show a grammar that parses JS and then how to derive from that a grammar that parses coffeescript (for example) would be very interesting :)
Su-Shee DrForr: or, I was thinking maybe something like making a JSON module live.. (not that there is another one needed probably..)
PerlJam: hey! order stuff for your yapc! ;) 17:38
vendethiel grammar that parses coffeescript? good luck with that...
(hint: indentation is by far not the hardest part :P)
DrForr Well, I'm putting the finishing touches on an ANTLR parser as I type. 17:39
Su-Shee DrForr: ok, stupid question: can it generate js as well?
hoelzro what's the difference between my $result = callsame; and my \result = callsame? what exactly is the latter doing? 17:40
PerlJam hoelzro: the first imposes an item context on the result, the latter does not.
vendethiel hoelzro: not containerizing
hoelzro ahhhhhh
thanks PerlJam, vendethiel
DrForr Not that I'm aware of, and that would probably get ito the mind-mangling realm. 17:41
DrForr hears a loud *pop* from the shelf that holds the *good* alcohol and notices for the first time how badly distorted the TV shelf looks. 17:42
tadzik Su-Shee: alrighty then 17:46
DrForr Well, generating JS from *perl* would be close to mind-mangling, generating JS from a data structure fairly simple but kind of pointless, as you'd have to ensure by hnad that you're generating syntactically valid JS, kind of defeating the purpose.
tadzik Su-Shee: although to be fair, your idea is more dogfooding than real world perl6 :) 17:47
Su-Shee DrForr: what I never get: why would I want to parse js from perl? what do I do with it then?
tadzik Su-Shee: I could talk about gamedev again, each time I do it forces me to work on my engine some more
Su-Shee tadzik: consider it an idea - you have done lots of relevant things and deployment/modules/version management is relevant to bring p6 into business 17:48
tadzik true
I can submit two :)
Su-Shee tadzik: most certainly you can do that!
DrForr JS is just an example, I'm sure there are more relevant languages or even just configuraiton file formats to work with.
Su-Shee DrForr: that was a real question, I never understood why I would want to parse lang x in perl. config files I understand of course. 17:49
DrForr Roundtripping JS-perl-JS might be more useful, and not *that* hard to do.
PerlJam Su-Shee: It's a middle step from auto-generating working software from a high-level description of what the software should do (and the generating software should clearly be written in perl ;) 17:50
Su-Shee PerlJam: yes, but what is it good for? what do I end up with what is useful to me?
PerlJam: and: is it relevant for many developers and their every day work? 17:51
PerlJam Su-Shee: "Computer, make me a cake" *poof* Cake. ;-)
DrForr It's a middle layer. On its own it's not terribly useful, but that doesn't mean it doesn't need to be done.
Su-Shee PerlJam: that is a totally useless metaphor for me. a real world example please.
DrForr And what I'm working on lets you bootstrap every language there's an ANTLR grammar for.
Su-Shee is too stupid to get from "Cake" to "useful software"
17:53 yqt left
Su-Shee what I'm getting at is having as many perl 6 talks as possible which are a good addition to the christmas announcement of "perl 6 production ready" and that means: anything useful for production. ;) I know that you can do cool brainfuck stuff with it. but how many times do you need that "in production"? 17:54
PerlJam good point. practicality first. Crazy futuristic stuff later. 17:55
Su-Shee yes. because if you want to convince anybody to _actually_ use perl6, it needs to be shown to people. 17:56
jnthn Or use the crazy futuristic stuff to solve very practical problems very well. :P
Su-Shee if you can bring it to the masses and make it useful for us common folk.. :) 17:57
DrForr If all you want is production tools rewritten in perl6 then why bother rewriting in the first place? Oh, I know, futuristic bugs!
Su-Shee DrForr: I'm caring only for the marketing, perception and success.
DrForr: but I'm happy to take any great idea making .. 24 year old javascript hipsters flock to perl 6 :) 17:58
DrForr Right. I guess what I'm really trying to say is there isn't a large enough userbase to know what the "killer app" is going to be just yet.
tony-o less the 24 yr old part some of us are already here Su-Shee
Su-Shee yes, and how would you create such a userbase?
tony-o free healthcare 17:59
DrForr Egg, meet chicken. Chicken, egg.
PerlJam Eggs came first.
17:59 eli-se left
Su-Shee oh well. submit whatever you want; in the end it's your community's choice on where to place yourself. 18:00
tony-o the async stuff in perl6 is what originally got me to kick the node fix
Su-Shee yes. I will cover that.
I submitted a talk "against node envy" in february already. ;)
jnthn will continue the async fixing/polishing work after his next couple of days teaching are done. :) 18:01
Su-Shee jnthn: your talk convinced me to add perl6 ;)
tadzik Su-Shee: ooh, that sound sinteresting :)
Su-Shee jnthn: the one yesterday.
vendethiel Su-Shee: hi, I'm a 20y-o *coffeescript* hipster that flocked to perl6 :)
Su-Shee awesome that makes two. now bring the other couple of hundred. ;)
vendethiel working on that. though I havn't written my fpw slides yet.. :) 18:02
tony-o Su-Shee: are those talks available online?
timotimo can a single person be said to "flock"?
PerlJam Su-Shee: I've occasionally thought the Perl 6 community should have a document that tries to answer why certain groups of people would use Perl 6. "I'm new to programming, why Perl 6?" "I'm a Java developer, why would I use Perl 6?" "I'm a Perl 5 programmer, why would I use Perl 6?"
18:02 larion joined
PerlJam etc. 18:02
vendethiel timotimo: i'm pretty fat :P
Su-Shee vendethiel: tony-o: then please please submit a talk "why I abondoned node.js for perl6" or something like it.
vendethiel Su-Shee: never would I use such a lowly tactic :) 18:03
Su-Shee tony-o: which talks?
timotimo vendethiel: wait what.
Su-Shee vendethiel: ?! why??
vendethiel because nodejs is web scale! you turn it up and it scales right up
tadzik vendethiel: why? It's okay, noders probably won't see it anyway :P
tony-o timotimo: i think a single person joining another group is considered flocking, if for no other reason just colloquial
vendethiel just like mongodb
tony-o Su-Shee: the against node envy
Su-Shee tony-o: I _submitted_ it - it's a talk for yapc::eu in september. ;) 18:04
haven't even written the code yet
tony-o ah!
DrForr Port Mojo, Dancer and whatever you like to perl6, you'll have Mojo and Dancer with different bugs. We've got to have time to find out what Perl6 will be good at, we're too busy finding and fixing bugs at the moment :)
Su-Shee tadzik: well people told me they liked the comparison aspect of the web frameworks thing, so I decided I'll compare the async stuff
vendethiel: so and why wouldn't you submit such a talk? 18:05
vendethiel Su-Shee: because it sounds like buzz titles
DrForr And I use 'we' to mean 'you' as I really haven't done a lot yet.
Su-Shee vendethiel: dude, that's why I asked for it.
18:05 rindolf left
vendethiel (I'd rather not have "dude" :P) 18:06
Su-Shee vendethiel: you can fill it with real content. it still IS a relevant thing.
vendethiel yeah, but I want people to be interested in what I have to say, not in some buzzfeed-like title
Su-Shee vendethiel: honey [] dear [] that's why I asked for it. ;)
PerlJam huh 18:07
Su-Shee vendethiel: you do realize that perl in general is desperate for interesting ideas how to make people be interested again?
it would be the perfect addition to lestrad's talk how he made the yapc::asia 1200 people large.. 18:08
DrForr One of the major idaes I remember from that is rebranding to more 'web development'. 18:09
18:10 cschwenz joined
Su-Shee well it's none of my business what perl 6 as a community thinks is a good idea, I'm just making sure that you actually submit talks and have a shot at convincing people... 18:12
18:15 yqt joined, eli-se joined
DrForr For my part the most convincing argument I saw was Curtis pointing out how much cr*p Java coders have to put up, and how much of that Perl6 gets rid of. 18:16
Su-Shee yeah, but on the other hand: java 8 is actually quite decent.. 18:17
DrForr And most of my time so far has been spent alternately finding compiler bugs and writng chunks of code while trying to ignore the "I'm about to get mugged" sensation that comes with new libraries and a new *language*. 18:21
18:22 raiph left
jdv79 mugged? 18:22
18:22 Hor|zon_ joined 18:24 raiph joined, kst`` joined
DrForr Hyperbole mostly, but when I was starting out I'd go along writing code, then find out "Oh, this doesn't work as it used to, what do I do *now*? Oh, nice, *that's* differ..wait, how do I... Oh, cute. 18:24
s/$/"
18:25 ggoebel2 joined, kbenson_ left, diana_olhovik_ left 18:26 betterwo1ld joined, tony-o_ joined, tony-o left, yakudza left, raydiak left, StavroMueller left, betterworld left 18:27 Patterner joined, StavroMueller joined, kbenson_ joined, diana_olhovik_ joined, agentzh joined, arnsholt joined, Khisanth joined, eternaleye joined, hernan joined, Possum joined, torbjorn joined, b^_^d joined, BooK_ joined, ponbiki joined, eternaleye left, yakudza joined 18:28 eternaleye joined, eternaleye left, eternaleye joined 18:29 raydiak joined 18:30 domidumont joined, b2gills joined
FROGGS TimToady: instead of reinterpreting a regex subset in INTERPOLATE's EVAL, can't we mess with EVAL's context so that it does not see e.g. $arg? 18:34
m: my $a = '<{ say $arg }> <?>'; '' ~~ / <$a> /
camelia rakudo-moar bcc8e2: OUTPUT«<{ say $arg }> <?>␤»
FROGGS m: my $x = 42; EVAL 'say $x' 18:35
camelia rakudo-moar bcc8e2: OUTPUT«42␤»
FROGGS m: my $x = 42; EVAL 'say $x', :!context
camelia rakudo-moar bcc8e2: OUTPUT«Cannot call EVAL("say \$x", :!context); none of these signatures match:␤ (Cool $code, Str :lang($lang) where { ... }, PseudoStash :context($context))␤ (Cool $code, Any :lang($lang) = { ... }, PseudoStash :context($context))␤ in block <unit> at…»
18:35 jgoff joined
FROGGS m: my $x = 42; EVAL 'say $x', :context(PseudoStash) 18:36
camelia rakudo-moar bcc8e2: OUTPUT«42␤»
timotimo :context(CORE) :)
er, i mean CORE::
FROGGS m: my $x = 42; EVAL 'say $x', :context(CORE::)
camelia rakudo-moar bcc8e2: OUTPUT«5===SORRY!5=== Error while compiling EVAL_0␤Variable '$x' is not declared␤at EVAL_0:1␤------> 3say 7⏏5$x␤»
FROGGS hehe
hmmmm
I'll try that
timotimo++
18:36 brrt joined
DrForr I,I "I got 99 problems but a switch ain't one." 18:37
18:38 rba_ joined
FROGGS ohh, I can perhaps pass the correct CALLER:: around 18:38
timotimo hmm 18:39
seems like current rakudo-moar on nom "writes to STDERR" on rc-forest-fire; maybe a fix is needed for latest flattening/itemization fixes/changes
brrt \o 18:40
yoleaux 15:38Z <FROGGS> brrt: I was able build: perl6 version 2015.04-221-gfce74e1 built on MoarVM version 2015.04-105-gbec36ae
16:50Z <jnthn> brrt: I can confirm it's foxed
16:50Z <jnthn> brrt: I can confirm it's *fixed*
18:37Z <PerlJam> brrt: Have you seen www.cs.tufts.edu/~nr/pubs/tiler.pdf
brrt yay
DrForr m: token f { 'a'? } "b" =~/<f>/; 18:41
camelia rakudo-moar bcc8e2: OUTPUT«5===SORRY!5===␤Strange text after block (missing semicolon or comma?)␤at /tmp/5AadvQoo4p:1␤------> 3token f { 'a'? }7⏏5 "b" =~/<f>/;␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end…»
brrt PerlJam - not read it, but have seen the strategy mentioned iirc
FROGGS m: token f { 'a'? }; "b" =~/<f>/;
camelia rakudo-moar bcc8e2: OUTPUT«5===SORRY!5===␤Unsupported use of =~ to do pattern matching; in Perl 6 please use ~~␤at /tmp/78rlyweHjj:1␤------> 3token f { 'a'? }; "b" =~7⏏5/<f>/;␤Other potential difficulties:␤ Useless declaration of a has-scoped method in mainlin…»
brrt thanks :-)
DrForr Oh, right. IN any case I was actually interested in whether optional terms now return 'Empty' as opposed to 'Nil'. 18:42
brrt fwiw, i made a tiny mistake, but it works out nicely; instead of 32 bytes of scratch space we now have 64 bytes of scratch space
18:42 jgoff left
brrt i.o.w. 'enough for every last caller-saved GPR' 18:42
dalek p/mast_localref_2: 34b1373 | timotimo++ | t/moar/02-qast-references.t:
returning a str from a localref scoped var is wrong
 18:43
tony-o_ m: ('a'..'z').pick(64).join('').chars.say 18:44
camelia rakudo-moar bcc8e2: OUTPUT«26␤»
18:44 spider-mario left
tony-o_ oh i need roll 18:44
m: ('a'..'z').roll(64).join('').chars.say 18:45
camelia rakudo-moar bcc8e2: OUTPUT«64␤»
18:49 espadrine_ joined
grondilu m: "hi"\ .say # suddenly wondering if unspaces work 18:49
camelia rakudo-moar bcc8e2: OUTPUT«hi␤»
tony-o_ m: "hi".\ s\ a\ y 18:50
camelia rakudo-moar bcc8e2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/qFB_lBm97z␤Two terms in a row␤at /tmp/qFB_lBm97z:1␤------> 3"hi".\ s\ 7⏏5a\ y␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ …»
grondilu I don't think you can put them inside what would otherwise be a token.
18:50 meisl joined 18:51 spider-mario joined
timotimo right 18:51
18:51 haidarian left
japhb timotimo: Did you create a GHI for that perl6-bench quickstart idea (bailing out if the perl5 does not have modules it needs)? 18:52
timotimo not yet 18:54
but if i do, it could also scan the benchmarks for use statements 18:55
Data::Alias and List::Util at least
oh, wait
those need to be installed in the component itself, not the system, right?
18:57 brrt left
FROGGS timotimo: seems I managed to get the context right 19:03
19:03 pdcawley_ left
FROGGS le spectest will tell 19:03
19:04 Sysaxed` joined, pdcawley joined
timotimo cool :) 19:19
japhb timotimo: Correct. 19:25
Something I started and never completed. :-(
19:25 domidumont left
dalek kudo/nom: f0e142b | FROGGS++ | src/ (2 files):
fix context of EVAL in regex interpolation
 19:30
Su-Shee so. I will remind you all about the deadline once or twice. ;) SUBMIT.
19:30 Su-Shee left 19:32 laouji joined
meisl Hi there, thx again for the kind help I received lately on the nqp toolchain :) 19:36
FROGGS hi meisl
meisl hi FROGGS
19:37 laouji left
meisl got a (very simple) compiler running, for my lambda calculator thingy :D 19:38
19:38 camelia joined
masak 'night, #perl6 19:39
meisl have a good rest, masak
FROGGS gnight masak
19:39 cschwenz left 19:40 ChanServ sets mode: +v camelia
FROGGS m: my $foo = 42; my $a = '<{ say $arg }> .'; say 'bar' ~~ / <$a> / # this will stop working in a bit 19:42
camelia rakudo-moar bcc8e2: OUTPUT«<{ say $arg }> .␤<{ say $arg }> .␤<{ say $arg }> .␤<{ say $arg }> .␤Nil␤»
FROGGS m: my $foo = 42; my $a = '<{ say $foo }> .'; say 'bar' ~~ / <$a> / # and this will work instead
camelia rakudo-moar bcc8e2: OUTPUT«5===SORRY!5=== Error while compiling EVAL_0␤Variable '$foo' is not declared␤at EVAL_0:1␤------> 3anon regex { <{ say 7⏏5$foo }> .}␤»
jdv79 what does "1<&1" mean? 19:43
meisl jdv79: on Windows I guess? 19:44
FROGGS I know about 2>&1, but not 1<&1
2>&1 does work on windows too btw
19:45 airdisa left
meisl without having tried: seems like a no-op to me...? 19:45
jdv79 just looking at rt.perl.org/Public/Bug/Display.html?id=77650
geekosaur I'd expect it to be a no-op as well 19:46
jdv79 ok
FROGGS jdv79: this is just some code to produce a warning
19:47 muraiki left
FROGGS m: sub 1 { 2 }; say &1 19:47
camelia rakudo-moar bcc8e2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ouL2IdpO42␤Missing block␤at /tmp/ouL2IdpO42:1␤------> 3sub7⏏5 1 { 2 }; say &1␤ expecting any of:␤ new name to be defined␤»
geekosaur oh, that's being evaluated by perl6, not by the shell. syntax error. 19:48
FROGGS ohh, &1 translates to $/[1] it seems 19:49
m: say "foo" ~~ /(.)(.)(.)/; say &1
camelia rakudo-moar f0e142: OUTPUT«｢foo｣␤ 0 => ｢f｣␤ 1 => ｢o｣␤ 2 => ｢o｣␤｢o｣␤»
FROGGS m: "foo" ~~ /(.)(.)(.)/; say &1
camelia rakudo-moar f0e142: OUTPUT«｢o｣␤»
FROGGS weird
jdv79 FROGGS: where'd you find that out?
FROGGS look at: perl6-m --target=ast -e 'say &2' 19:50
jdv79 anyway that ticket seems bogus, right? closable as per last comment?
FROGGS - QAST::Op(call &say) say &2
- QAST::Op(call &postcircumfix:<[ ]>)
- QAST::Var(lexical $/ :decl())
- QAST::WVal(Int)
meisl so, to be clear - this has absolutely nothing to do with shell/cmd redirection, right? 19:51
FROGGS hmmm, I dunno if these test function are supposed to swallow stderr... so I cannot judge
meisl: correct
[Coke] m: rakudo: use Test; eval_lives_ok '1<&1' 19:52
camelia rakudo-moar f0e142: OUTPUT«not ok 1 - ␤␤# Failed test at /tmp/GwlaEa0J1v line 1␤# Error: Whitespace required before < operator␤»
jdv79 iirc, and the last comment confirms, tests do generate STDERR in the p5 world. so how to find out if that is correct in this world? 19:53
[Coke] m: rakudo: use Test; eval_lives_ok '1 <&1'
camelia rakudo-moar f0e142: OUTPUT«Use of Nil in numeric context at EVAL_0 line 1 in any <unit-outer> at EVAL_0:1␤␤ok 1 - ␤»
FROGGS m: my $foo = 42; my $a = '<?{ say $foo }> .+'; say 'bar' ~~ / <$a> / # \o/
camelia rakudo-moar f0e142: OUTPUT«42␤｢bar｣␤»
19:54 Possum left
[Coke] I *think* the original complaint was that test builtins shouldn't be complaining to stderr. I think if the test does something that leaks to stderr, that's fine. 19:54
19:56 |Tux| joined
meisl I think this test is really about what leaks to stderr or not, &1 being just a thing to make eval complain 19:56
geekosaur I think most of us expected that to be a shell thing, not a perl6 thing
meisl Coke: yes, just what you said
geekosaur someone probably noticed this when they botched quoting of a redirected perl6 :)
jdv79 yeah, ok. i don't get why there should be that distinction but sure:) 19:57
meisl well, that's what I'd call a bad test: whatever the result should be or not - you're being led to a completely wrong path at first sight 19:58
so instead of using &1 in that (alleged) comparison I'd propose using $IamNotNumeric 19:59
jdv79 stderr is potentially poisonous in any case. i could better understand filtering all of it to prepend a # but *.
meisl and then the thing about ws before <, is the test about that or the non-numeric in numeric context?! 20:00
20:09 colomon left
literal W 4 20:11
oops
20:11 atweiden joined 20:12 Possum joined 20:13 airdisa joined 20:14 eli-se left 20:26 airdisa left 20:27 lizmat joined
dalek ast: f962253 | (Justin DeVuyst)++ | S24-testing/3-output.t:
Tests for RT #77650.
 20:28
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=77650
20:30 eli-se joined 20:31 espadrine_ left, colomon joined 20:34 laouji joined 20:36 molaf_ left 20:38 laouji left 20:39 cygx joined
cygx o/ 20:40
20:42 ufobat left 20:43 darutoko left 20:44 brrt joined
jdv79 is rt.perl.org/Ticket/Display.html?id=77350 misclassified as closable with tests? 20:44
FROGGS m: grammar A { token TOP { <so> }; token so { 'foo' | 'bar' } }; say A.parse('foo') 20:45
camelia rakudo-moar f0e142: OUTPUT«Too many positionals passed; expected 1 argument but got 2␤ in regex so at /tmp/4ceGnSSjO9:1␤ in regex TOP at /tmp/4ceGnSSjO9:1␤ in block <unit> at /tmp/4ceGnSSjO9:1␤␤»
FROGGS jdv79: seems like
cygx I've been playing around with macros and implemented assert at github.com/cygx/p6-assert
should I add that to the ecosystem, or are the module names (assert, DEBUG, NDEBUG) too invasive? 20:46
FROGGS hmmm, I guess these names are fine
geekosaur we could do better than names lifted directly from C
FROGGS true 20:47
but I would not call them invasive
colomon as module names? 20:48
20:48 kaare__ left
cygx colomon: yes 20:48
[Coke] well, it's "testneeded", but the test will currently fail if you use the updated one, aye.
(jdv79)
cygx colomon: they intentionally look like pragmas
[Coke] add a fudged failing test and then we can unmark test needed. 20:49
colomon cygx: that seems mildly wrong to me.
I mean, they’re not pragmas, right?
cygx colomon: well, they are not proper modules either 20:50
jdv79 [Coke]: ok, cool.
colomon why not?
cygx DEBUG and NDEBUG just set environment vars and can be called repeatedly
assert exposes a single macro
20:51 dolmen joined
colomon hurm, does “use DEBUG” actually turn it on? 20:51
colomon is thinking this sounds quite cool, even if he’s not 100% convinced by the names
cygx colomon: yes, in contrast to C assert, you need to enable debug mode 20:53
dalek ast: ffe2c59 | lizmat++ | S24-testing/3-output.t:
Update test count
colomon cygx: is it a general framework or just assert? 20:54
cygx colomon: right now,just assert - but DEBUG and NDEBUG can easily be repurposed
20:56 larion left
colomon cygx: feels to me like you’re right on the verge of something awesome. My instinct is the names still need a tweak, but you might do very well to get it in the Ecosystem as is, and figure out the tweaks later. 20:56
dalek kudo/nom: 536c26f | lizmat++ | src/Perl6/ (2 files):
Make 'use cur' / @?INC operational (attempt #2)
 20:57
meisl I've got some questions on QAST::Stmt(s) - anyone? 20:58
20:58 telex left
lizmat note: this does not fix panda, getting less sure that's my fault 20:58
21:00 larion joined, telex joined
timotimo lizmat++ # perseverance 21:01
FROGGS meisl: ask 21:03
meisl well, actually I'm wondering what they are for at all...? (I do get working code if I stick my Ops and even just Vars right under the Block) 21:06
21:07 raiph left
FROGGS meisl: github.com/perl6/nqp/blob/master/d...d-qaststmt 21:10
21:10 larion left 21:12 colomon left 21:13 fhelmberger joined
jdv79 lizmat: thanks 21:14
meisl FROGGS: hm, yes I've read this, or similar elsewhere, and it appears reasonable at 1st sight to explicitly group a nr of statements under Stmt - but... 21:16
21:17 fhelmberger left
meisl FROGGS: ...what else could be inside a Block than a seq of stmts? even more so given that any expression is also a stmt; plus... 21:18
FROGGS meisl: maybe jnthn can explain :o) 21:19
21:19 sftp joined
meisl FROGGS: ...plus that it *does* work *without* the extra Stmt(s) around it, just right under Block (some magic going on that sanitizes my illformed QAST tree?) 21:20
FROGGS hmmm, not sure that this happens
meisl well, I have to admit that I didn't check that (magic?) thoroughly, just judging from my toy examples 21:22
21:23 larion joined 21:26 dolmen left
meisl FROGGS: ok, anyways. maybe you can elucidate "[...] QAST::Stmt [...] can lead to better code [...] if used correct. Incorrect use can, of course, lead to incorrect code" 21:27
lizmat FROGGS jnthn : do you know of a way inside World.nqp to find out the backend we're running under ? 21:28
FROGGS meisl: ahh, QAST::Stmt marks a register allocation boundary, so there is at least one difference
lizmat: gimme a sec 21:29
lizmat: if nqp::getcomp('perl6').backend.name ne 'moar' {
lizmat FROGGS++
meisl FROGGS: yes, it's something about locals but I still don't really get it 21:30
FROGGS yeah, perhaps talk with jnthn about it 21:31
meisl FROGGS: in particular I would really like to know what exactly is *incorrect use*
FROGGS me needs to go to bed now
well, trying to share temporaries across QAST::Stmt's
lizmat hopes FROGGS will get a good night's rest
FROGGS ohh, I will :o) 21:32
gnight #perl6
meisl o yea, thx a lot, FROGGS
21:32 FROGGS left 21:33 agentzh_ joined 21:40 diana_olhovik_ left, agentzh_ left 21:41 colomon joined
cygx what's a good name for a logging function that doesn't clash with the logarithm? 21:41
eli-se logging::log 21:43
colomon still thinks there should be a namespace, just doesn’t know what 21:44
lizmat nsa::log
eli-se Namespaces are one honking great idea.
21:44 brrt left
cygx dbg:: ? 21:47
meisl cygx: log::log 21:53
timotimo how about łog? 21:55
kurahaupo shouldn't the function name for "logarithm" be "ln"? 21:56
21:56 firecat joined
timotimo well, log is usually log10 and ln is usually log_e 21:56
meisl or lg 21:57
timotimo m: say log 10; say ln 10
camelia rakudo-moar 536c26: OUTPUT«5===SORRY!5=== Error while compiling /tmp/pEGbZqU7i1␤Undeclared routine:␤ ln used at line 1. Did you mean 'lc', 'on'?␤␤»
timotimo we don't seem to have ln under that name
but we do have the natural log, right?
m: say log 10; say lg 10
camelia rakudo-moar 536c26: OUTPUT«5===SORRY!5=== Error while compiling /tmp/FwZsYdItH3␤Undeclared routine:␤ lg used at line 1. Did you mean 'lc'?␤␤»
kurahaupo timotimo: it's just a fixed ratio; log_10(x) = ln(x)/ln(10)
timotimo of course
cygx well, at least some mathematicians often denote log_e as log as well 21:58
meisl m: say ln 2.7
camelia rakudo-moar 536c26: OUTPUT«5===SORRY!5=== Error while compiling /tmp/MSXR6OmJHr␤Undeclared routine:␤ ln used at line 1. Did you mean 'lc', 'on'?␤␤»
kurahaupo timotimo: and "length" works as ceil(ln(x)/ln(10))
err, floor
cygx personally, I prefer log for the generic function, lg for base 10 and ln for base e
at least some mathematicians _consistently_
anyway, I went with logger for now (as in `use logger`) 21:59
kurahaupo Maybe log-base-10 should be written Num.length ? :-)
meisl well, you can call it logg, as opposed to a big chunk of wood
22:00 firecat left, firecat joined, RabidGravy left
meisl hehe, `use log` somehow reminds of Monkey Island days 22:03
cygx if someone wants to keep bikeshedding, the code is at github.com/cygx/p6-debug
geekosaur l7m :p
cygx use DEBUG/use NDEBUG now set environment vars in the PERL6_DEBUG_* namespace 22:04
assert works as in C, logger just does a &note
timotimo Pod::To::HTML just OOM'd on my laptop in its t/02-code.t
and it could very well be that it's going to do the same in the 04-lists.t as well 22:05
22:07 lolisa left, BenGoldberg joined 22:08 lolisa joined
timotimo this time i may kill it before it OOM's and gets other things on my machine killed 22:08
jdv79 m: class foo is Any; say foo.^methods # why nothing? 22:09
camelia rakudo-moar 536c26: OUTPUT«␤»
lizmat m: class foo is Any; say foo.^methods(:all)
camelia rakudo-moar 536c26: OUTPUT«Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new Method+{<anon>}.new ACCEPTS Method+{<anon>}.new Method+{…»
22:11 larion left 22:14 bjz left 22:23 spider-mario left, BenGoldberg left
timotimo hmm 22:23
so, what were all the things happening last week
timtoady was at some event, there was the GPW and OSDC.no?
22:23 spider-mario joined
lizmat TimToady was at Collision in Las Vegas 22:24
as was pmichaud
eli-se timotimotoady
lizmat gist.github.com/lizmat/c7f53fa206da900c2b42 # status update @?INC related work
22:24 BenGoldberg joined
lizmat on that note, I wish all of #perl6 a good night! 22:25
&
eli-se goodbye! 22:26
22:26 tony-o_ is now known as tony-o
timotimo lizmat: are the : in the "include spec format" thingie new or outdated? i thought we used # in there? 22:26
lizmat yes, they are outdated...
fixing 22:27
timotimo thank you :)
lizmat fixed (I think) 22:29
afk& 22:30
tony-o m: use MONKEY-TYPING; augment class IO::Handle { augment method read(IO::Handle:D: Int(Cool:D) $bytes) { die $bytes; }; }; $*IN.read(1); die "dye"; 22:34
camelia rakudo-moar 536c26: OUTPUT«dye␤ in block <unit> at /tmp/gjyby1osNF:1␤␤»
22:35 lizmat left
TimToady timotimo: other than Collision, I split off Empty from Nil, and fixed warnings to include file and line number 22:36
22:36 cognominal left, lizmat joined 22:37 cognominal joined, eli-se left
TimToady other than that, mostly de-LTAing messages 22:38
22:44 andreoss joined 22:46 muraiki joined 22:47 firecat left
timotimo right, good :) 22:48
dalek ast: 74bedac | (Justin DeVuyst)++ | S05-grammar/inheritance.t:
Tests for RT #77350.
 22:49
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...l?id=77350
jdv79 idk if that's really a bug but now there's a todo for it in any case 22:54
TimToady that's weird, my laptop, with 100% charge, nothing making heat, just turned off, bang (well, not with any real sound) 23:00
23:00 kurahaupo left
jdv79 cosmic rays 23:01
TimToady you'd think it'd have a more temperamental crash if that were the case
andreoss m: my (Any, Any, $c) = 1...3; say $c
camelia rakudo-moar 536c26: OUTPUT«3␤»
TimToady unless the cosmic ray in question hit the power switch itself...
jdv79 the spec says .assuming is on all Callables but Block seems to lack it - is that correct? 23:02
tony-o TimToady: i had a lenovo that would do that because it thought the cpu fan was failing
TimToady so have I, but then it usually starts carping about FAN FAILURE or so on bootup 23:03
starting a compile, and the fan revs up okay 23:05
meisl jdv79: I stumbled over this as well, and ended up writing my own currying lib 23:09
23:13 Isp-sec left 23:15 raiph joined 23:20 agentzh_ joined 23:24 vytas joined
timotimo procrastinates blog post writing by implementing a possibly worthless optimization .. 23:24
23:25 cognominal left, ][Sno][ joined
tony-o timotimo: i think i found another symptom of that IO::Socket/Async thing where it can't read from the handle 23:26
23:27 spookah joined 23:28 gfldex left
timotimo uh-oh 23:28
tony-o github.com/tony-o/perl6-event-emit...ss.pm6#L49
that will consistently read if the process is the main process but if i use Proc::Async and write to the STDIN of the sub process, it will only read $*IN maybe 1 in 100 times 23:29
excuse the monkey typing, i was experimenting with stuff but wanted to upload it to show
23:30 airdisa joined
timotimo oof, looks like a bit much to look at at once 23:31
can you explain to me how to read the output of cachegrind?
tony-o cachegrind ? 23:32
timotimo i'm used to it only showing a single line regarding "I", which i learned was "instruction count"
but now i have I refs, I1 misses, LLi misses, D refs, D1 misses, LLd misses, LL refs and LL misses
tony-o beyond me :) 23:33
timotimo OK 23:34
well, the numbers went down because of my optimization, so that's probably good
I refs went from 6,918,462,233 to 6,917,415,091 23:38
23:38 muraiki left
timotimo while I1 misses went from 82,161,276 to 80,275,325 23:38
which is a much more drastic change, and i suppose misses are much more interesting than refs are 23:39
tony-o i wonder if tux's csv times will go down now 23:42
23:48 larion joined 23:51 cygx left

Powered by Cro and the Raku® Programming Language.

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

Thank you!