Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2011-06-21

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:03 f00li5h is now known as f00li5h[HireMe]
00:09 slontec joined #perl6
00:25 daniel-s left #perl6
00:25 kborer joined #perl6
00:27 jnthn sleep &
00:44 sorear What's the most idiomatic way to write "{$from}::$_" ?
00:48 sftp joined #perl6
00:48 thou ::($from)<$_>
00:48 thou no idea if that works
00:48 thou :-)
00:49 thou o/ hello!  and, ... \o goodbye, #perl6 :-)
00:49 sftp_ left #perl6
00:49 thou left #perl6
00:51 sorear not what I meant :)
00:52 sorear I want $from ~ "::" ~ $)
00:52 sorear $_
00:57 sorear niecza v7 release on the 27th
00:57 icwiener left #perl6
01:00 cotto_work left #perl6
01:01 sftp_ joined #perl6
01:02 sftp left #perl6
01:05 noganex left #perl6
01:06 noganex joined #perl6
01:10 TimToady you mean more like ::($from)::{$_}, or maybe ::($from)::($_)
01:12 TimToady or ::($from ~ '::' ~ $_)
01:12 sorear No
01:12 sorear I do not want to look up a name
01:12 sorear I just want to construct a string
01:13 TimToady then the way you wrote it first seems pretty okay
01:14 TimToady but I don't know what you mean by 'idiomatic'
01:14 sorear "{$package}::$name" just looks a bit ugly to me
01:15 TimToady perl6: my $from = "abc"; $_ = "def"; say "$from\::$_"
01:15 p6eval pugs, rakudo 248244, niecza v6-177-g365e216: OUTPUT«abc::def␤»
01:15 TimToady that works too
01:17 TimToady but $from ~ '::' ~ $_ ain't all that bad either
01:18 sftp_ left #perl6
01:23 sftp_ joined #perl6
01:23 woosley joined #perl6
01:40 lichtkind TimToady: sorry for being little bit grumpy on that sin thing, i dont take my position to seriously only if other people suffer :)
01:44 lichtkind i mean not too seriously :)
01:47 lichtkind TimToady: other than that i just want to know if there are any changes in the IO sector underway, because that tablet is gona to be wrtten next
01:51 whiteknight left #perl6
01:52 gh joined #perl6
01:53 Chillance joined #perl6
01:56 cognominal_ joined #perl6
02:00 molaf joined #perl6
02:00 d4l3k_ joined #perl6
02:00 cognominal left #perl6
02:00 kst` joined #perl6
02:01 nsh joined #perl6
02:01 beekar joined #perl6
02:02 scottp_ joined #perl6
02:02 miso2217__ joined #perl6
02:03 uniejo_ joined #perl6
02:03 diakopte1 joined #perl6
02:04 orafu left #perl6
02:04 _patch_ joined #perl6
02:04 Gruber joined #perl6
02:04 orafu joined #perl6
02:04 xinming_ joined #perl6
02:08 miso2217_ left #perl6
02:08 kst left #perl6
02:08 xinming left #perl6
02:08 scottp left #perl6
02:08 araujo left #perl6
02:08 dalek left #perl6
02:08 jasonmay left #perl6
02:08 starcoder left #perl6
02:08 hugme left #perl6
02:08 diakopter left #perl6
02:08 uniejo left #perl6
02:08 Helios left #perl6
02:08 Grrrr left #perl6
02:08 _patch left #perl6
02:08 nsh| left #perl6
02:09 d4l3k_ is now known as dalek
02:13 gh left #perl6
02:23 Guest42133 joined #perl6
02:27 stepnem left #perl6
02:27 Guest42133 left #perl6
02:29 envi_laptop joined #perl6
02:30 stepnem joined #perl6
02:34 chitragupt left #perl6
02:36 ab5tract joined #perl6
02:46 cooper left #perl6
02:47 lichtkind left #perl6
02:47 cooper joined #perl6
02:55 tokuhirom left #perl6
02:57 molaf left #perl6
03:02 slavik1 left #perl6
03:03 chitragupt joined #perl6
03:20 dalek nqp-rx: 1dc87d4 | pmichaud++ | src/stage0/ (3 files):
03:20 dalek nqp-rx: Update bootstrap files to take advantage of new PAST::Stmt nodes and
03:20 dalek nqp-rx: better register allocation.
03:20 dalek nqp-rx: review: https://github.com/perl6/nqp-rx/commit/1dc87d48d6
03:23 Su-Shee_ joined #perl6
03:24 pmichaud all parrot tests pass with new nqp-rx bootstrap... building rakudo
03:24 pmichaud oops, wrong chan
03:24 agentzh joined #perl6
03:24 agentzh left #perl6
03:24 agentzh joined #perl6
03:27 Su-Shee left #perl6
03:30 JimmyZ joined #perl6
03:32 jaldhar joined #perl6
03:41 sftp_ left #perl6
04:08 colomon in a rather abrupt change in personal plans, I just booked a flight to YAPC::NA.  I don't like flying, but it turns out I like the idea of driving 12 hours by myself, part of it through mountains (which I hate) even less.  Plus the cost of the flight was almost the same as the gas would have cost.
04:08 colomon afk # bed
04:16 ab5tract left #perl6
04:23 TimToady http://rosettacode.org/wiki/Arbitrary​-precision_integers_(included)#Perl_6
04:24 TimToady uses niecza...didn't mention it takes 2m 17s on my T61...  :)
04:25 TimToady niecza: say 5 ** 262144;   # this is the slow part  :)
04:25 p6eval niecza v6-177-g365e216: OUTPUT«(timeout)»
04:26 satyavvd joined #perl6
04:32 sorear hehe, I can imagine :)
04:32 sorear the ** routine is... old.
04:33 sorear pmurias wrote it, iirc, before niecza even had bitops
04:34 TimToady it could likely be sped up by squaring 5 repeatedly and factoring out large pieces, the doing it again for the remainder
04:35 sorear yeah
04:35 sorear wait.
04:35 sorear 262144 is a power of 2
04:36 TimToady niecza: say 4**3**2
04:36 p6eval niecza v6-177-g365e216: OUTPUT«262144␤»
04:36 TimToady yes
04:36 sorear niecza: say (5, { $^a * $^a } ... *)[18]
04:36 p6eval niecza v6-177-g365e216: OUTPUT«(timeout)»
04:36 sorear niecza: say (5, { $^a * $^a } ... *)[1]
04:36 p6eval niecza v6-177-g365e216: OUTPUT«25␤»
04:36 sorear niecza: say (5, { $^a * $^a } ... *)[8]
04:36 p6eval niecza v6-177-g365e216: OUTPUT«8636168555094444625386351862800399571116​00036443628138502370347016859180316242705797150​75034722882265605472939461496635969950989468319​466936530037770580747746862471103668212890625␤»
04:37 sorear niecza: say (5, { $^a * $^a } ... *)[18].chars
04:37 p6eval niecza v6-177-g365e216: OUTPUT«(timeout)»
04:37 sorear niecza: say (5, { $^a * $^a } ... *)[16].chars
04:37 p6eval niecza v6-177-g365e216: OUTPUT«(timeout)»
04:37 sorear niecza: say (5, { $^a * $^a } ... *)[10].chars
04:37 p6eval niecza v6-177-g365e216: OUTPUT«716␤»
04:37 sorear niecza: say (5, { $^a * $^a } ... *)[13].chars
04:37 p6eval niecza v6-177-g365e216: OUTPUT«5726␤»
04:38 sorear fail
04:42 * sorear decides to forbid class MY::Foo { }
04:43 birdwindupbird joined #perl6
04:52 TimToady hmm, BigInteger has a Pow routine, but it looks like maybe you're not using it
04:56 sorear right
04:57 sorear you want lib/CORE.setting:291
04:57 sorear I don't really understand how ** is supposed to work, typologically
04:58 sorear I kindof like the Haskell solution there...
04:58 sorear I wonder things like whether -2 ** 4e0 should be Complex
04:59 sorear rakudo: say (-2 ** 4e0).WHAT
04:59 p6eval rakudo 248244: OUTPUT«Num()␤»
05:03 TimToady niecza: say 2 ** .5
05:03 p6eval niecza v6-177-g365e216: OUTPUT«2␤»
05:03 TimToady heh
05:05 koban joined #perl6
05:06 sorear but 4e0 is often considered to have an <=epsilon (real) error
05:06 dalek nqp: 6e4f4db | pmichaud++ | / (3 files):
05:06 dalek nqp: Add PAST::Want nodetype, allows selection of a PAST tree based
05:06 dalek nqp: on rtype signature requested by the caller.
05:06 dalek nqp:
05:06 dalek nqp: This will likely move into Parrot's source tree eventually; I've
05:06 dalek nqp: added it here now so we can test and iterate it more directly
05:06 dalek nqp: until it stabilizes.
05:06 dalek nqp: review: https://github.com/perl6/nqp/commit/6e4f4db287
05:08 * sorear tries to figure out how to type lessequals again
05:12 tomize left #perl6
05:12 Trashlord left #perl6
05:12 tomize joined #perl6
05:13 dalek rakudo/nom: d3bd796 | pmichaud++ | / (2 files):
05:13 dalek rakudo/nom: Bump NQP_REVISION to get PAST::Want.  Use PAST::Want on constant
05:13 dalek rakudo/nom: nodes to avoid serializing and unboxing constant int/num operands
05:13 dalek rakudo/nom: going directly to pir:: opcodes.
05:13 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d3bd796843
05:14 sorear What is PAST::Want?
05:15 Trashlord joined #perl6
05:20 birdwindupbird left #perl6
05:26 kaare_ joined #perl6
05:32 gottreu left #perl6
05:43 Chillance left #perl6
05:44 stephanepayrard_ joined #perl6
05:47 cognominal_ left #perl6
05:50 envi_laptop left #perl6
05:50 starcoder joined #perl6
05:55 wtw joined #perl6
06:03 Sarten-X left #perl6
06:03 wamba joined #perl6
06:10 birdwindupbird joined #perl6
06:10 Sarten-X joined #perl6
06:22 Su-Shee_ is now known as Su-Shee
06:24 fhelmberger joined #perl6
06:26 tomize left #perl6
06:26 tomize joined #perl6
06:27 benabik left #perl6
06:37 benabik joined #perl6
07:03 scottp_ is now known as scottp
07:14 bkolera left #perl6
07:17 dukeleto left #perl6
07:17 dukeleto joined #perl6
07:33 araujo joined #perl6
07:34 mj41 joined #perl6
07:39 araujo left #perl6
07:42 daxim joined #perl6
08:00 envi_laptop joined #perl6
08:00 wamba left #perl6
08:00 tzhs joined #perl6
08:05 Gruber is now known as Grrrr
08:05 REPLeffect left #perl6
08:19 moritz sorear: it's probably detecting sink context
08:20 JimmyZ Walk ant?
08:20 REPLeffect joined #perl6
08:26 dakkar joined #perl6
08:26 xinming_ is now known as xinming
08:32 sorear moritz: the pod in the commit explains it
08:32 sorear it selects one of N children based on INSP result type
08:33 sorear to make constants more efficient in native type code
08:33 sorear (grr... efficient native types will set me *so far back*)
08:37 moritz sorear: no, it will set rakudo forth, not niecza back
08:39 Sarten-X left #perl6
08:42 VXZ left #perl6
08:43 mberends joined #perl6
08:44 mberends I recall pmichaud++ was famous for DO NOT WANT want() ;)
08:45 sorear o/ mberends
08:45 mberends o/
08:45 Shozan joined #perl6
08:45 moritz mberends: indeed, but the sink/notsink distinction can actually be made, and should (for performance and eagerness reasons)
08:46 moritz mberends: contrary to want(), which wasn't computable
08:46 sorear I hate sink/notsink because it brazenly violates the "do not pay for what you do no use" rule
08:47 sorear $x = foo($y,$z); # this has to call $x.sink
08:47 sorear er wait
08:47 sorear foo($y,$z); # this has to call .sink
08:48 sorear I can arrange to short-circuit it by testing a flag on the STable, but still
08:48 Sarten-X joined #perl6
08:49 icwiener joined #perl6
08:49 sorear or I could add PL_sawlazylist and only generate .sink calls if lazy lists are being used.  That's sure to elicit a response from TimToady:D
08:55 Sarten-X left #perl6
08:55 s1n left #perl6
09:04 mberends left #perl6
09:05 sorear *out*
09:08 Sarten-X joined #perl6
09:13 saaki left #perl6
09:14 saaki joined #perl6
09:14 dalek rakudo: f234be6 | moritz++ | build/PARROT_REVISION:
09:14 dalek rakudo: bump PARROT_REVISION to 3.5.0
09:14 dalek rakudo: review: https://github.com/rakudo/rakudo/commit/f234be6859
09:16 daniel-s joined #perl6
09:19 dalek rakudo: 7570802 | moritz++ | docs/ChangeLog:
09:19 dalek rakudo: update ChangeLog a bit
09:19 dalek rakudo: review: https://github.com/rakudo/rakudo/commit/757080260f
09:22 toebu joined #perl6
09:36 kanishka joined #perl6
09:45 woosley left #perl6
10:07 agentzh left #perl6
10:18 stepnem left #perl6
10:25 fbass joined #perl6
10:27 stepnem joined #perl6
10:37 araujo joined #perl6
10:37 araujo left #perl6
10:37 araujo joined #perl6
10:40 mkramer1 joined #perl6
10:44 mkramer joined #perl6
10:46 mkramer1 left #perl6
10:56 s1n joined #perl6
11:15 xinming left #perl6
11:16 xinming joined #perl6
11:17 jedai left #perl6
11:19 envi_laptop left #perl6
11:19 avar left #perl6
11:21 avar joined #perl6
11:32 wamba joined #perl6
11:40 pnu left #perl6
11:40 pnu joined #perl6
11:42 satyavvd left #perl6
11:48 cotto left #perl6
11:49 colomon moritz++
12:03 stepnem left #perl6
12:04 stepnem joined #perl6
12:06 Helios joined #perl6
12:06 takadonet morning all
12:06 colomon o/
12:07 moritz \o
12:07 lichtkind joined #perl6
12:08 orafu left #perl6
12:08 orafu joined #perl6
12:11 aloha left #perl6
12:12 yves joined #perl6
12:13 aloha joined #perl6
12:18 daniel-s is it bad to write the opening curly bracket on the next line, rather than as: if (a<b) {
12:18 birdwindupbird left #perl6
12:19 moritz it's much more important to put whitespace around the infix operator
12:19 moritz and you don't need the ( ) either
12:19 moritz so just write   if a < b {   and be happy :-)
12:19 * moritz is great in not actually answering questions :-)
12:20 molaf joined #perl6
12:22 dalek rakudo: 3fc36ba | pmichaud++ | docs/announce/2011.06:
12:22 dalek rakudo: Add a draft 2011.06 release announcement.  (Still needs updates.)
12:22 dalek rakudo: review: https://github.com/rakudo/rakudo/commit/3fc36ba58d
12:31 JimmyZ good evening, #perl6
12:31 pmichaud JimmyZ: o/
12:31 pmichaud good morning, #perl6
12:31 takadonet JimmyZ: o/
12:31 JimmyZ && good morning, pmichaud
12:32 woosley joined #perl6
12:32 moritz pmichaud: do you want to mention the nom branch in the release announcement?
12:32 pmichaud I think it would be good to mention it, yes.
12:32 JimmyZ takadonet: o
12:38 lichtkind moritz: if you would be so kind make a quick check if I did no major BS http://perlpunks.de/paste/show/4e00909e.9d1.4e
12:49 bluescreen10 joined #perl6
12:51 moritz lichtkind: not bullshit, but I'd like to remind you that only names and substantives are capitalized in German
12:51 lichtkind moritz: sounds plausible :)
12:52 PerlJam greetings #perl6lings
12:53 pmichaud PerlJam: o/
12:53 PerlJam This month's release is the "hitchhikers guide to the galaxy" release   :)
12:53 pmichaud yes, I believe that's why jdhore++ claimed it. :)
12:53 pmichaud We should put "Don't Panic" in the announcement.  :-)
12:53 * moritz wonders why lue wasn't faster :-)
12:53 PerlJam indeed
12:54 PerlJam had lue gotten it, there would have been too much HHGTTG
12:54 PerlJam :-)
12:54 mberends joined #perl6
12:54 pmichaud oh, and all-caps.
12:54 moritz INDEED
12:55 mberends \o/ biggish int long division works: 0x906c93c2fa29608fc6f2cad3ef5 / 0x11112222333344445555 == 0x87654321 # will push after $work
12:56 JimmyZ_ joined #perl6
12:58 JimmyZ left #perl6
12:58 JimmyZ_ is now known as JimmyZ
12:59 fbass left #perl6
13:00 PerlJam I wonder if someone on the "outside" is looking at the Rakudo release announcements and wondering what happened to that Worhtington guy ... he hasn't contributed in a while   ;-)
13:00 pmichaud Slacker.
13:02 birdwindupbird joined #perl6
13:03 sftp joined #perl6
13:03 Holy_Cow joined #perl6
13:03 birdwindupbird left #perl6
13:03 birdwindupbird joined #perl6
13:04 smash joined #perl6
13:04 smash hello everyone
13:05 sunnavy left #perl6
13:05 MayDaniel joined #perl6
13:07 takadonet smash: hey
13:08 smash hmm.. i'm a bit lost in these new nqps' (my fault of course)
13:08 mberends left #perl6
13:11 woosley left #perl6
13:12 PerlJam smash: lost how?
13:13 moritz it's quite easy
13:13 moritz there's the old-old NQP. It's dead
13:14 moritz the next one was NQP-rx, on which rakudo master is built
13:14 moritz it ships with parrot
13:14 * smash nods.
13:14 smash nqp-rx was the last one i looked at
13:14 moritz then there's the new nqp, which comes with 6model, and on which rakudo/nom is being built
13:14 koban left #perl6
13:14 sunnavy joined #perl6
13:14 moritz the only really confusing thing is that the nqp-rx executable is called 'parrot-nqp'
13:16 smash moritz: thank you for the summary
13:16 moritz smash: you're welcome
13:33 aidan2011 joined #perl6
13:34 lichtkind left #perl6
13:39 tokuhirom joined #perl6
13:40 daniel-s left #perl6
13:41 daniel-s joined #perl6
13:52 wamba left #perl6
13:58 Shozan left #perl6
14:00 jasonmay joined #perl6
14:06 Holy_Cow left #perl6
14:09 kfo joined #perl6
14:09 mkramer left #perl6
14:13 kfo_ left #perl6
14:14 colomon left #perl6
14:15 Chillance joined #perl6
14:17 wtw left #perl6
14:17 leprevost joined #perl6
14:20 bluescreen10 left #perl6
14:23 jaldhar left #perl6
14:23 jaldhar joined #perl6
14:24 MayDaniel left #perl6
14:27 slavik left #perl6
14:30 mkramer1 joined #perl6
14:32 molaf left #perl6
14:38 daniel-s left #perl6
14:38 daniel-s joined #perl6
14:39 slavik joined #perl6
14:51 thou joined #perl6
14:56 dalek nqp: bcfe1f0 | pmichaud++ | tools/build/PARROT_REVISION:
14:56 dalek nqp: Bump PARROT_REVISION to 3.5.0 release.
14:56 dalek nqp: review: https://github.com/perl6/nqp/commit/bcfe1f009e
14:56 dalek nqp: c62322b | pmichaud++ | / (5 files):
14:56 dalek nqp: Add initial version of nqp:: pseudo-opcode space for low-level primitives.
14:56 dalek nqp: review: https://github.com/perl6/nqp/commit/c62322b88c
14:59 moritz pmichaud: do the control nqp:: opcodes take closures?
15:00 moritz pmichaud: also you wrote that the p6 terms are prefered, but use nqp::upcase instead of nqp::uc
15:03 tokuhirom left #perl6
15:03 tokuhirom joined #perl6
15:05 moritz pmichaud: and we might want an open() opcde (?)
15:06 moritz ah, not closures, just statements
15:07 moritz thunks
15:07 mkramer1 left #perl6
15:08 moritz pmichaud++
15:09 moritz though why do we need nqp::if and nqp::: when we already have the ternary? bootstrapping thingies?
15:20 daniel-s left #perl6
15:21 toebu left #perl6
15:22 Tedd1^ left #perl6
15:22 wamba joined #perl6
15:28 donri joined #perl6
15:28 kaare_ left #perl6
15:36 strider__ joined #perl6
15:36 strider__ left #perl6
15:36 strider__ joined #perl6
15:37 tzhs left #perl6
15:38 Tedd1 joined #perl6
15:38 cotto joined #perl6
15:39 strider__ left #perl6
15:39 stormcrow joined #perl6
15:47 stormcrow left #perl6
15:47 strider__ joined #perl6
15:50 strider__ left #perl6
15:50 strider__ joined #perl6
15:53 pmichaud moritz: yes, bootstrapping.
15:54 pmichaud eventually ?? !! will be defined in terms of nqp::if() .
15:54 pmichaud Agreed on "uc" versus "upcase" -- fixing.
15:54 slontec left #perl6
15:54 moritz pmichaud: but isn't control flow usually bootstrapped by the code emitting part?
15:55 pmichaud nqp::if() is just another way of saying  PAST::Op(:pasttype<if> ... )
15:55 moritz I mean, the PAST if-block takes care of that
15:55 moritz ok
15:55 pmichaud all of the nqp:: "opcodes" are really just ways of requesting specific PAST nodes
15:55 pmichaud it even occurred to me earlier that we can do  nqp::stmts(...)   and create PAST::Stmt blocks
15:56 pmichaud i.e., eventually it would be possible to write all of nqp using nqp:: ops :)
15:56 pmichaud (well, maybe not all, but we can get darn close)
15:57 pmichaud what nqp:: really does is give us a functional syntax to map to the capabilities of whatever virtual machine we happen to be on
15:57 pmichaud (as exposed by PAST)
15:57 pmichaud eventually PAST will be written in nqp, at which point we'll be very self-hosted
15:58 cotto left #perl6
15:58 pmichaud About 20 minutes ago I was thinking of nqp::parrot(...), which performs an operation on Parrot but is a no-op everywhere else.
15:58 moritz might be useful, might be dangerous
15:59 pmichaud well, I was thinking of the case of something like the 'compreg' opcode.
15:59 pmichaud Eventually HLL::Compiler will have its own "lookup a compiler" equivalent, that can be independent of compreg
16:00 pmichaud i.e., it'll keep its own table of compilers
16:00 pmichaud but if we're running on parrot, we'd like to go ahead and register the compiler via compreg anyway
16:01 pmichaud so:   nqp::parrot(pirop=>'compreg', ...args...)   can do that for us.  It lets us execute a parrot opcode, but gets ignored by non-parrot vms.
16:01 Yaakov joined #perl6
16:01 pmichaud anyway, yes, there's a lot of new capability here to be explored :)
16:02 pmichaud I'm eager to start converting nom to using nqp:: instead of pir:: where we can
16:03 moritz shouldn't be too hard, once you've bootstrapped
16:03 Yaakov left #perl6
16:03 pmichaud oh, I can add nqp:: syntax to nom now :)
16:03 pmichaud no bootstrap necessary, at least for the CORE.setting part
16:03 moritz ah, I misread
16:03 moritz I thought you meant nqp
16:04 moritz for nqp you'd need bootstrapping
16:04 pmichaud nqp already can start using nqp::
16:04 pmichaud that's how it's able to test it :)
16:04 pmichaud (nqp:: is already built into nqp.)
16:04 moritz but not in the stage1 compiler, is it?
16:04 pmichaud rakudo uses the stage2 compiler
16:05 pmichaud anyway, "make bootstrap-files"  will put it into stage 0 :)
16:05 moritz but you couldn't use nqp:: in NQP's Action.pm unless you bootstrap, right?
16:05 * moritz just wants to check his sanity
16:05 pmichaud right
16:05 pmichaud s/unless/until, at this point.
16:05 moritz right
16:07 pmichaud doing that now :)
16:09 pmichaud the opcodes I'm not entirely happy with are nqp::bindattribute and nqp::getattribute.  They feel "long", and I'm not fond of "get" either.  I'm thinking of just  nqp::bindattr and nqp::getattr
16:09 pmichaud or even just nqp::attr instead of nqp::getattr
16:10 PerlJam whenever I see "get...", I tend to expect there to be a corresponding "set..."
16:10 PerlJam (just a data point)
16:10 mj41 left #perl6
16:10 strider__ left #perl6
16:10 strider__ joined #perl6
16:11 pmichaud PerlJam: I'm the same way.  I'm still stung by Parrot's over-misuse of that term.
16:11 MayDaniel joined #perl6
16:12 pmichaud and I think I'd be much more explicit by using "bind*" instead of "set*", so that means "get*" should probably disappear in favor of something else
16:12 dalek nqp: 3bb5b1c | pmichaud++ | / (3 files):
16:12 dalek nqp: nqp::downcase/nqp::upcase should be nqp::lc/nqp::uc.  moritz++
16:12 dalek nqp: review: https://github.com/perl6/nqp/commit/3bb5b1cbac
16:12 dalek nqp: 900cf14 | pmichaud++ | src/stage0/ (6 files):
16:12 dalek nqp: Re-bootstrap to enable nqp:: ops within NQP itself.
16:12 dalek nqp: review: https://github.com/perl6/nqp/commit/900cf14206
16:15 spq1 joined #perl6
16:16 strider__ left #perl6
16:16 strider__ joined #perl6
16:17 strider__ left #perl6
16:19 * PerlJam suffers a failure of imagination on replacement candidates for "get*"
16:20 REPLeffect left #perl6
16:20 dukeleto left #perl6
16:21 dukeleto joined #perl6
16:22 pmichaud "fetch"  is the best I've come up with thus far
16:23 Trashlord left #perl6
16:23 pmichaud "grab"
16:23 PerlJam "reap" might work if you're so inclined  (it certainly won't have any cognitive overload in this context ;)
16:23 pmichaud "reap" implies to delete or remove, to me.
16:24 pmichaud pmichaud@kiwi:~/nom$ ./perl6
16:24 pmichaud > nqp::say(nqp::add_i(2,3))
16:24 pmichaud 5
16:24 pmichaud \o/
16:25 JimmyZ \o we open the door to other vms :)
16:25 JimmyZ not we, I mean pmichaud++
16:25 pmichaud it's a collective effort, certainly.
16:26 JimmyZ though I don't like C# and Java :)
16:26 Trashlord joined #perl6
16:26 pmichaud well, if this goes well, then you'll be able to use Perl on the CLR without having to deal with C# :-)
16:27 PerlJam "grab" works for me (I tend to favor something with 4 letters to match "bind" in the absence of some mnemonic device)
16:27 * JimmyZ never learns   commercial language
16:27 pmichaud I'll brainstorm a bit more.
16:27 JimmyZ I would like fetch
16:31 PerlJam so will there be 3 variants or just 2?
16:31 pmichaud ?
16:32 PerlJam bind, fetch, and "change the thing bound" ?
16:32 PerlJam or just bind?
16:32 pmichaud bind, fetch, and maybe store
16:32 pmichaud where store means "put into the thing"
16:32 PerlJam ok
16:32 pmichaud i.e., we bind a lexical symbol, and we can store into a container
16:32 pmichaud (also following some of the p6 terminology here :)
16:34 PerlJam You can't go too far wrong following the lead of a linguist when doing languagey things  ;)
16:35 pmichaud ...how does this look to folks?   http://gist.github.com/1038259
16:36 PerlJam less verbose than before.
16:37 PerlJam stupid question:  We can't have ops that operate on boxed ints?
16:37 JimmyZ maladjustment
16:38 pmichaud in parrot, that would mean going through the vtables
16:38 pmichaud which would mean creating a vtable for every type
16:40 pmichaud which currently means that you either write that code in C (not portable), or you suffer an absolutely gigantonormous performance penalty as you jump through inferior runloops to get to the pir that implements your vtable
16:41 aidan2011 left #perl6
16:41 pmichaud if you mean "can we create ops that automatically unbox the ints for us"... we might be able to do that, but jnthn++ feels pretty strongly at this point that we should keep coercions a bit more explicit (I agree with him for now)
16:41 pmichaud s/ops/pseudo-ops/
16:42 pmichaud s/coercions/(un)boxing/
16:45 PerlJam I was thinking ahead a little to the mythical VM that caters to Rakudo's needs and wondering how this code would change to run on both that VM (or a VM like it) and Parrot
16:45 pmichaud I think the code might be the same -- on that VM the unbox_i opcodes would be no-ops (and thus optimize out)
16:46 cdarroch joined #perl6
16:46 cdarroch left #perl6
16:46 cdarroch joined #perl6
16:52 dalek rakudo/nom: 2089c06 | pmichaud++ | / (3 files):
16:52 dalek rakudo/nom: Bump NQP_REVISION, add nqp::ops.
16:52 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/2089c0612a
16:52 dalek rakudo/nom: b513e34 | pmichaud++ | src/ (2 files):
16:52 dalek rakudo/nom: Proof of concept of Int arithmetic ops using nqp:: opcodes instead of pir:: .  Win.
16:52 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b513e348fd
16:53 kst` left #perl6
16:54 pmichaud I *like* this layer of abstraction.  :)
16:55 pmichaud afk, lunch
16:58 kst joined #perl6
16:58 tokuhirom left #perl6
16:59 cotto joined #perl6
16:59 birdwindupbird left #perl6
17:04 daxim left #perl6
17:05 dakkar left #perl6
17:09 dual joined #perl6
17:14 Chillance left #perl6
17:16 Chillance joined #perl6
17:18 Sarten-X left #perl6
17:18 jdhore1 pmichaud, I could've done the announcement :(
17:19 PerlJam jdhore1: a) it's not finished and b) why the :( ?  this is what open source and community is about.  Everyone does a little bit to make the whole better.
17:19 jdhore1 True, but I sort of like writing them
17:20 PerlJam you've got the more important and difficult task though ... choosing a name :)
17:20 jdhore1 I already have that
17:21 y3llow left #perl6
17:21 jdhore1 Since it's release #42, i'm going to be making all sorts of Hitchiker's Guide references. The release name is one of them.
17:21 y3llow joined #perl6
17:25 Sarten-X joined #perl6
17:26 smash left #perl6
17:26 kanishka left #perl6
17:28 spetrea sorear: hey
17:28 phenny spetrea: 19 Jun 18:14Z <sorear> tell spetrea to check out niecza if performance is a major concern
17:29 spetrea sorear: I always wonder around, write Perl5 a lot at work and at home, but I come here from time to time because I want to know the future
17:29 PerlJam so, we have a higher concentration of the future here than other places?
17:30 JimmyZ maybe we can change the future here
17:30 JimmyZ error, not me
17:30 JimmyZ you guys
17:31 PerlJam we can change other people's future, just not our own
17:32 pyrimidine left #perl6
17:38 cotto left #perl6
17:45 pmichaud jdhore1: I only created an initial draft.
17:45 pmichaud jdhore1: you're welcome to change it however you want.  :)
17:46 jdhore1 :D
17:46 pmichaud I do have one more change to make
17:46 pmichaud (with a hitchhiker's reference :)
17:48 JimmyZ left #perl6
17:50 jdhore1 :D
17:51 jdhore1 I'm going so obscure with the releasename that even i didn't get the reference till someone explained it to me...but it's perfect.
17:52 cotto joined #perl6
17:52 Chillance left #perl6
18:02 dalek rakudo: 7175e54 | pmichaud++ | docs/announce/2011.06:
18:02 dalek rakudo: Add a DON'T PANIC note referencing "nom".
18:02 dalek rakudo: review: https://github.com/rakudo/rakudo/commit/7175e54022
18:03 PerlJam pmichaud: awww, you didn't like my "contribution"?  :)
18:04 cotto left #perl6
18:06 pmichaud PerlJam: I missed it, I guess
18:07 PerlJam I only added my name to the release doc to claim the Oct release  :)
18:07 PerlJam so that lumps me in with "and others" I guess :)
18:07 pmichaud oh, that
18:08 pmichaud You're welcome to edit the announcement and add yourself if you want.  I think that's worth a mention :)
18:08 masak joined #perl6
18:08 masak oh hai #perl6
18:08 pmichaud masak: oh hai!
18:09 moritz \o
18:09 PerlJam masak: greets!
18:09 * masak is having laptop trouble
18:10 masak worse, I gave it to a computer repairman yesterday, but I had to take it back because it turns out I can't live without it... :/
18:10 masak but enough about me. how are you guys doing?
18:10 pmichaud I'm preparing my #phasers report
18:12 masak ooh, thanks for the reminder! :)
18:19 kboga joined #perl6
18:33 lichtkind joined #perl6
18:34 Tene masak: I've got a spare laptop laying around; obviously you should come hang out at my place.
18:37 pmichaud I've got two or three.  Of course, some of them are woefully underpowered.  :)
18:38 masak I can use this one, but only with a USB keyboard plugged in.
18:38 masak Tene: thanks for the offer. I wish it were easier to realize :)
18:39 pmichaud I even have a spare usb wireless keyboard or two :)
18:40 masak I have a model M one, but I'll need to buy a smaller one before going to FPW.
18:40 pmichaud bbi20
18:40 colomon joined #perl6
18:40 dukeleto left #perl6
18:41 dukeleto joined #perl6
18:41 colomon sorear: I'm going to take a stab at properly implementing infix:<**> on the drive back up.  (Unless I successfully manage to download a lot of useful $work stuff from this cafe in the next five minutes.)
18:49 kborer left #perl6
18:50 sorear good * #perl6
18:51 cotto joined #perl6
18:52 mberends joined #perl6
18:53 Chillance joined #perl6
18:54 colomon left #perl6
18:55 colomon joined #perl6
18:57 colomon left #perl6
18:59 jaldhar left #perl6
19:03 sorear spetrea: o/
19:08 Woodi left #perl6
19:09 sivoais left #perl6
19:09 Woodi joined #perl6
19:25 bakedb joined #perl6
19:27 moritz std: $::a
19:27 p6eval std 37a0cdd: OUTPUT«ok 00:01 118m␤»
19:29 masak there's an OO pattern called Memento of which I've learned recently.
19:29 masak it uses an inner class to do serialization without breaking encapsulation/data hiding.
19:30 robinsmidsrod left #perl6
19:30 masak it's common in Java and C#, it seems.
19:30 moritz and I thought it involved reversed time (like the movie)
19:30 masak :P
19:30 masak I was wondering if the same could be done in Perl 6, and I think it can; if the inner class trusts the outer.
19:31 masak and then I got to thinking -- maybe inner (physically nested) classes should trust outer ones by default.
19:31 masak s/\./?/
19:35 masak std: $::"foo"
19:35 p6eval std 37a0cdd: OUTPUT«[31m===[0mSORRY![31m===[0m␤Two terms in a row at /tmp/cz8DfJ2CCB line 1:␤------> [32m$::[33m⏏[31m"foo"[0m␤    expecting any of:␤    POST␤   bracketed infix␤        infix or meta-infix␤    name␤   postfix␤        postfix_prefix_meta_operator␤   statement modifier
19:35 p6eval ..loop␤Parse failed␤FAILED 00:01 119m␤»
19:36 sorear std: $::("foo")
19:36 p6eval std 37a0cdd: OUTPUT«ok 00:01 119m␤»
19:37 masak std: $::($::($::("foo")))
19:37 p6eval std 37a0cdd: OUTPUT«ok 00:01 119m␤»
19:38 birdwindupbird joined #perl6
19:38 pmichaud blog post draft -- comments welcome:  http://pmichaud.com/sandbox/lists.html
19:41 moritz nice one
19:41 sorear TimToady: 'no strict; $x = 5; my $x; say $x' # What should this do in Perl 6?
19:41 pmichaud feels a little airy, but maybe I can add more down-to-earth stuff in a future post
19:43 masak pmichaud: reading that draft, I'm reminded of what I like about the Perl 6 design process: we're all building towards Perl 6, but we don't exactly know what it is yet. but "we know it when we see it" -- we know what Perl behaves like, and we often realize we have a number of expectations on Perl 6 too.
19:43 masak how list laziness interacts with arrays is one example.
19:43 masak it feels like it "falls out" of our expectations and our vague notions of what "Perlish" behavior is.
19:43 robinsmidsrod joined #perl6
19:44 Raynes left #perl6
19:44 pmichaud in this case it's more like "here's how we think it works, nope -- that's definitely not perlish, let's try this inste.... oops, that doesn't fit either, how about ..."
19:44 PerlJam I like the paragraph with the analogy of quantum physics and chemists
19:44 mberends pmichaud: a few short code examples (you're very good at those) would reduce the airiness
19:44 masak once I designed a chess variant. it made me realize that I have a lot of vague notions of "Chessish" behavior that made the variant "design itself" in spots. :)
19:45 ingyfoo joined #perl6
19:45 pmichaud mberends: I'm trying to decide if it's better to put this up now and concrete examples later, or to have this wait for the concrete examples.  It might be a day or two before I can get the examples in, which makes me think "post now and again later" may be the better approach.
19:45 Raynes joined #perl6
19:45 Raynes left #perl6
19:45 Raynes joined #perl6
19:45 tadzik pmichaud: I'm sligthly confused by the post. It starts with "what's up in nom", and then describes what are the two branches about
19:45 sunnavy left #perl6
19:46 ingy left #perl6
19:48 pmichaud tadzik: hmm, the post is really more about the state of Perl 6 list implementations :)
19:49 tadzik oh, ok :)
19:50 TimToady sorear: obviously it should say "'no strict' not allowed in strict mode"  :)
19:50 PerlJam pmichaud: will S07 mention all of those listy things you talk about in this post?
19:50 pmichaud PerlJam: definitely.
19:50 pmichaud it pretty much has to
19:51 TimToady sorear: seriously, rather than just throwing variables into the current package like P5, we should consider whether assuming an implicit 'our' makes more sense
19:51 pmichaud they all relate pretty tightly --- indeed, most of them were developed in response to "hey, lists don't work"
19:52 Raynes left #perl6
19:52 Raynes joined #perl6
19:53 PerlJam The only one I don't really have a firm handle on are LoLs.  I'd like to see a blog post about them at some point if you're taking suggestions
19:53 pmichaud Yes, I think that's a blog post on its own.  :)
19:53 sunnavy joined #perl6
19:53 pmichaud And I can write that post as soon as I have a handle on LoLs.
19:53 masak PerlJam: I think LoLs are inherently sludgy, not just for you.
19:53 pmichaud Which will probably happen just about after the time I finish the first draft implementation of them :)
19:53 thou pmichaud: i'm looking forward to your S07 rewrite; there's some stuff in S02 that is hard to understand without knowing the difference between LoL context, list context, list vs Array vs. Seq, etc., and I've still not seen a good description of how all those terms relate
19:54 pmichaud thou: as masak++ said, some of them are still fuzzy -- they're waiting for an implementation to sort them out.  nom is likely to be that implementation that does the first batch of sorting :)
19:54 masak how do LoLs and captures interact? &zip takes a LoL as an argument, doesn't it?
19:54 thou yeah, it's exciting. jnthn++ pmichaud++ !
19:54 PerlJam thou: it's easy ... all you have to do is hang out on #perl6 and participate in every discussion relating to lists and then pay attention to what the collective has decided does and does not work.   ;-)
19:55 thou hehe
19:55 thou working on it.
19:55 pmichaud zip takes a capture, its signature binds the capture into a LoL
19:55 masak PerlJam: I wish that weren't so true... :)
19:55 cbk joined #perl6
19:55 masak pmichaud: yeah, I guess that is pretty easy, actually.
19:55 pmichaud well, there's still some trickiness in that we have to make sure that &infix:<;> dtrt  :-)
19:56 pmichaud er,  semicolon in arglists, too
19:56 thou and the relationship of get() and getarg() may be a good way to explain things, too -- dive down to that lower level with an example or two
19:56 masak pmichaud: did you catch TimToady's comment about ; not being an infix?
19:56 pmichaud there is no get() and getarg() in the new implement
19:56 thou oh, well, nevermind then
19:56 thou hehhe
19:56 pmichaud masak: yes, thus my correction.  it's an argument separator
19:56 pmichaud as well as a statement terminator
19:57 masak pmichaud: it's only the latter, but the constant folder catches it.
19:57 masak IIUC.
19:57 pmichaud token semiarglist {
19:57 pmichaud <arglist> ** ';'
19:57 pmichaud <.ws>
19:57 pmichaud }
19:57 pmichaud (from STD.pm6)
19:57 masak oh, ok.
19:58 PerlJam The last sentence of the 5th paragraph may come back to haunt you.  "...while Perl 6 programmers can do a lot of useful chemistry without having to be aware of the gory details underneath."    I'm *so* reminded of leaky abstractions there.
19:59 PerlJam As long as they can be gradually aware of the gory details thought, everything is fine.
19:59 PerlJam s/ght/gh/
19:59 pmichaud PerlJam: some chemists do need to know about the quantum mechanics underneath, yes.  :)
19:59 masak it's hard to get a precise understanding of it without going into the gory details, probably.
19:59 pmichaud I didn't say you can do all chemistry without having to be aware of the details, just that you can do a lot.
20:00 TimToady y'all make it sound like chemistry isn't Compleatly Gory in its own right...
20:00 pmichaud anyway, it certainly wouldn't be the first statement that comes back to haunt me.  :-)
20:01 pmichaud I think one of my first mis-statements was:  "Perl 6 compiler pumpking?  Sure, why not?"  :-P
20:01 * masak hugs pmichaud
20:01 TimToady that wasn't a statement; it was a question
20:01 pmichaud see!  that's the sort of haunting I get!  :-P
20:01 masak mis-question... :)
20:01 tadzik it's just that no one answered
20:02 masak no-one knows the answer.
20:02 TimToady well, the Spirit of Christmas Future answered by pointing at your grave..
20:02 pmichaud and if I understand correctly, it's really an "interrogatory statement".
20:02 masak it's really an interrogatory statement?
20:02 masak :P
20:02 PerlJam heh
20:03 TimToady it was a derogatory instatement...
20:03 * pmichaud looks over the precipice of having any sort of discussion with TimToady++ on the meaning of words and language and decides to back away slowly.
20:04 TimToady .oO(if you can't confuse 'em...er...confuse 'em...)
20:04 masak lingrist -- implementor: 1 -- 0 ;)
20:05 pmichaud I'm going back to writing code, where I don't have to deal with TimToady's ideas on langua.... er, wait.
20:05 TimToady you can rewrite the rest of the Synopses too :)
20:05 * PerlJam notes that TimToady is haunted as well  ;)
20:06 TimToady well, a haunted house still needs Walls
20:06 masak *groan*
20:06 mberends with broken Windows?
20:07 masak that's most houses these days...
20:07 PerlJam TimToady: speaking of which!   How come http://www.wall.org/~larry/natural.html is disappeared?
20:09 PerlJam in fact, many (if not all) of the wall.org/~larry/*.html documents that I think should be there are actually not.
20:10 PerlJam pm.html -- gone.   perl.html --- gone.
20:11 pochi joined #perl6
20:12 TimToady still recovering from a headcrash several months ago, and using it as an excuse not to carry over stuff that isn't neede anymore
20:12 TimToady but haven't copied over everything that I do want to carry over yet
20:13 PerlJam bummer
20:13 TimToady but natural is there now
20:15 PerlJam makes it easier to spread the good word if it's there for people to read  :)
20:16 Su-Shee PerlJam: let me guess: you never throw away stuff and everyone can see for how long by looking into your garage/cellar/etc? :)
20:17 PerlJam Su-Shee: I have been conditioned to occasinally throw stuff away.  I recently threw away my high school physics notes even!
20:17 Su-Shee .oO(I know how old he is.. ;)
20:19 whiteknight joined #perl6
20:19 PerlJam Su-Shee: my wife and I go good together ... she's always pressuring to throw away old stuff, I'm always pressuring to keep it ... at some point we arrive at an equilibrium where we keep stuff for as long as needed but it eventually migrates to elsewhere  (other people, junk yard, good will, whatever)
20:21 PerlJam though there have been some snafus  (my fishing rods were sold in a garage sale without my knowledge and once a pair of shoes I'd only had a few weeks were given to good will)
20:21 mj41 joined #perl6
20:21 Su-Shee PerlJam: ah. yes. we have the same roles here. we solved the problem by mr shee has a room for him, I have one for me. guess which one has 7 monitors and 3 printers and 5 database servers and shelves of steel?
20:27 * sorear doesn't know much about oraperl and sybperl
20:29 jedai joined #perl6
20:33 Woodi left #perl6
20:38 bakedb left #perl6
20:39 masak sorear: aren't those pre-p5 "distributions" which threatened to fracture the codebase?
20:39 masak Perl 5 solved the original issue by having a module system.
20:41 Sarten-X left #perl6
20:46 Woodi joined #perl6
20:46 diakopte1 is now known as diakopter
20:47 mberends o/ diakopter
20:47 diakopter o
20:47 phenny diakopter: 11 Jun 20:45Z <moritz> ask diakopter any idea what's up with the appflux server? I can't seem to ssh to it, and dalek often times out
20:47 diakopter wow that was awhile ago
20:47 moritz \o diakopter. Ignore that, issue is long since gone
20:48 diakopter I wonder how long I was diakopte1
20:48 impious joined #perl6
20:49 diakopter I recall helping maintain an oraperl
20:49 diakopter *installation
20:51 diakopter TimToady: a haunted house can have a genius or two, too
20:52 Sarten-X joined #perl6
20:52 birdwindupbird left #perl6
21:02 colomon joined #perl6
21:03 leprevost left #perl6
21:03 MayDaniel left #perl6
21:15 wamba left #perl6
21:27 masak diakopter: ah, Maniac Mansion... :)
21:28 masak good times.
21:37 donri left #perl6
21:37 cotto left #perl6
21:40 donri joined #perl6
21:40 mj41 left #perl6
21:42 sorear diakopter: you remember oraperl?  how did that work?
21:44 * sorear is interested in the history of it
21:44 Patterner left #perl6
21:45 sorear pmichaud: I hear you want a 'rakudo-nom:' p6eval target
21:46 benabik nom: say 'hi'; # I thought one got installed.
21:46 p6eval nom: OUTPUT«hi␤»
21:46 sorear benabik: that was supposed to be my line :(
21:46 * benabik didn't mean to spoil it.  :-(
21:46 Psyche^ joined #perl6
21:46 Psyche^ is now known as Patterner
21:47 MayDaniel joined #perl6
21:48 PerlJam sorear: as I recall, oraperl just baked in some custom functions that talked to an oracle database. The functions were just there like the rest of perlfunc.  But that's about all I know.
21:48 PerlJam I think I compiled it once to see what all of the fuss was about, but I never used it.
21:48 masak http://search.cpan.org/~timb​/DBD-Oracle-1.16/Oraperl.pm
21:49 masak those seem to be the functions.
21:49 pmichaud sorear: what I'm looking for is a way to manually prod p6eval into rebuilding rakudo-nom right away instead of waiting for the next update
21:49 pmichaud the next timed update
21:50 jferrero joined #perl6
21:51 pmichaud I already know about "nom: " on p6eval.
21:53 masak pmichaud: ...while avoiding race conditions and p6eval locking up...
21:53 pmichaud if it's hard or can't be done, then I can live with the current system, yes.
21:54 pmichaud I figured it was an easyish sort of thing to do.  If not, no biggie.
21:54 masak maybe I'm just being overly pessimistic.
21:54 pmichaud I know that I see other people restart it from time to time.
21:55 pmichaud (and it seems to get the update when they do their magic)
21:57 PerlJam it's too bad there isn't some sort of smart queue that you can just inject a "rebuild nom" message or even "build nom SHA1"
21:57 masak blog post! http://strangelyconsistent.o​rg/blog/june-20-2011-hangman
21:59 dalek rakudo/nom: 047ab6a | pmichaud++ | src/ (3 files):
21:59 dalek rakudo/nom: More nqp:: updates.  For now I've put some 'generic' nqp:: ops
21:59 dalek rakudo/nom: into Perl6/Actions.pm for testing, when they stabilize they can
21:59 dalek rakudo/nom: migrate to the nqp repo.
21:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/047ab6af06
22:00 tylercurtis left #perl6
22:00 tylercurtis joined #perl6
22:02 sorear hmm.  might be doable
22:02 sorear although you'd have to wait a bit
22:02 sorear and you wouldn't know how long to wait, without login access
22:05 lichtkind thou: the article is ready if you want to look at it
22:06 * sorear is looking for good shell-accessible mutual exclusion programs
22:11 felher masak++ for blog post
22:13 sorear 6model: hi
22:14 alester joined #perl6
22:14 thou hi, lichtkind. ok, sure
22:15 lue hello world! o/
22:15 * sorear would like to at least try :)
22:15 sorear hello lue!!
22:16 daniel-s joined #perl6
22:16 masak hi lue!
22:17 masak 'night, #perl6
22:18 masak left #perl6
22:20 [particle]1 joined #perl6
22:22 diakopter sorear: too long ago; a very distant memory
22:22 diakopter I touched only one script I think
22:22 [particle] left #perl6
22:23 PacoLinux left #perl6
22:25 MayDaniel left #perl6
22:29 cotto joined #perl6
22:34 sorear evalbot control restart
22:34 p6eval left #perl6
22:34 p6eval joined #perl6
22:39 p6eval left #perl6
22:40 p6eval joined #perl6
22:41 sorear niecza: say 2 + 2 # pmichaud
22:41 p6eval niecza v6-177-g365e216: OUTPUT«4␤»
22:41 sorear evalbot rebuild niecza
22:41 p6eval OK (started asyncronously)
22:41 sorear niecza: say 2 + 2
22:41 p6eval niecza v6-177-g365e216: OUTPUT«Rebuild in progress␤»
22:41 sorear niecza: say 2 + 2
22:41 p6eval niecza v6-177-g365e216: OUTPUT«4␤»
22:42 pmichaud +1
22:42 pmichaud +2
22:42 benabik sorear++
22:46 kboga pmichaud: is this helpful? http://pastebin.com/mNwkfhTz
22:49 sorear moritz: is there anything special I have to do to push from the evalbot machine?
23:03 p6eval left #perl6
23:03 p6eval joined #perl6
23:04 daniel-s left #perl6
23:06 dalek evalbot: 5e15117 | sorear++ | / (4 files):
23:06 dalek evalbot: Allow issuing restart commands, and do a bit of mutual exclusion locking
23:06 dalek evalbot: review: https://github.com/perl6/evalbot/commit/5e151174e6
23:06 dalek evalbot: 20fc7c7 | sorear++ | lib/EvalbotExecuter.pm:
23:06 dalek evalbot: Drop max_output_len a bit to stop Bot::BasicBot from making orphans
23:06 dalek evalbot: review: https://github.com/perl6/evalbot/commit/20fc7c7201
23:07 pmichaud kboga:  cool, thanks for the patch!  I'll review and apply shortly.
23:07 kboga also, i think commit 047ab6af0681a427d739 removed the Num's modulo infix operator
23:08 pmichaud kboga: it appears part of the patch may be missing at the end?
23:08 kboga hmm, let me see
23:10 Tene left #perl6
23:10 kboga resubmission: http://pastebin.com/SavXvfVU
23:13 TimToady sorear: if you happen to notice in passing where the (timeout) overwrites the output...
23:15 sorear TimToady: I'm fixing that now
23:16 sorear mildew: say 2 + 2
23:16 p6eval mildew: OUTPUT«sh: /home/mildew/perl5/perlbrew/bin/perl: not found␤»
23:16 sorear is mildew likely to ever work again or should I just axe it now?
23:17 pmichaud reviewing...
23:17 whiteknight left #perl6
23:19 gfldex left #perl6
23:19 gfldex joined #perl6
23:20 pmichaud compiling
23:21 p6eval left #perl6
23:21 p6eval joined #perl6
23:23 cdarroch left #perl6
23:23 TimToady sorear: I'd say ax it
23:24 pmichaud kboga: applied, thanks!  kboga++
23:25 kboga pmichaud: np, thanks for reviewing
23:25 pmichaud afk, fetching dinner
23:27 dalek rakudo/nom: 4272fe4 | pmichaud++ | src/core/ (13 files):
23:27 dalek rakudo/nom: Patches to convert many pir:: ops into nqp:: ops.  Patch courtesy kboga++ via IRC.
23:27 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4272fe4bdd
23:30 sorear hehe...
23:30 sorear rakudo: say "hi"; 1 while 1
23:30 p6eval rakudo 7175e5: OUTPUT«hi␤(timeout)»
23:30 sorear fixed? well...
23:31 sorear rakudo: say "$_ " for 1..*
23:31 p6eval rakudo 7175e5: OUTPUT«1 ␤2 ␤3 ␤4 ␤5 ␤6 ␤7 ␤8 ␤9 ␤10 ␤11 ␤12 ␤13 ␤14 ␤15 ␤16 ␤17 ␤18 ␤19 ␤20 ␤21 ␤22 ␤23 ␤24 ␤25 ␤26 ␤27 ␤28 ␤29 ␤30 ␤31 ␤32 ␤33 ␤34 ␤35 ␤36 ␤37 ␤38 ␤39 ␤40 ␤41 ␤42 ␤43 ␤44 ␤45 ␤46 ␤47 ␤48 ␤49 ␤50 ␤51 ␤52 ␤53 ␤54 ␤55 ␤56 ␤57 ␤58 ␤59 ␤60 ␤61 ␤62 ␤63 ␤64 ␤65 ␤66 ␤67 ␤68 ␤69 ␤
23:31 sorear I don't think cutting off (timeout) is the most useful behavior.
23:32 tokuhirom joined #perl6
23:36 whiteknight joined #perl6
23:36 lichtkind thou: got it?
23:37 thou looking...
23:37 lichtkind thanks :)
23:39 kboga oh typo in 047ab6af0681a427d739: pow_n        => 'mod__Nnn', (patch coming up)
23:39 TimToady sorear: nevertheless, it's more useful than overwriting the front
23:40 p6eval left #perl6
23:40 TimToady esp when you already know it's gotta be infinite
23:40 p6eval joined #perl6
23:41 TimToady maybe s/OUTPUT/TIMEOUT/ in that case
23:41 TimToady TIMEOUTPUT«1 ␤2 ␤3 ␤4 ␤5 ␤6 ␤     :)
23:43 p6eval left #perl6
23:43 p6eval joined #perl6
23:43 kboga small patch to fix some typo and the missing % operator: http://pastebin.com/Gd1kynKh
23:43 sorear rakudo: "$_ ".say for 1..*
23:43 p6eval rakudo 7175e5: OUTPUT«(timeout)1 ␤2 ␤3 ␤4 ␤5 ␤6 ␤7 ␤8 ␤9 ␤10 ␤11 ␤12 ␤13 ␤14 ␤15 ␤16 ␤17 ␤18 ␤19 ␤20 ␤21 ␤22 ␤23 ␤24 ␤25 ␤26 ␤27 ␤28 ␤29 ␤30 ␤31 ␤32 ␤33 ␤34 ␤35 ␤36 ␤37 ␤38 ␤39 ␤40 ␤41 ␤42 ␤43 ␤44 ␤45 ␤46 ␤47 ␤48 ␤49 ␤50 ␤51 ␤52 ␤53 ␤54 ␤55 ␤56 ␤57 ␤58 ␤59 ␤60 ␤61 ␤62 ␤63 ␤64 ␤65 ␤66 ␤67 ␤68
23:43 sorear rakudo: my $x; $x := $x; $x = 5
23:43 p6eval rakudo 7175e5: OUTPUT«Segmentation fault␤»
23:44 sorear wtf
23:44 sorear I just added code to print "(signal " . ($? & 127) . "?"
23:45 sorear oh, hmm, I didn't actually have the & there
23:45 sorear maybe $? is a dualvar
23:45 TimToady I believe it is
23:46 TimToady or maybe not
23:46 TimToady $! is
23:47 sorear well I won't complain. :)
23:47 TimToady it would be OT :)
23:47 lichtkind what is a dualvar?
23:48 TimToady a p5 var with both a string and numeric value
23:48 dalek evalbot: 94a9b38 | sorear++ | / (2 files):
23:48 dalek evalbot: Fix (timeout) overwriting, generate errors for signal-exit
23:48 dalek evalbot: review: https://github.com/perl6/evalbot/commit/94a9b383b1
23:48 icwiener left #perl6
23:48 lichtkind TimToady: and both are different?
23:48 mberends left #perl6
23:49 TimToady they're not autmatically derivable from the other value
23:49 lichtkind i never noticed that, thank you
23:50 TimToady any scalar can have both a numeric and string value, but that's just caching a coercion
23:51 TimToady you can set $! to an errno, and it will give you back a string with the error message (in P5)
23:51 sorear buubot_backup: eval: (kill 9, 1), [ 0 + $!, "" . $! ]
23:51 buubot_backup sorear: [1,"Operation not permitted"]
23:51 spetrea I shaved my head http://i.imgur.com/MIqkW.jpg
23:51 tokuhir__ joined #perl6
23:52 sorear buubot_backup: eval: (unlink "/tmp/pies"), [ 0 + $!, "" . $! ]
23:52 buubot_backup sorear: [2,"No such file or directory"]
23:52 sorear that's how it works
23:53 sorear TimToady: who is responsible for PerlIO?  Should p6 try to offer something similar?
23:54 TimToady um, that's what all the discussion of the S32/IO has been about
23:54 sorear I think I've improved p6eval enough for today.
23:54 TimToady you could put ...⌛  on the end if it times out :)
23:54 TimToady .u ⌛
23:54 phenny U+231B HOURGLASS (⌛)
23:55 Tene joined #perl6
23:55 sorear S32/IO doesn't seem to have any STREAMSish features
23:55 sorear at most it handles encodings
23:56 tokuhirom left #perl6
23:59 TimToady there are other possible ways to model that in p6, I suppose, anywhere from a fancy "encoding" to a variant on feed operators

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

Perl 6 | Reference Documentation | Rakudo