Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2013-12-05

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

All times shown according to UTC.

Time Nick Message
00:17 FROGGS joined #moarvm
00:35 colomon joined #moarvm
01:56 jnap joined #moarvm
03:32 ssutch joined #moarvm
04:16 jnap joined #moarvm
07:19 jnap joined #moarvm
08:00 ssutch joined #moarvm
08:16 FROGGS joined #moarvm
11:22 jnap joined #moarvm
12:18 odc joined #moarvm
13:24 jnap joined #moarvm
13:35 woolfy left #moarvm
14:14 camelia joined #moarvm
15:14 jnap joined #moarvm
16:00 TimToady .oO(MoarVM is...resting...right?)
16:04 _sri psssst
16:14 diakopter :)
16:14 nwc10 MoarVM is blotto?
16:14 TimToady I shoulda said #moarvm
16:53 jnap joined #moarvm
16:54 jnthn Well, the Next Problem is hunting something that looks like a closure serialization issue, and I ended up working on indy stuff 'cus I needed to brush up on it for my talk next week. :)
16:54 * nwc10 is currently testing the NQP indy branch on Moar with moar-support
16:54 * diakopter dares to ask what indy is
16:54 diakopter "I was named after the dog!?!"
17:02 nwc10 jnthn: gets to
17:02 nwc10 ./perl6-m --target=mbc --output=blib/Test.moarvm lib/Test.pm
17:02 nwc10 Unhandled exception: Cannot invoke this object
17:02 * nwc10 thinks that that is where we were before
17:03 jnthn Yeah.
17:03 jnthn Sounds like, anyway.
17:03 jnthn diakopter: invoke dynamic
17:06 TimToady maybe in-dy would produce a less FAQ
17:10 colomon joined #moarvm
17:10 jnthn Wasn't my choice of shortening ;)
17:28 diakopter TimToady: it wouldn't be any less frequently asked by me, since I now realize I'd asked it (at least!) once before already
17:51 FROGGS[mobile] joined #moarvm
19:14 cognominal joined #moarvm
20:11 FROGGS joined #moarvm
20:44 dalek MoarVM: bd782aa | jnthn++ | src/core/exceptions.c:
20:44 dalek MoarVM: Exception category should default to catch.
20:44 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/bd782aa821
20:48 FROGGS jnthn: how do we want to get rid of that gc bug?
20:48 FROGGS I am really out of ideas on that topi
20:48 FROGGS topic*
20:48 jnthn I need to reproduce it, I guess...
20:49 jnthn More pressing on getting setting loading working at the moment, though...
20:50 FROGGS yeah
20:51 FROGGS hmmm, I could even hack on windows if I had to
20:53 jnthn Or just build with a lower opt level, iirc?
20:54 FROGGS true, that should still work
20:55 FROGGS jnthn: you want to show perl6-m somewhere?
20:58 jnthn FROGGS: No, just would really like to get to "hello world". :)
20:58 FROGGS k
21:10 cognominal joined #moarvm
21:14 dalek MoarVM: 6f955c8 | jnthn++ | src/core/frame.c:
21:14 dalek MoarVM: Make sure scalar containers are fresh per frame.
21:14 dalek MoarVM:
21:14 dalek MoarVM: Also stub the place we'll later need to provide some support for state
21:14 dalek MoarVM: variables.
21:14 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/6f955c8b2d
21:17 jnthn We now make it to the $PROCESS::TZ = get-local-timezone-offset(); at 14731
21:17 FROGGS yay!!
21:19 jnthn We don't need that yet so will just comment it out...
21:19 FROGGS :S
21:20 jnthn Well, #?if !moar it
21:20 jnthn It hits some binding thing, but it will also hit the decodelocaltime not being done yet, I'd guess...
21:20 FROGGS yeah
21:23 jnthn Did exactly that for JVM port too :)
21:23 jnthn TZ? Yah, later... :)
21:25 FROGGS hehe
21:25 FROGGS the spec tests will force us to fix it then :o)
21:26 jnthn aye
21:26 jnthn Well, we gotta manage sanity test first... :)
21:27 jnthn I'm gonna probably worry about the GC bug, and the missing GC opt that makes setting comp take ages, before digging too far into that, though.
21:27 jnthn 'cus otherwise it's a long build cycle.
21:29 jnthn 16614...
21:38 FROGGS :o)
21:38 FROGGS happy FROGGS is happy
21:41 jnthn 16640
21:50 FROGGS rebase -i is pain
21:51 timotimo it is?
21:51 timotimo i love rebase -i
21:53 FROGGS well, I wanna fixup 15 commits into one, and have conflicts here and there
21:53 * jnthn likes it too, but it took a little learning the first time :)
21:54 FROGGS mostly in src/core/Array.pm with diffs that does not fit on my screen
21:54 timotimo oh, oops :)
22:06 tgt joined #moarvm
22:16 jnthn *drumroll*
22:17 jnthn C:\consulting\rakudo>perl6-m -e "say 'hi moar'"
22:17 jnthn WARNING: State vars NYI
22:17 jnthn WARNING: State vars NYI
22:17 jnthn hi moar
22:18 diakopter wat.
22:18 diakopter time? :)
22:20 FROGGS \o/
22:20 FROGGS I can haz "hi moar"?
22:21 jnthn Yes.
22:21 diakopter what's the startup/run time like?
22:22 jnthn diakopter: 0.72s, probably muchly coming from the time we waste decoding the huge setting blob to NFG only to immediately ASCII it again...
22:23 diakopter well the nfg reader knows to make ascii strings when possible, I thought
22:23 diakopter oh, but yes.
22:23 diakopter :)
22:23 jnthn Well, it does, but it doesn't know it's not UTF-8
22:23 jnthn We actually could consider indicating "assume ascii when decoding" in the bytecode file at the point we create it.
22:24 diakopter I thought I made it assume ascii by default to start, then demote to bigger thing when necessary
22:24 jnthn Ah...I dunno, I thought it only did that when decoding ascii. I may be wrong...
22:25 jnthn Either way it's doing the work to look for potential utf-8...and vice versa when it encodes again...
22:25 diakopter when decoding ascii it never makes anything but ascii
22:25 jnthn Right.
22:25 jnthn But the constant pool in the bytecode file is currently defined as "a bunch of things in utf-8"
22:26 jnthn Goes without saying we still get this thing started/run faster than JVM :)
22:26 FROGGS *g*
22:26 jnthn Anyway, we can optimize later. For today, I'm just happy we can actually do "hello world" :)
22:27 FROGGS yeah
22:27 FROGGS jnthn++
22:54 jnthn Guess I should have a nice beer to mark getting to this point... :)
22:56 FROGGS[mobile] I'll enjoy a nice water then :S
23:08 jnthn No beer?
23:39 woolfy joined #moarvm

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