timotimo a ... set? 00:04
but sets are unordered %)
diakopter theoretically! 00:05
konobi yay... looks like there's an answer to fixing JIT on illumos 00:09
illumos/solaris even
timotimo i'll go to bed; don't have anything to show yet 00:31
cognominal m: .say for IO::Handle.^methods.>>.candidates.flat.>>.perl.sort.map: { m/ (.*?) '#`(' /; "$0... }" } 01:58
camelia rakudo-moar 04791b: OUTPUT«method FALLBACK (RESTRICTED-CLASS $: | is raw) { ... }␤method gist (RESTRICTED-CLASS $: | is raw) { ... }␤method new (RESTRICTED-CLASS $: | is raw) { ... }␤»
cognominal héhé :) 01:59
03:41 vendethiel joined 08:30 lizmat joined 08:31 virtualsue joined 08:42 lizmat joined 08:54 lizmat joined 09:02 vendethiel joined 09:07 lizmat joined 09:19 lizmat joined 09:33 leont joined 09:35 domidumont joined 09:41 domidumont joined 10:06 vendethiel- joined, ggoebel9 joined 10:07 colomon_ joined 10:24 Ven joined 10:55 leont joined 11:15 rurban joined 13:32 dalek joined 14:38 lizmat joined 14:47 Ven joined, leont joined 15:13 cognominal joined 15:40 Util joined 15:41 domidumont joined 15:49 lizmat joined 15:51 leont joined 15:54 rurban joined 16:38 JimmyZ joined
dalek arVM/multidim_references: 95745ec | timotimo++ | src/ (4 files):
nativeref multidim struct, repr functions, hllconfig entries
 17:22
timotimo ^- that's the first little step
jnthn Will look after shopping, now I've finished being indecisive about what to make for dinner :) 17:26
bbiab
17:34 flussenc1 joined 17:36 ilmari_ joined
nwc10 jnthn: indecisive - as in "what sort of curry?" ? 17:39
17:44 harrow joined 17:47 JimmyZ joined
dalek arVM/multidim_references: eef1604 | timotimo++ | src/6model/containers.c:
native_ref_fetch_* learn about multidim
 18:08
arVM/multidim_references: 37e2091 | timotimo++ | / (6 files):
add multidimref_* ops to the interpreter
18:11 [Coke] joined
jnthn nwc10: Ended up doing chili con carne. :) I'm missing ingredients for most of the curries I tend to do, and didn't fancy the trek to the bigger store that sells them. 18:15
timotimo jnthn: where in Array.pm do i put the AT-POS for multidimref_*? there's a ShapedArray and a TypedArray; maybe i'll have to add another role that does "only" the AT-POS that returns the reference and have that in three different versions for each register type 18:22
jnthn You don't 18:23
Array.pm is not native
It's only AT-POS in native_array.pm's shaped role that'd need it 18:24
timotimo oh! 18:25
perfect
that's probably a bit easier
especially since it already has the three roles 18:26
jnthn: did it even make sense that i made a multidim reference for Str? 18:29
hm. i'll prolly have to fudge out the multidim refs for jvm 18:31
and the usage there, too
jnthn timotimo: Yeah, I can imagine us supporting my str @arr some day 18:40
timotimo it's not quite "packed" then, though? 18:41
jnthn Yeah :)
But it's the kind of thing join maybe uses to be a bit smarter
And similar things 18:42
timotimo mhm
jnthn So I can see we may want it from a "Perl 6 in Perl 6" perspective
timotimo timo@schmetterling ~/p/rakudo (nom)> perl6 -e 'my int @foo[2;2]; @foo[1;1]++; say @foo' 18:54
[[0 0] [0 1]]
dalek arVM/multidim_references: b804b13 | timotimo++ | src/ (3 files):
allow multidim nativerefs to be registered in hll
 18:55
timotimo i think i can merge this branch into moar now
nwc10 does it pass spectest? 18:57
timotimo should i try with or without the additional nqp and rakudo patches? 18:58
nwc10 I don't know enough about the branch to know the answer to that.
jnthn Just test the bunch together to make sure no regressions 18:59
timotimo right
jnthn I can have a look, but my chili just reached the point of nommability :)
so bbi20 or so :)
timotimo the moar branch adds alternative code paths, doesn't really change existing ones
timotimo specs tests 19:00
gist.github.com/timo/48fadfb2da0ff6c840a8 - does that look regression-free to you? 19:16
nwc10 I don't recognise the failure for /spec/S32-array/adverbs.t 19:18
timotimo i do 19:19
dalek arVM: 95745ec | timotimo++ | src/ (4 files):
nativeref multidim struct, repr functions, hllconfig entries
 19:21
arVM: eef1604 | timotimo++ | src/6model/containers.c:
native_ref_fetch_* learn about multidim
arVM: 37e2091 | timotimo++ | / (6 files):
add multidimref_* ops to the interpreter
arVM: b804b13 | timotimo++ | src/ (3 files):
allow multidim nativerefs to be registered in hll
jnthn back 19:40
timotimo welcome back, jnthn 19:41
i went ahead and merged in moar and bumped nqp
and nqp now maps the ops, too
i have a commit ready to be pushed in rakudo
jnthn + MVM_ASSIGN_REF(tc, &(ref->common.header), ref->body.u.multidim.obj, obj); 19:46
+ ref->body.u.multidim.indices = indices;
timotimo oh, i didn't copy-paste that line?
jnthn Second one should be MVM_ASSIGN_REF too
timotimo i swear i did! :)
i mean ... obviously not
will fix
i guess i confused myself ... i duplicated the MVMROOT and perhaps my brain thought that'd be enough for both things 19:48
dalek arVM: e7edb06 | timotimo++ | src/6model/reprs/NativeRef.c:
add a missing ASSIGN_REF for multidim refs
timotimo jnthn++
jnthn Don't see anything else in the Moar patches 19:50
NQP one is fine too
timotimo great! i'll bump again in nqp and fixup the pending rakudo commit 19:51
whew. i finally did something actually productive again 19:53
badly needed that. 19:54
then i should also write some spec tests
jnthn :) 19:55
timotimo tonight is a celebration-like eventy thingie
jnthn timotimo++
timotimo but it's the kind where bringing your laptop and hacking will be acceptable
jnthn Sounds nice :)
timotimo yup. it takes place in the local hackerspace 19:56
hitting the road soon
20:10 xiaomiao joined 21:50 vendethiel joined 22:10 leont joined 22:43 vendethiel joined
konobi howdy all 22:44
timotimo howdy konobi 22:45
23:13 vendethiel joined 23:21 leont joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!