01:45 vendethiel joined 02:24 Peter_R joined 03:13 BinGOs joined, leedo joined
dalek arVM: 0324fe3 | (Jimmy Zhuo)++ | tools/ucd2c.pl:
add some const to global static vars
06:18
06:35 FROGGS joined 07:11 oetiker joined 07:34 brrt joined
jnthn JimmyZ: that doesn't update the C files, and if you had, you'd have seen a bunch of warnings. That patch as it stands just creates warnings for somebody to have to clear up down the line, so I'm -1 to it. 07:34
brrt \o 07:35
nwc10 o/
jnthn o/ 07:37
07:39 dalek joined 07:41 zakharyas joined
JimmyZ jnthn: so it is fine if I clear the warnings? 07:44
I just need UNIDATA/extracted/DerivedBidiClass.txt or somebody updates it ... 07:45
jnthn JimmyZ: You just need to download the Unicode 7.0 database from the Unicode site; it's a ZIP file with all UNIDATA in. 07:47
Assuming Unicode isn't bourgeoisie... :)
07:53 Ven joined
JimmyZ hmm, if I update it, it causes unicode_db.c a big change 07:59
jnthn Sure it's not just line endings changes that go away on git add? 08:01
It should only change the lines you changed.
JimmyZ jnthn: turns out only one warning if somebody updates the C files. 08:12
I download the zip, update the C files and cause many uneeded C file changes, maybe I downloaded a old one? 08:15
it even cause many chnages to unicode_gen.h
jnthn That sounds wrong... 08:21
Can you do a fix for the warning? And I can do the update. 08:22
JimmyZ yes, I can
Thanks.
jnthn I've been running that script lots lately anyway due to working on NF*
JimmyZ I download it from www.unicode.org/Public/UNIDATA/UCD.zip 08:23
dalek arVM: ce0878e | (Jimmy Zhuo)++ | src/strings/unicode_ops.c:
fixed warnings
08:26
JimmyZ jnthn: It is no hurry to update the C files.
because I fixed the warnings. :P
jnthn ok :) 08:27
08:30 dalek joined 08:35 Ven joined 09:16 Ven joined 09:47 vendethiel joined 10:04 Ven joined 11:11 vendethiel joined 11:13 dalek joined 11:54 vendethiel joined 12:03 Ven joined 12:22 vendethiel joined 13:20 vendethiel joined 13:25 Ven joined 13:37 ShimmerFairy joined 13:41 Ven joined 13:46 vendethiel joined 14:05 Ven joined 14:33 vendethiel joined 14:43 Ven joined 16:24 vendethiel joined 16:49 vendethiel joined 17:33 vendethiel joined 17:45 FROGGS joined 17:54 zakharyas joined
dalek arVM: 2bad607 | FROGGS++ | src/ (2 files):
don't lie about MVM_BOOL_MODE_UNBOX_STR_NOT_EMPTY (_OR_ZERO)

Now MVM_BOOL_MODE_UNBOX_STR_NOT_EMPTY basically delegates to istrue_s and MVM_BOOL_MODE_UNBOX_STR_NOT_EMPTY_OR_ZERO considers "=" to be False.
19:18
psch tpyo :D 19:19
FROGGS where? 19:20
psch "="
FROGGS damn
nwc10 oooh, OK, stage0 bump 19:21
FROGGS yeah, had to :o(
nwc10 so, "this week" the new serialisation version will be 15?
and we can toss the 13 code? :-)
FROGGS you probably know more than me :o)
nwc10 slightly, but I'm not confident that I only guessed the right question and answer pairs when I was chatting with jnthn 19:22
anyway, new stage0 is smaller
19:40 FROGGS joined 19:47 nwc10_ joined
timotimo that's very good 20:03
TimToady wow, that nozzle is red hot 20:15
FROGGS that's fun without context :o) 20:18
20:28 brrt joined
brrt \o 20:28
time for my weekly jit article dump: github.com/burningmime/curves/blob...JITPerf.md
nwc10 o/
brrt tl;dr - SIMD is really impressive
\o nwc10 :-)
FROGGS yay! something to read! 20:29
brrt :-) 20:30
hmm, i'm going to spoil it for you, but i think i have a theory why the above-normal fetching-and-storing for the SIMD doesn't mattter much, and it is pipelining
and rsp aliasing too 20:34
21:02 lizmat_ joined 22:40 colomon joined 22:47 colomon joined 23:07 colomon joined