|
00:22
colomon joined
03:46
lue joined
06:04
zakharyas joined
06:45
kjs_ joined
06:56
kjs_ joined
07:05
kjs_ joined
07:12
kjs_ joined
07:47
FROGGS joined
08:52
FROGGS joined
09:19
FROGGS joined
11:04
ggoebel111111117 joined
13:56
nebuchad` joined
14:19
woolfy left
14:20
woolfy joined
15:07
kjs_ joined
15:46
brrt joined
16:31
brrt left
16:42
kjs_ joined
16:43
colomon joined,
d4l3k_ joined
16:50
colomon joined
18:00
kjs_ joined
18:15
FROGGS joined
|
|||
| dalek | arVM: eb49939 | TimToady++ | src/6model/reprs/NFA.c: reduce fates to 24 bits |
19:45 | |
| FROGGS | "The high 40 bits of the fat encodes[...]" | 19:48 | |
| fat? | |||
| jnthn | Encodes can become obese too... | 19:49 | |
| japhb | .oO( phat encodes ) |
19:50 | |
|
19:59
kjs_ joined
|
|||
| TimToady | I sincerely doubt that we need 40 bits for the literal length, but apparently parrot is having difficulty with numbers in the neighborhood of 2^32 | 20:14 | |
| plus, if we ever try to squeeze these down to 32 bit numbers, it at least leaves use 8 bits for the literal length | |||
| I doubt we'll see any literals even that long, and we could spec that they max out at 127 or 255 | |||
| *us | 20:15 | ||
|
20:20
colomon joined
|
|||
| TimToady | only charlists are using pointers, and those could be indices into any array of bitmaps instead | 20:24 | |
| which would probably run faster anyway | |||
| ('course, bitmaps might not work well with NFG) | 20:26 | ||
| FROGGS | PNG is superior to bitmaps anyway | ||
| FROGGS .oO( don't gif up the puns ) | 20:27 | ||
| jnthn | .oO( If I weren't behind with work, I could rif off a few more... ) |
20:43 | |
.oO( If I don't get this done, there'll be a tiff at best, or somebody psd at worst... ) |
20:45 | ||
| FROGGS | just png me when that here is over | 20:59 | |
| japhb | Don't mov, I'm sure it'll end soon | 21:03 | |
|
21:38
colomon joined
23:05
colomon joined
|
|||
| dalek | arVM: eeb5c92 | (Timo Paulssen)++ | src/jit/graph.c: jit cmp_I |
23:34 | |