« first day (717 days earlier)      last day (1777 days later) » 

05:06
@Dennis When you can, can you pull Attache?
05:18
@ConorO'Brien Done.
 
4 hours later…
09:31
@Dennis Can you please replace Braille with [this version](https://tio.run/##jVXfb9owEH5O/oprp1KbwEqgW7uF8DCplSZN2tOeOjQZxwFrmRM5gVbb@NvZnZO0hUasiAd8933fnX0/kMOllLvdG21ktk4UTMsq0fnb1czfM2V6QTZ/bUq9NCoBuRIWZG42ylbswMr9P75nVbW2BpiEHoweLkccplMIfQ/o87e1j66c/fLQfo342eylfXLt7JPI3/r@JtcJqAcl15U6yKFfiUINoNS/1Y8KksKltA/RpjQDEHYZ@V6D0wXEMIp837tf6UwBCwJdcECqR2h0kuydLuYumRShnocKcOgZpSTilfe6kitgRK5lPClKRYDRR8xcVwzZPIKFVeJn9Mwfot/dAYJGOikgoGznHejxI3r4CvQE0fsgTN/RO8CXj9KPyvOa1QF@h2ASjSGkqnaD3teg4VHQFYK0U6LnDSDswFzXmGGLGXdgPhAmBXbi7sD/qyle4o/qLxr8K@XlC/hR9aSrUktVUf@yrrZRSCjWtX@f2IVOu@StMAnjcEb@9Blp6@MXh06b
 
4 hours later…
13:53
@stasoid Can you put it in a GitHub repo?
 
4 hours later…
17:44
Or I'll make one for the organization. In any case, I'll add it tonight.
@stasoid Is rand() % 0xff correct? I assume it should be either rand() % 0x100 or rand() & 0xff.
18:17
@Dennis Yes, I prefer it to be on TIO's github.
@Dennis I don't know.
18:52
I wonder if & 0xff is faster than % 0x100 or if they both get optimized to the same thing
tape is of type unsigned char *, so it can be removed entirely. gcc probably does.
@Dennis could we add Swift 3 to TIO? Swift 3 tolerates quite a few useful things for golfing that Swift 4 doesn't
@Downgoat I don't believe Swift 3 runs on Linux
@Pavel wait really? H*ck
Actually ignore me I'm wrong
19:00
The only reason we have Swift 4 is because somebody made a repo for it. Before that, I tried (and failed, miserably) to compile it myself.
hm ok
There are tarballs at swift.org/download/#releases
19:13
@Dennis Surprisingly, it does not. rand() & 0xff gets optimized to rand(), but rand() % 256 actually modifies the return value (not using div, of course) before the assignment.
I don't think rand's return value is marked as unsigned
So it's probably preserving that thinking it might be signed
rand() returns an int, yes. But that shouldn't matter if the result gets cast to unsigned char.
19:28
Is it (unsigned) (rand() % 256) or ((unsigned) rand()) % 256
 
1 hour later…
20:33
@quartata It's (unsigned char) (rand() % 256), but gcc "knows" about the cast. (unsigned char) (rand() & 255) eliminates the AND.
clang removes the % as well.
Wot? Clang doing better optimization than GCC?
@Pavel why are you so surprised
@Downgoat I've always heard that GCC is better at optimization than Clang

« first day (717 days earlier)      last day (1777 days later) »