@neauoire excellent sin LUT

@jonbro you should see my tri LUT

@tty 🤍 it's your day

@tty It's a single .Screen/sprite DEO using the screen's auto byte.

It draws a vertical column of the flag in a single draw call.

@neauoire Ok, I clearly don't grok the auto byte yet. The rest of the code makes sense though -- most of the work here is just doing the sine table lookups & adding them together :)

@neauoire is there an easy way to plug that raw data from the source into a program to see the output? is that was nasu does? (i haven't really messed with any of your tooling)

@neauoire #f5 in the auto byte is length 16 (ff) and 5 (bit 1 and 3), which maps to.. auto-x and auto-addr? I'm not clear on what auto-addr does I think.

@neauoire ok! I think I understand just about everything.

- ;col-icn encodes 16 8x8 2bpp sprites, in vertical sequence (2 blue, 3 pink, 1 transition tile, 2 white, etc)
- "auto addr" means "automatically advance the .Sprite/addr by 4 on each draw"

I'm still not clear why #f5 is used for .Sprite/auto, since the low nibble *should* mean auto-x, right??

@tty the auto-x is not really necessary in this case, it's vestigial of another design where I didn't re-set the x position on each draw.

@neauoire oh. but how is it advancing downwards if auto-y isn't being set? wouldn't auto-x be causing tiles to draw horizontally instead?

@tty Yeah, it's a bit odd. The length is the direction that is not auto-advancing, like:

@neauoire oooooh, I see!

@neauoire ok, I think that maybe wasn't clear from the varvara docs!

@neauoire but now this makes total sense

@tty I will try to rephrase the docs, it's a bit confusing. I have to re-explain this everyday recently haha. I ought to make a diagram on the site or something.

@tty yes, create a new nasu file, and paste the col-icn data in it, and you'll see it :)