> But personally, I don't use tell-dont-ask. I do look to co-locate data and behavior, which often leads to similar results. One thing I find troubling about tell-dont-ask is that I've seen it encourage people to become GetterEradicators, seeking to get rid of all query methods.
During the middle ages, France and Spain (and probably Sicily/Southern Italy) used two numerals: Arabic and Roman. In the end, Arabic numerals gave our (western) current numbers and Romans are mostly kept in titles and names.
Some watches still use Roman numerals (XII, III, VI, IX), Swatch specify here that those are Arabic (12, 3, 6, 9).
There was an example, of ChatGPT copying and responding in the speakers voice mid conversation, on OpenAI blog. This was presented an example on non-alignment.
Last line immediately reminded of Star Citizen which was first announced in 2012 and people are still paying for the promises/hypes the company keeps making.
I believe you, my brain doesn't. Those movements are so smooth. The line where two different colours merge have no sharp edges and look perfectly curved. Only thing edgy was the very last blinking animation where there weren't any colours.
Thinking more, if there were only 4 LEDs I can imagine how they would look like diffused.
Initially I was thinking if 8x8 can show all those smooth details and motions, can it be used to show any other higher resolution imagery instead of just moving colours.
I think the thing that fools us here is the fact that we're used to pixels being discrete, so it doesn't "look" low-res because the colors blend into each other. If you imagine that each LED produces a small circle onto the diffuser, and that this circle overlaps the ones around it a bit, it gets easier to see through the illusion.
It's much easier to understand when you can change the distance of the diffuser to the panel (which I did when testing), because then you can see the lights go from little squares with lots of dark space around them, to this, to big blobs of one color.
A diffuser is basically applying a convolution to the underlying pixel array (moving the diffuser away from the LED increases the convolution width, but unfortunately also reduces the intensity).
I don't think you can exploit that to show something higher resolution than the original pixel array,
Or is it that the example in the article is a bit poor?
reply