JPEG XS - Wikipedia
0 points by HugoDaniel
0 points by HugoDaniel
I don't find non sequitur Wikipedia links that interesting. What's the point of posting this?
No idea but I had never heard of it. It's even 2 years older than JPEG XL, which seemed to be all over the news. I found it interesting. https://xkcd.com/1053/
The Talk page implies it's a slight rewrite of a commercial white paper.
Dunno about the patent encumbrance situation either. What are "RAND Terms"?
Reasonable and Non-Discriminatory; ie, everyone gets the same price and the price is "reasonable".
OK. Though it was from RAND Corp for some reason.
So if I create software that uses this protocol, can it be FLOSS?
In general, if some technology’s marketing talks about “RAND licensing” that typically indicates that it is patented and you can’t use it without paying. This is common in graphics and cabling/connector standards, where the standards consortium is also a patent cross-licensing pool or operates alongside one. (And it isn’t a cartel at all, honest.) This article is a good example.
If the patent holders have released an implementation under a vanilla BSD or MIT licenses (ie, without patent-related restrictions), then that implementation is freely usable so long as the patent holders agree they really did mean what the licenses say. But you can’t create your own without permission from the holders of the government-granted monopoly.
I suspect it probably came from a conversation around https://github.com/raphlinus/nibnib which appeared over the weekend - @raph wrote that to do fast image decompression on low powered systems like his 'vic-2350' (reimagining of a c64 on an rp2350). But while this thing says it'd "allow lightweight encoder and decoder implementations" I don't think it's in the same ballpark as nibnib, which is even simpler than QOI.