Classic strawman. The claim I've consistently challenged is that you said the output can be bigger than the input after lossless compression:
As I said, that is mathematically inherent when you have a finite compression algorithm and completely arbitrary content (both is the case here).
At best the protocol can then switch to the original uncompressed content (with some overhead added on top) but that option does not exist when a fixed compression ratio is mandated by limited channel bandwidth.
Then lossy compression is the only option, as is the case with GPUs which can't drive the full resolution uncompressed to Thunderbolt because they don't support the full data rate.
The original contention was that:
a) DSC was completely lossless – it
isn't (actual lossless compression is used as far as
possible, but it
isn't always possible, in particular not with high-contrast, all-channel-full-spectrum-noisy content)
b) Apple mandated GPUs always needing to support DSC for the XDR (and by externsion for the ASD) – they
don't (because in both cases they very much support the full, uncompressed image signal if the GPU can actually provide that through Thunderbolt, which some smaller / older GPUs just can't, so only
those need DSC to make up for the shortfall, but with the risk of compression artefacts – the more the higher the compression ratio)
Both points have now indeed been conceded exactly as I had been saying from the start, so agreement at long last.
There's nothing weasly here. VESA's been up-front from the start that their algorithm is an attempt to make a codec that IS NOT informationally lossless, but IS it visually lossless. Indeed, that's the entire point of the codec. Everyone understands that it is not informationally lossless, so it's laughable to claim that they're somehow hiding something, or confusing people, by using that term. So stop making stuff up. The irony is you're the one that's blowing smoke here.
Nowhere clearly communicating that DSC does indeed come with artefacts under certain circumstances but in the PR copy incorrectly suggesting perfect quality is simply
dishonest.
This kind of misdirection is quite common, unfortunately, and in this discussion it has caused that confusion about the Apple displays.
Even some technical information must be read very carefully like a lawyer would, because it is not written for clarity, but for obfuscation.
Unfortunately even Apple isn't adding that much to explain the issue.