Mozilla Eyes Removal Of Theora Support In Firefox

[uxmkt]

What is, if there are Theora videos on some websites?

GIFs and dynamic pages are already outrageous, don't even mention full-motion videos!

That said, shouldn't ffmpeg's integration into firefox (on Linux anyway) solve most of the codec troubles?

[Quackdoc]

Mozilla probably thought that if VR takes off it would be really bad for them to not have support ready for it. To be honest, HDR is a gimmick similar to VR and not widely supported either. Your fake HDR display most likely has its fake HDR disabled by default and you don't even know it. Enabling it would result in awful picture quality and you'd hate your life.

HDR support in Firefox is not really needed.

I strongly disagree with this, decent HDR displays are getting cheaper and cheaper, and HDR is genuinely a better viewing experience for games movies and images.

Apple could make the start on their websites and maybe push others to adopt JXL + a WASM decoder. Safari natively supports JXL, so it would make it appear to be the fastest browser, and they could advertise some neat benchmarks like "Safari renders apple.com x times faster than Chrome".

i've setup jxl wasm on a couple sites, sadly it has a tendency to crash firefox when more then 5 images are on a page, and chrome slows to a crawl, I plan on re-writing the jxl-oxide wasm decoder's JS interface to make it more usable, then test that sometime however

[curfew]

My display, yes it has some fake HDR, which doesn't work correctly, even on Windows, but my TV has proper HDR support.

And, yes, you are in the tiny minority who might sometimes use their HDR-enabled TVs for watching videos from your computer's web browser.

HDR televisions also come with built-in web services that virtually every consumer will use instead of Firefox on PC.

[Artim]

What is, if there are Theora videos on some websites?
Then they can no longer be seen with both browsers.

Incorrect. As already described in the Chrome thread of this "issue" several times, Theora was never really supported by any HTML5 player, and also Safari never supported it at all. So if there's a webpage out there still serving Theora, that's not unmaintained for a decade now, will most likely already use a JS player, so the decoder inside the browser is never used anyways.

[Artim]

0.09% of content browsed by users is Theora?

Nice privacy you got there champs, you love that telemetry?

Good thing they can't know your absolute window positions with Crapland though. Now that's a huge privacy issue indeed! But what videos you watch, who cares, right?!??

Because it's impossible to disable "Allow Firefox to send technical and interaction data to Mozilla" in the settings or what? Also, they never claimed to be able to tell which videos you watch. They probably added a feature a while ago when they first considered removing the Theora crap that would count how often the Theora decoder was used (and maybe what other formats nobody uses) compared to the more common codecs and regularly send that number to Mozilla. There's not that much you can learn from that number.

[Weasel]

Because it's impossible to disable "Allow Firefox to send technical and interaction data to Mozilla" in the settings or what? Also, they never claimed to be able to tell which videos you watch. They probably added a feature a while ago when they first considered removing the Theora crap that would count how often the Theora decoder was used (and maybe what other formats nobody uses) compared to the more common codecs and regularly send that number to Mozilla. There's not that much you can learn from that number.

Cool, now where's the option to enable absolute window positioning in Crapland? Even if it's not on by default.

I could say the same thing. There's not much you can learn from those numbers (positions)…

[pong]

To some extent, yes, halting problem, etc. etc. code analysis of complex code is hard / expensive.

But part of the benefit of things like orthogonality, design by contract, etc. is that one can just
have a relatively small set of inputs, define "obviously correct" preconditions for the ranges of those input data,
and explicitly code the body so that the entire stated acceptable range of input values are handled safely,
and bounds checking is done to ensure that memory safety is obeyed regardless of that the index or whatever is.
Keep the safety checks separate from / prerequisite to the business logic and they can be verified easily so
that it's memory safe regardless of what the bad data / other business logic is.
One can possibly even automatically prove certain things are going to have certain invariants / post conditions /
safety guarantees particularly in simple cases where again one is just constraining the bounds of access to
match the explicit bounds of a data structure.

The main issue is : _proving_ that a code is safe is very very expensive and costly.