5 ms·

How is jpeg xl “hostile”, it’s just a new format. Does Chrome really need to support every new image format?

uoaei 3y ago
Can you expand on what you mean? As it stands your questions seem in incredibly bad faith, especially considering the market share Chrome has in the browser space.
- yankput 3y ago
  The article says Chrome sabotaged JPEG XL. I am saying that JPEG XL is a new format, which it is; any additional format that is supported means more code, more potential for bugs, more security issues (format parsing is one of the biggest sources of security issues). Chrome had experimental support for it and then they dropped it. They still support PNG, normal JPEG. I agree with the arguments against the current DRM and Manifest 3 content blocking was kind of misdirected (however… look at how Safari does it, it’s essentially the same) Also I don’t think the same argument applies for Chrome as for OSes for example. You are not really locked in. Just use a different browser. It’s just web.
detuur 3y ago
It was/is a very well thought-out format that had a lot of industry signalling they were willing or even enthusiastic about supporting it, which happens virtually never for new image formats. All the major browsers were working on support, many image processing pipelines were being updated. Everyone agreed that it was not just a major improvement, but actually the best option out of all the contenders out there. Even the Chrome team. The only ones who didn't agree were the AVIF team at Google, who had developed a competing standard. Well, whoever was on that team had some pull, because not long after Chrome landed JXL support in stable and everything was about to pick up some serious pace, they suddenly landed a commit that reverted everything to do with JXL support in Chrome, and that was that. An immense, cross-industry effort undone by internal Google politics. That's what's hostile about this situation. JXL is still limping along, but Google's unilateral reversal hurt everyone's confidence in the project but also the entire process. Apparently it doesn't matter what everyone in the world thinks is the most appropriate format. What matters is Lord Google's favour.
- jillesvangurp 3y ago
  A possible reason might be some patents owned by Microsoft: https://www.theregister.com/2022/02/17/microsoft_ans_patent/ https://www.theregister.com/2022/02/17/microsoft_ans_patent/ That's also a good reason for others to drop support until licensing of these patents is clarified.
  - bob1029 3y ago
    This is also why the original jpeg was forced to use a less-than-ideal encoding scheme (Huffman vs arithmetic). Most libraries have flags that enable the alternative mode. Maybe we should stop issuing patents for math problems?
  - pilif 3y ago
    > That's also a good reason for others to drop support until licensing of these patents is clarified. If that were true, then browsers should drop support for VP8, VP9 and indeed AV1 because the MPEG-LA is claiming ownership of patents covering those codecs.
    - jillesvangurp 3y ago
      The rollout of that stuff took some time precisely for this reason.
  - jonsneyers 3y ago
    Those Microsoft patents are not relevant for JPEG XL. Their licensing is therefore also not relevant for JPEG XL.
- magicalist 3y ago
  > Everyone agreed that it was not just a major improvement, but actually the best option out of all the contenders out there. Wikipedia tells me the first of four parts of the jpeg xl standard was finalized less than two years ago, and the last part (barely) less than a year ago. The idea that a) browsers are bad if they're not adding a year-old image format to the web forever and b) "everyone" knows what they're talking about with an image codec finalized less than a year ago is pretty ludicrous. It's a good format, it'll gain adoption.

Can you expand on what you mean? As it stands your questions seem in incredibly bad faith, especially considering the market share Chrome has in the browser space.

yankput 3y ago
The article says Chrome sabotaged JPEG XL. I am saying that JPEG XL is a new format, which it is; any additional format that is supported means more code, more potential for bugs, more security issues (format parsing is one of the biggest sources of security issues). Chrome had experimental support for it and then they dropped it. They still support PNG, normal JPEG. I agree with the arguments against the current DRM and Manifest 3 content blocking was kind of misdirected (however… look at how Safari does it, it’s essentially the same) Also I don’t think the same argument applies for Chrome as for OSes for example. You are not really locked in. Just use a different browser. It’s just web.

detuur 3y ago

It was/is a very well thought-out format that had a lot of industry signalling they were willing or even enthusiastic about supporting it, which happens virtually never for new image formats. All the major browsers were working on support, many image processing pipelines were being updated. Everyone agreed that it was not just a major improvement, but actually the best option out of all the contenders out there. Even the Chrome team. The only ones who didn't agree were the AVIF team at Google, who had developed a competing standard. Well, whoever was on that team had some pull, because not long after Chrome landed JXL support in stable and everything was about to pick up some serious pace, they suddenly landed a commit that reverted everything to do with JXL support in Chrome, and that was that. An immense, cross-industry effort undone by internal Google politics. That's what's hostile about this situation. JXL is still limping along, but Google's unilateral reversal hurt everyone's confidence in the project but also the entire process. Apparently it doesn't matter what everyone in the world thinks is the most appropriate format. What matters is Lord Google's favour.

jillesvangurp 3y ago
A possible reason might be some patents owned by Microsoft: https://www.theregister.com/2022/02/17/microsoft_ans_patent/ https://www.theregister.com/2022/02/17/microsoft_ans_patent/ That's also a good reason for others to drop support until licensing of these patents is clarified.
- bob1029 3y ago
  This is also why the original jpeg was forced to use a less-than-ideal encoding scheme (Huffman vs arithmetic). Most libraries have flags that enable the alternative mode. Maybe we should stop issuing patents for math problems?
- pilif 3y ago
  > That's also a good reason for others to drop support until licensing of these patents is clarified. If that were true, then browsers should drop support for VP8, VP9 and indeed AV1 because the MPEG-LA is claiming ownership of patents covering those codecs.
  - jillesvangurp 3y ago
    The rollout of that stuff took some time precisely for this reason.
- jonsneyers 3y ago
  Those Microsoft patents are not relevant for JPEG XL. Their licensing is therefore also not relevant for JPEG XL.
magicalist 3y ago
> Everyone agreed that it was not just a major improvement, but actually the best option out of all the contenders out there. Wikipedia tells me the first of four parts of the jpeg xl standard was finalized less than two years ago, and the last part (barely) less than a year ago. The idea that a) browsers are bad if they're not adding a year-old image format to the web forever and b) "everyone" knows what they're talking about with an image codec finalized less than a year ago is pretty ludicrous. It's a good format, it'll gain adoption.