Add support for the JPEG XL format about performance HOT 7 CLOSED

As it's been a while since the initial comments here at the beginning of 2022, and you probably noticed that JPEG XL has gotten really good industry and software adoption in recent months. The list here is getting pretty long. This image codec's support has spread like wildfire on Linux, all the major apps and desktops I've encountered support at least viewing (if not also encoding) those images now, it "Just Works", out of the box, and it all happened within the span of a few months... that's gotta say something, if it can offset your perception that "it isn't even clear if this format will ever see the light of day!"

I would also like to add an extra piece of information that might be relevant in making your job easier here. In addition to imagemagick already supporting it, and gd being interested in supporting this whenever possible, I recently discovered that there is also this convenience library for PHP that would surely be relevant to WordPress: https://github.com/joppuyo/jpeg-xl-encode

WordPress has tremendous marketshare weight on the web. It would be really good if it could drive innovation forward on this front, as it has a unique opportunity to shape the web for the better. One part of the Chrome/Chromium team pretends that "the industry is not interested" in this, which makes zero sense given the current high rate of adoption and endorsement. Statements of interest in this ticket by WordPress contributors might help.

https://cloudinary.com/blog/the-case-for-jpeg-xl is also a very good read in response to what has been observed in that ticket, and so is https://jpegxl.info/why-jxl.html as a shorter "features-centric" read.

from performance.

While I am a JPEG XL proponent, I don't really think it makes much sense to add JPEG XL support to WordPress until the format support has been added to browsers. The only browser that supports JPEG XL today is Pale Moon which has a negligible market share. Why should we generate images when barely no browser will be able to display them?

Furhermore, the multi-format image support in WordPress is currently a mess. Since there's no picture tag support, WordPress core developers have been coming up with all kinds of hacks when implementing WebP support for unsupported browsers, including rewriting image tags using JavaScript or using a JavaScript polyfill for converting WebP images to JPEG in the browser.

I think adding JPEG XL or even AVIF support can't be done properly until proper picture tag support has been added to WordPress.

I appreciate linking my library for converting JPEG and PNG images to JPEG XL, but I don't think running the cjxl binary on the command line is a good solution for the general public. WordPress image processing traditionally uses ImageMagick or GD and while ImageMagick has added JPEG XL support already, it has not landed yet in GD.

Like WebP support right now, JPEG XL support will depend on the build flags used when building ImageMagick and GD. So we also need to wait until operating systems and web hosting companies start shipping JPEG XL support so it becomes mainstream.

from performance.

From what I can tell, this format is barely supported - could be cool to provide a sandbox for it and do some initial benchmark - couldnt tell from the trac ticket if that had indeed been done, but willing to give it a stab to see if its as good as everyone says it is!

from performance.

From what I can tell, this format is barely supported

True, and it isn't even clear if this format will ever see the light of day! I opened this ticket mostly for long term tracking, I expect AVIF will be the next format we work on.

from performance.

Agreed! It would be an interesting exercise to see the difference in improvement, if I get some time I will try setup a sandbox and compare JPEGXL, AVIF and WebP on a WordPress site!

from performance.

I agree with @joppuyo ... adding support for this format to the performance lab is premature. https://caniuse.com/jpegxl

I should think a standalone plugin would be a good way to proceed in late 2022. When and if browser support becomes widespread then the decision can be revisited.

from performance.

While I am a JPEG XL proponent, I don't really think it makes much sense to add JPEG XL support to WordPress until the format support has been added to browsers.

@joppuyo
And Chromium claims, that it wont add support until JPEG XL becomes more adopted in web ecosystem (that includes Wordpress). How do you propose we should get out of this paradoxical situation?

A claims they wont add JPEG XL until B does.
B claims they wont add JPEG XL until A does.

Seems like Catch-22.

from performance.