> JPEG XL is the future of all image formats. It is a new royalty-free image codec targeting the image quality as found on the web, providing about 60% size savings when compared to original JPEG at the same perceptual quality, while supporting modern features like HDR, animation, alpha channel, lossless JPEG recompression, lossless and progressive modes. It is based on Google's PIK and Cloudinary's FUIF, and is an ISO standard.
It can even losslessly recompress existing JPEG images (though Tiki could probably get even better-optimized results by re-generating all the thumbnails sizes to do it "cleanly" without worrying too much about backward compatibility).
Explanation here: https://www.slideshare.net/cloudinarymarketing/imagecon-2019-jon-sneyer
Reference implementation here: https://github.com/libjxl/libjxl/
Adoption:
--
The JPEG XL community is very motivated and vocal. If they see you making progress about it, they will retweet (and they monitor any mention of jpeg xl or #jpegxl on twitter, and they also have a subreddit). If you need help, their experts will happily advise. There's at least two of them who are very easy to identify due to their Twitter activity.
--
Marketing it:
- Once thips is implemented in Tiki, there needs to be a (joint?) press release announcement of some sort to leverage this, and again when it lands in a stable version, and again when it becomes enabled/enforced by default
- Being featured on the list of apps that implement it, in https://en.wikipedia.org/wiki/JPEG_XL#Official_support, and thus crosslinking to the Tiki wikipedia page, would make a lot of sense.
Timeline:
- I think it would be worthwhile for Tiki to prepare for this and maybe announce it in time for the 20th anniversary. Once it is supported across browsers (see https://caniuse.com/jpegxl particularly its "notes" and "resources" tabs) I believe it should be the default public-facing format for Tiki.
- No rush to do it "urgently this week" however, as the EvoluData website is not ready so we don't really have a nice home to receive public attention
- If we miss Tiki's 20th anniversary timeframe/deadline for this, no big deal, it could come in later as its own big piece of announcement, though in that case it would be PR-beneficial to publicly announce it as being on our roadmap. Not just for PR, but also commenting on browsers' bug trackers upstream to "show traction" to help convince them that they should indeed intend to deploy the damned thing into production. We would be joining big guys like Adobe, Facebook, etc. in publicly stating our support, which certainly counts for something.