A tweet led me to a pair of articles about a new file format called FUIF. That stands for “Free Universal Image Format.” Jon Sneyers describes it in a series of articles which so far include a Part 1 and Part 2.
It’s “responsive by design”; a single image file can be truncated at various offsets to produce different resolutions. Sneyers says FUIF meets JPEG’s criteria for a new format that provides “efficient coding of images with text and graphics” and “very low file size image coding.”
He has experience in designing file formats; his FLIF (Free Lossless Image Format) has gotten some good reactions from people who have noticed it. Getting noticed in the world of file formats is the hard part. No one develops software for a format that isn’t well known, and a format won’t become well known if there’s no software for it. That’s why we’re still stuck with decades-old formats like TIFF and JPEG, even though much better ones exist. From a preservation standpoint, it’s a good thing if formats don’t change too often, but it means progress is slow and requires getting the interest of big stakeholders.
The two articles published so far give only vague clues about how it works. This looks like one of the central ideas:
One of the main motivations for FUIF is to have an image format that is responsive by design, which means it’s no longer necessary to produce many variants of the same image: low-quality placeholders, thumbnails, many downscaled versions for many display resolutions. A single file, truncated at different offsets, can do the same thing.
This may mean that the pixels are distributed through the file in such a way that one truncation point gives you a low-resolution version, and subsequent truncation points build up the image with more pixels, more bits in the same pixels, or both. It might mean something I haven’t thought of. So far there just isn’t enough information for me to tell.
It’s too early to tell where FUIF is going, if anywhere. But you may have heard about it here first!