Re: FYI: Practical uses of OpenType fontes as another higher level protocol
"Joao S. O. Bueno via Unicode" <[email protected]>
| Newsgroups | gmane.text.unicode.general |
|---|---|
| Message-ID | <CAH0mxTSBTuT_4R2J+v5mTL-PXwQ2qQcR-26dnKSDQb0UvMhgBQ@mail.gmail.com> |
On Wed, Aug 21, 2024 at 2:13 PM Mark E. Shoulson via Unicode <[email protected]> wrote: > > Be careful. OpenType fonts are so powerful these days they are > essentially Turing-complete and can become an arbitrarily complex > "protocol". See https://fuglede.github.io/llama.ttf/ which is a > complete Large Language Model AI stuffed into a font, renderable by > (almost) ordinary tools. It's an extreme case, and meant to be such, > but still. Sure. Anyway, a lot of the (~30 YOLD) recent font file formats would be turing complete themselves - maybe lacking the ability to perform any I/O out of glyph parameters - but "type 1" fontes were full postscript programs, weren't they? > ~mark > > On 8/20/24 08:56, Joao S. O. Bueno via Unicode wrote: > > Hi - > > > > Just got this news yesterday which I believe might be of interest for > > some of the participants on this list - > > People had successfully made use of mechanisms in OpenType fonts > > to be able to colorize text according to context. The practical use > > is for simple syntax highlighting. > > > > https://blog.glyphdrawing.club/font-with-built-in-syntax-highlighting/?utm_source=tldrnewsletter > > > > This is a protocol sitting between unicode and markup languages, > > at the cost of having to be properly configured in any displaying app - > > (however, any app using proper libraries for rendering the fonts, and > > enabling font selection and parametrization would work "out of the box") > > > > > > This can also work as one more item to list in the "proper ways to do it > > in higher protocols" when > > people involved or interested in Unicode are queried about > > including mechanisms such as enabling color attributes. > > > > Regards, > > > > Joao