Clojurians Log v2
emacs 2026-01-20
Is anybody else noticing changes to emacs syntax highlighting recently? Mine recently (within the last 2 weeks) changed dramatically in a pleasing way, but it seems as if it does not work very well. It's like the algorithm for the highlighting was changed, but it's not quite working right. Does anybody have knowledge of what this could be?
I'm a little embarrased to just be sharing a screenshot of work-code here, but this screenshots shows what I mean.
If I revert the buffer, it looks like this, with the highlighting fixed/correct.
Some minutes later little errors like 1st screenshot come back.
to be complete, here are pointers to the problems I see:
• near the top with-meta goes from white to greet midway through the symbol
• In the comment below create-trip-state-changes! Accepts has the first 4 characters highlighting differently
Running GNU Emacs 30.2 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.24.50, cairo version 1.18.4) with latest versions of all my packages...
I've noticed that sometimes I have to re-enable clojure-mode to get the good highlighting
I'll see what this does if I see the issue again 👍🏽
maybe lsp-semantic-tokens-mode from lsp-mode?
I don't know what this mode does or how it is activated, but when I revert the buffer, this mode is enabled. If I disable the mode with M-x it reverts to a style that subjectively looks familiar to me... maybe you are on to something
I know there was a PR recently related to semantic tokens that's why I thought, you can try disable to make sure it's that via variable lsp-semantic-tokens-enable
Can you perhaps give me a pointer to the project; maybe they would appreciate a bug report or at least I could educate myself some
@sebastiansturm is the maintainer of that part in lsp-mode, it's the most complex elisp code I already seen so I know just a little bit of it, but he should be able to help