Dear all, I’m experimenting with using Claude to develop TeXmacs. I do no know what are the feelings of the others developers in this direction, so I kept my AI work in my github (https://github.com/mgubi/texmacs <http://github.com/mgubi/texmacs>) and haven’t committed anything to Savannah.
There are various directions: 1) I’m sweeping over all the bugs in Savannah which are not assigned and let Claude analyze, test and propose PR for them. (filed as GH issues) 2) I’m writing new test suites and in the process finding bugs (branch wip_more_tests) 3) I’ve completed the OpenType MATH support (branch wip_opentype) 4) I’ve added some support for variable fonts on top of wip_opentype (branch wip_variable_fonts) 5) I’ve added a new Cocoa backend and finished my work on a immediate GUI backed on top of SDL called Vue (branch wip_other_guis) + new set of icons 6) I’ve packaged many of these improvements and adapted to work in the browser via Webasm (branch wip_wasm_vue). The app is called TeXmacs Vue, and can be tested here: https://mgubi.github.io/texmacs/ In the process I’m recording the issues in the GH issue tracker as mostly they are generated by Claude, so I want to keep them separated by the bugs in savannah. Overall I’m quite satisfied, but I want to review the code before proposing any patch to TeXmacs and even before this discuss together what are the policy of the project wrt. to AI content. Would be happy to hear what you think and agree on some workflow and rules. Best, Max
_______________________________________________ Texmacs-dev mailing list [email protected] https://lists.gnu.org/mailman/listinfo/texmacs-dev
