On 11/03/2026, Arash Esbati wrote: > The function doesn't work as expected due to the check above. I think > the test should be adjusted. What do you and Al think? >
Ah I see. The purpose of this check was to avoid a loading error in case the package was not installed. This is especially relevant since the function gets called on point movement and signalling too many errors would be undesirable. Nevertheless, I agree that failing silently is not a good option either. We can do the following: - Add a setter code to `preview-at-point-placement` to require buframe and issue an error if it's not available. - To deal with the case of the variable being set directly (without calling the setter code), we can issue an error and override the value of preview-at-point-placement back to its default value to avoid further errors. What do you think? -- Al _______________________________________________ bug-auctex mailing list [email protected] https://lists.gnu.org/mailman/listinfo/bug-auctex
