[email protected] (Alex V. Koval) writes: > Hello Michael and others,
Hi Alex, [your email address doesn't work] > I’ve just upgraded to 26.0.90 recently to do some pre-release testing of > new Emacs and I think I’ve found a bug. > > Expected behavior: > > 1. When I press Ctrl+C twice before TRAMP has been sending one Ctrl+C > signal to remote side. This was sending INT to the *local* ssh process, which has established the remote connection. It wasn't sent to the remote process itself. That's why the implementation has been changed. > 2. It does not do it now. It does complain though: >> tramp-interrupt-process: Process shell did not interrupt Well, this happens due to a timeout. Tramp waits for up to one second whether the interrupt succeeds. Most of the cases it is sufficient, but there might be situations it takes more time to interrupt, and shows you this error. OTOH I don't want to increase the timeout, it could make the default case worse. I've committed a patch to the emacs-26 branch, that Tramp falls back to the previous behavior if the interrupt doesn't show success in time. Could you, please, check? > WBR & Thank you for a great module which improved my work & life for many > years > Alex Best regards, Michael. _______________________________________________ Tramp-devel mailing list [email protected] https://lists.gnu.org/mailman/listinfo/tramp-devel
