There are no guidelines, but there are existing users, who will find this change "disruptive and unintiutive" on Windows.. I'm a bona-finde user who uses this OS everyday, and our tools there, too. I also happen not to have a Break on my keyboard, and it seems that Break is now well hidden on keyboards. Ctrl-D does not work. Ctrl-Z does not work. I would rather not not write "quit". So, of all Windows users we have feedback so far, none found the old behavior disruptive, and the all vocal ones found this proposed change disruptive.

So here is the deal - can we do exactly nothing about it Windows in 10.4, or whatever old version it was predestined to be fixed. It would be impolite to break tools in the very last version before release is going EOL. In 10.11+ I'll file a followup, to break mysql on Ctrl-D or Ctrl-Z or both. Because this version already uses ReadConsoleW, making it possible to catch Ctrl-<something> for <something> that is not "C" (via last parameter, CONSOLE_READCONSOLE_CONTROL , which only works in Unicode "W" version)

• agree about 10.4
• I couldn't understand whether you agree with the suggestion 'Ctrl-C and Ctrl-Break do not exit the tool, Ctrl-D does", could you state that explicitly, please?

Yes, Ctrl-D interrupts, Iood enough to me, although python uses Ctrl-Z on Windows some reason. I have no idea about Ctrl-Break, I can't generate it, so I do not care either way.

Checked statement: "cmd.exe, powershell, ftp, netsh — none of them exists on either Ctrl-C or Ctrl-Break." netsh does actually

Implemented this by only changing Linux behaviour for CTRL-C and adding some extra windows support for non-interactive signals sent by the OS during terminal close, log off or shut down, like in MySQL's client implementation.