On 2025-01-06 10:25, Stefan Monnier wrote:
<feedback trimmed>

Thanks for the feedback. Patch attached with most feedback attached.  
Not addressed feedback follows:

> - The above enables `xterm-mouse-mode` which will
>   then proceed to *not* obey `xterm--auto-xt-mouse-allowed-*` on any
>   further ttys used in the same session.
> 
>   IOW, I think the above test belongs in
>   `turn-on-xterm-mouse-tracking-on-terminal` rather than here.
> 
> - Maybe instead of `xterm-mouse-mode-called` we should test
> 
>     (eq xterm-mouse-mode 'when-safe)
> 
>   or maybe rename the var to something more explicit like
>   `xterm-mouse-mode-only-when-safe`.

Yeah, I wanted to implement auto enabling like this, but could not 
figure out a way that seemed ergonomic to a user.  The problem I could 
not solve is that I wanted to ensure that a user running M-x 
xterm-mouse-mode still toggled from whatever auto-enabling logic ended 
up choosing.  That is, if xterm-mouse-mode was auto-enabled, running M-x 
xterm-mouse-mode should disable it and if xterm-mouse-mode was not 
auto-enabled, running M-x xterm-mouse-mode should enable it.

> - Of course, the previous code was also very weird:
> 
>     (when xterm-mouse-mode (xterm-mouse-mode 1))
> 
>   says to enable the mode, but only when it's already enabled?!

Yeah, it's weird. I followed the pattern from term/linux.el.  That file 
has the exact same pattern for gpm-mouse-mode, for the exact same 
purpose.

   -- MJF