Re: Possible bug in “kmous” in xterm-256color description

[Thomas Dickey]

On Mon, Aug 02, 2021 at 10:37:19AM -0400, Bill Gray wrote:
> On 8/1/21 5:29 AM, Thomas Dickey wrote:
> 
> > mouse-position reporting is rarely used, judging from comments
> > made by various people (it would be a performance issue on any remote
> > connection).
> 
>    With 'any-event' (xterm-1003) tracking,  you get events if
> the mouse moves at all.  With 'button-event' (xterm-1002)
> tracking,  you only get them if the mouse moves to another
> character cell.  1003 does get you a blizzard of events
> from even a small mouse movement.  Even if you're running
> locally,  it can be hard to keep up.  Events can be 'lost'.
> I have a program that enables any-event tracking,  so that
> it can detect if the mouse is hovering over a control and
> supply hint text... but the unreliability of mouse control
> in that situation has been a problem.
> 
>    Is there a way to blend 1003 and 1002 to ask for "any mouse
> event,  but only if we're in a new character cell"?  I think

That's terminal-specific (nothing that ncurses can do anything about).

xterm only (is supposed to) send events when there's a change.

That dates back to August 1998, so arguably other developers have
had a chance to test the feature.

Other programs don't necessarily do that (in a quick check,
I see that PuTTY, KDE Konsole send multiple events in a cell).

> most programs really don't care about mouse moves that remain
> in the current character cell.
> 
>    Alternatively,  perhaps mouse motion events could be capped
> at,  say,  20 a second ("we just reported a mouse move at time
> t0;  we've got another one at t0 + 0.02 seconds,  but don't
> report another move until t0 + 0.05,  as we may not be done
> moving yet").
> 
> -- Bill
> 

-- 
Thomas E. Dickey <dickey@invisible-island.net>
https://invisible-island.net
ftp://ftp.invisible-island.net

signature.asc
Description: PGP signature