Option "Emulate3Buttons" comes to mind. but how would you set it if you
dont have an XF86Config?

On Fri, 4 Jan 2002, Greg Hughes wrote:

>
> On Fri, 4 Jan 2002, Andrew Diller wrote:
>
> <snip>
>
> > Now -- one last thing: how can you emulate the 3rd mouse button in X?
> >
> > Blackbox really needs that 3rd button, and I would like to cut and paste.
> > Anyone know a good way?
>
> You can emulate the 3rd button at two levels -- in the mouse driver or in
> the X server.  The X server is probably the better of the two places to do
> it (truthful hardware and all that).  I have no idea if it's supported
> already (ala XFree86) or how difficult it would be to add it.  (sigh... so
> little time...)
>
> Adding it in the kernel mouse driver would probably be more of a headache
> than it's worth since you have to worry about things like how quickly the
> buttons need to be pressed to be considered a chord, or interfacing with
> the keyboard driver if you want a key-click modifier for it.  These kinds
> of things are really UI issues that should be taken care of by the X
> server or even the window manager.
>
> > Lastly, I'm amazed at Greg's Aug 17 TP kernel. It works amazingly well on
> > my z50 with 1.5.2. I'm even pulling down a large package (mozilla) and
> > sshing into the z50 and the cursor behaves very well except during high
> > network activity. The disk writes don't seem to bother it.
>
> Thanks for the moral support.  :)
>
> The mouse gets flakey under heavier load because the driver is missing
> interrupts and gets serial overruns on the TP port.  The VR4121 CPU
> supports a serial debug port with a 1-byte buffer which is what the TP is
> connected to.  Unfortunately, because the buffer is 1 byte long (i.e.
> there is no FIFO) if you miss any interrupts you start dropping bytes.
>
> I really don't know why this is the case, however, since WinCE seems to
> handle the mouse alright and I'm leaving the TP in the mode WinCE
> configures it to be in (which is a streaming mode at a fixed rate).
> Perhaps NetBSD just takes longer to handle PCMCIA interrupts than does CE?
>
> It should be possible to eliminate such issues by rewriting the mouse
> driver to set the TP to polling mode.  In this mode, it is possible to
> request a resend of the last data packet if an error (such as
> overrun) occurs.  However, I would like to find out how WinCE can work
> (presumably) in streaming mode rather than implementing this 'polling
> hack'.
>
> > Now I just need to be able to use the two buttons on the z50 as a chorded
> > middle button and I'm ready to really use this.
>
> I suspect it should just be a configuration option in the X server....  In
> any case, I would be interested to know how to get this to work as well (I
> don't have the time to investigate myself at the moment!)
>
> - Greg Hughes
>

-- 

John L. Utz III
john@utzweb.net

Idiocy is the Impulse Function in the Convolution of Life