Crossposted-To: comp.windows.x
From: olsen@HING.LCS.MIT.EDU (James Olsen)
Subject: xfree86-1.1 & Logitech Trackman: no middle button
Date: Mon, 19 Oct 1992 15:48:03 GMT
I'm running the xfree86-1.1 X386mono server under Linux 0.98 on my
laptop with a Logitech TrackMan trackball, and it's running just fine
except for one problem: the middle button isn't being recognized.
My Xconfig is set to: Microsoft "/dev/ttys1". From the source
code, it appears that the xfree86 mouse driver expects an extra
(fourth) byte in the data packet from the mouse when the midddle
button changes state, of hex value 20 or 00, depending on whether the
middle button is pressed or not.
Looking at the trackball packets directly from the serial port at 1200
baud (immediately after aborting the X server), the fourth byte value
appears to be poorly framed, being usually f4 or d1 when the middle
button is pressed, and c0 or f0 when it is released. I can sometimes
get the trackball into a state where it sends the expected 20/00 byte,
but it's never in this state after the X server has been running.
Does anyone know about this problem? Unless I'm missing something
(which is highly probable), there is some problem with setting up the
trackball into the state where it sends the expected data packet
format. Any help would be greatly appreciated.
--
Jim Olsen olsen@mit.edu "Tache d'etre heureux."
From: chrisz@tabaqui.informatik.rwth-aachen.de (Christian Ziemann)
Crossposted-To: comp.windows.x
Subject: Re: xfree86-1.1 & Logitech Trackman: no middle button
Date: 20 Oct 92 10:42:41 GMT
olsen@HING.LCS.MIT.EDU (James Olsen) writes:
>I'm running the xfree86-1.1 X386mono server under Linux 0.98 on my
>laptop with a Logitech TrackMan trackball, and it's running just fine
>except for one problem: the middle button isn't being recognized.
>My Xconfig is set to: Microsoft "/dev/ttys1". From the source
>code, it appears that the xfree86 mouse driver expects an extra
>(fourth) byte in the data packet from the mouse when the midddle
>button changes state, of hex value 20 or 00, depending on whether the
>middle button is pressed or not.
>Looking at the trackball packets directly from the serial port at 1200
>baud (immediately after aborting the X server), the fourth byte value
>appears to be poorly framed, being usually f4 or d1 when the middle
>button is pressed, and c0 or f0 when it is released. I can sometimes
>get the trackball into a state where it sends the expected 20/00 byte,
>but it's never in this state after the X server has been running.
>Does anyone know about this problem? Unless I'm missing something
>(which is highly probable), there is some problem with setting up the
>trackball into the state where it sends the expected data packet
>format. Any help would be greatly appreciated.
>--
>Jim Olsen olsen@mit.edu "Tache d'etre heureux."
OK, eventually someone got the same problem as I did. I called Logitech,
Switzerland, and I also mailed a message to Thomas Roell, so this is my
collection of info:
1) There are *different versions* (USA / Europe) of the TrackMan and
MouseMan, so some of us do experience the problem while others don't.
2) The following info goes for the *European* version. (I don't know
exactly in which way the US version differs, but maybe this helps.)
- all commands to the Track/MouseMan have to be sent at 1200 baud,
so you have to select the protocol first, then switch to 9600 baud,
if desired.
- whenever the Track/MouseMan receives a '*' (indicating the beginning
of a command sequence) it switches back to 1200 baud.
- only the following protocols are supported:
MouseSystems 5 byte protocol (3 button mouse)
Microsoft 3 byte protocol (2 button mouse)
Microsoft 3 + 1 byte protocol (3 button mouse)
The 3+1 protocol is the default. It is also used by the Xserver, as
one can see in mit/server/ddx/x386/common/x386Io.c. It should work,
as long as you use 1200 baud only, but there is another problem (see
5, below).
3) Only 1200 baud and 9600 baud are supported. (Old Logitech mice supported
2400 baud, too.)
4) Selection of a sample rate is not supported.
5) My (European) TrackMan sends a 0x21/0x01 pair for the 3rd button, while
the MouseMan sends 0x20/0x00.
Now, what is the bottom line? Well, I created a diff and mailed it to Thomas
Roell a couple of weeks ago, but I guess it just wasn't in time to make it
into the XFree86 1.1 release. I will mail another copy of that diff to the
XFree86 team, hoping to get a decent solution with the next official patch.
Oh, BTW, I am not (yet) a linux user. I experienced the problem on an ISC
2.2.1 machine, but I think it doesn't look too machine specific.
I hope this helps a bit.
Chris (chrisz@pool.informatik.rwth-aachen.de)
Crossposted-To: comp.windows.x
From: dwex@cbnewsj.cb.att.com (david.e.wexelblat)
Subject: Re: xfree86-1.1 & Logitech Trackman: no middle button
Date: Tue, 20 Oct 1992 13:10:49 GMT
In article <chrisz.719577761@tabaqui>
chrisz@tabaqui.informatik.rwth-aachen.de (Christian Ziemann) writes:
[ description of Logitech TrackMan problems deleted ]
> Now, what is the bottom line? Well, I created a diff and mailed it to Thomas
> Roell a couple of weeks ago, but I guess it just wasn't in time to make it
> into the XFree86 1.1 release. I will mail another copy of that diff to the
> XFree86 team, hoping to get a decent solution with the next official patch.
>
Thomas Roell has no connection with XFree86 except as a friend and
sometimes confidant. If you have problems with XFree86, report them
(and any fixes - save us some work :->) to <xfree86@physics.su.oz.au>.
If it is a bug in X386 as well as XFree86, we will report it to SGCS.
The flow of information between the two groups is fairly small, as one
would expect. They are in business to make money, and we are a competitor,
although a relatively friendly one.
The bottom line is that this is the first we have heard of this, so send
the patch to the above address, and it go into the next update.
> Oh, BTW, I am not (yet) a linux user. I experienced the problem on an ISC
> 2.2.1 machine, but I think it doesn't look too machine specific.
>
> I hope this helps a bit.
>
> Chris (chrisz@pool.informatik.rwth-aachen.de)
>
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
David Wexelblat | dwex@mtgzfs3.att.com | Somebody get me a
AT&T Bell Laboratories | ...!att!mtgzfs3!dwex | cheeseburger!
200 Laurel Ave - 4B-421 | |
Middletown, NJ 07748 | (908) 957-5871 | --Steve Miller Band
Crossposted-To: comp.windows.x
From: chrisz@tabaqui.informatik.rwth-aachen.de (Christian Ziemann)
Subject: Re: xfree86-1.1 & Logitech Trackman: no middle button
Date: 21 Oct 92 12:57:04 GMT
dwex@cbnewsj.cb.att.com (david.e.wexelblat) writes:
[ my own previous stuff deleted ]
>Thomas Roell has no connection with XFree86 except as a friend and
>sometimes confidant. If you have problems with XFree86, report them
>(and any fixes - save us some work :->) to <xfree86@physics.su.oz.au>.
>If it is a bug in X386 as well as XFree86, we will report it to SGCS.
>The flow of information between the two groups is fairly small, as one
>would expect. They are in business to make money, and we are a competitor,
>although a relatively friendly one.
David, the problem was that I found the bug before I got to know about
XFree86. So it was originally a bug in X386 (that's why I reported it to
Thomas Roell) and now it's also a bug in XFree86.
I did in fact expect that Thomas would at least give you a hint, because
he claimed to have fixed the bug a long time ago. Otherwise I would
definitely have sent you my patch much earlier - sorry.
>The bottom line is that this is the first we have heard of this, so send
>the patch to the above address, and it go into the next update.
OK, please wait for the patch coming via e-mail probably tomorrow.
Chris (chrisz@pool.informatik.rwth-aachen.de)