• Please review our updated Terms and Rules here

Win 3.0 Serial Mouse - Only Moves when Hourglass Cursor is Displayed

bensinc

Experienced Member
Joined
Jul 22, 2014
Messages
64
I'm trying to figure out an odd issue on a pretty weird setup. I'm tinkering with Windows 3.0 on a Zeos Pocket PC, which runs surprisingly well. Unfortunately I can't seem to get a serial mouse to work right. It's a Logitech 3 button, and it works fine under DOS using the cutemouse driver. On Windows 3.0 using the MS driver, and an older Win 2.x MS driver, it only works while the hourglass cursor is displayed! When it's the normal pointer nothing works, not even the buttons.

Windows 3.0 does have an included "Logitech Serial" driver, but Windows never loads when I use that one.

Does anyone know what I might check? I know this is a pretty crazy setup! Thanks!
 
Could you post what MSD reports for the address of the serial port? IIRC, the MS mouse drivers had problems if the serial port wasn't set to the usual place.

Does the Windows driver fail when cutemouse is not loaded? Just in case cutemouse is intercepting something.

I had a Logitech mouse with my Windows 3.0 setup on a generic AT clone. Worked fine so I would suspect something specific to the Pocket PC. https://gadgets.notjakob.com/forums/ is a forum that specializes in the palmtop machines including the Tidalwave model that Zeos rebadged.
 
MSD shows COM1 at 0x3F8 and IRQ 4, which I believe is normal. MSD even detects the mouse as a Logitech and it controls the cursor in MSD just fine! I guess MSD has its own mouse driver, as I hadn't loaded cutemouse beforehand.

I have tried Windows without the cutemouse driver, and it doesn't seem to make any difference. I actually did cutemouse after finding this issue just to verify the mouse did work.

It's just so weird that it works while the hourglass cursor is displayed, but stops otherwise. Open something from Program Manager and you get to move the mouse while you wait, but that's it!

I also just found an old copy of the real Logitech driver disk, but it has the same behavior.

Thanks for the forum suggestion too!
 
Last edited:
MSD shows COM1 at 0x3F8 and IRQ 4, which I believe is normal.
Years ago, there was a thread in which an information reporting tool reported 3F8/IRQ4, but it turned out that the hardware was using a different interrupt. The tool must be assuming, 'All serial ports at 3F8 are set to IRQ4, so don't bother checking, just report IRQ4'.

There is a DOS tool named SERTEST at [here]. After selecting 3F8 as the port to test, the 'Interrupt test' menu option will see if either IRQ3 or IRQ4 responds.
 
Thanks, I’ll check that out! But, if it were on the wrong settings, why would it work when the hourglass cursor is displayed?
 
Thanks, I’ll check that out! But, if it were on the wrong settings, why would it work when the hourglass cursor is displayed?
I think we have to be careful in ruling out things based on limited/preconceived knowledge. Do we really know what is going on 'under the hood'? For example, someone saying to themself, "The CMOS battery is only there to preserve the CMOS configuration when the computer is powered off. Therefore, a dead battery in the Dallas CMOS module cannot possibly be the reason why the CMOS SETUP screen is not letting me type in values."
 
I ran sertest.exe, and the interrupt test did show IRQ 4 responding for com1 at 03F8. I'm guessing it's just some weirdness with this machine. I'm betting there are very few people that have tried running Windows on this!

If I keep alt+tabbing around I can use the mouse briefly. I just have to keep the system under load! I wonder if disk access has something to do with it, like the serial port interrupt isn't working unless something else is going on. It's probably not worth debugging much further!
 
Hey, I got it! It turns out that there is a "power save" option that was on! I'm guessing when under load it turns off automatically, and when it's idle it does something that saves power but messes with the serial port.

So now I'm enjoying Windows 3.0 solitaire in all of its B&W CGA glory!
 
Hey, I got it! It turns out that there is a "power save" option that was on! I'm guessing when under load it turns off automatically, and when it's idle it does something that saves power but messes with the serial port.
We see weird stuff. Like the member who had floppy drive issues, that were fixed when he/she turned off the CMOS option of "video cacheable option".
 
Back
Top