Won't it be quite a few parts to run something like a hc165? you have to clock it a bunch of times to run the register out, and that probably means a real oscillator to be consistent enough to rely on timing alone to read the bits on the 100.
Plus it would use that clock to poll the buttons to tell the difference between pressing and holding so it can send again on release. And if you're relying on the interrupts instead of polling on the 100, then the stick better arbitrarily generate a resync packet periodically, just resending the current state whatever it is even if no state changed. In the end the 100 is still polling, just now being driven by the stick, except possibly less often. This assumes part of the goal is to avoid just throwing a magic modern part in the joystick. If that's allowed then there are whole little microcontrollers that can read all the pins and output full rs232. Bcr would still be cool though because it leaves all the more useful ports free, uses a port that otherwise gets very little use, provides the power needed, and unlike the bus port, every version of the machine except KC-85 has one and uses the same connector. (kc85 has no bcr port), so it would work on Olivetti and NEC and both 100 and 200. bkw On Wed, Sep 24, 2025, 7:27 AM Stephen Adolph <[email protected]> wrote: > Both are great approaches. What's nice about using the barcode reader port > is when the joystick is static, no CPU effort is expelled to manage the > joystick. If you use the parallel port, you would need to add a task to > the 8085 to regularly poll the parallel port. > > > On Tue, Sep 23, 2025 at 11:31 PM Kenneth Pettit <[email protected]> > wrote: > >> Oh yeah, I suppose you could do that. Just be sure to do it with >> interrupt disabled because those outputs are also used for other purposes. >> >> Ken >> >> On 9/23/25 8:19 PM, Mike Stein wrote: >> >> So, can't you scan it like a keyboard, i.e. output one of the 8 bits at a >> time and check if it appears on the input pin? >> >> What am I missing? >> >> m >> >> On Tue, Sep 23, 2025 at 11:08 PM Kenneth Pettit <[email protected]> >> wrote: >> >>> The printer port is output only except for 2 inputs: BUSY and /BUSY >>> >>> Ken >>> >>> On 9/23/25 7:59 PM, Mike Stein wrote: >>> >>> Couldn't you just scan the printer port looking for 1 out of 8 switch >>> closures?. >>> >>> m >>> >>> On Tue, Sep 23, 2025 at 9:42 PM Scott McDonnell < >>> [email protected]> wrote: >>> >>>> I am back to thinking about a joystick for the Model 100. I have a >>>> specific reason in mind and even a specific joystick. >>>> >>>> My intent is to target the Atari/Commodore 64 digital joystick >>>> interface >>>> mainly because my inspiration is the Suncom Icontroller which mounts to >>>> the side of the machine. >>>> >>>> This interface is a series of 5 discrete switches to control direction >>>> and a fire button. Very basic. >>>> >>>> While the parallel port would have been ideal for something like this, >>>> it is not bidirectional and offers only two inputs. >>>> >>>> The serial port provides 3 inputs; still not enough >>>> >>>> While some multiplexing strategies might allow this to work, both of >>>> those ports are also often used for other important purposes. >>>> >>>> So I started thinking about the barcode port. It only has one input, >>>> but >>>> the joystick could be encoded into a serial stream. Maybe even using >>>> the >>>> original barcode drivers. I don't know the specifics of the BCR >>>> implementation here, but often they are just keyboard wedges. >>>> >>>> Any thoughts on this or a better solution? >>>> >>>> >>>> >>> >>
