|
Post by dshadoff on Dec 1, 2019 22:12:54 GMT
Are you talking about a 'standard' device ? (Kind of curious what you plan to do with the keyboard if it did exist)
There are many different ways a keyboard can communicate...
Back in the days when this machine was out, keyboards on home computers were often a row-column matrix which you would scan row-by-row, in order to determine which key(s) were pressed. I'm assuming that you don't want that kind of overhead... are you looking for: a) a scan code or an ASCII value ? b) notification of press and release, or one-shot ?
And lots of ways to interface, not just through the joypad port - since we are talking about custom hardware now.
|
|
|
Post by elmer on Dec 2, 2019 6:35:00 GMT
Are you talking about a 'standard' device ? (Kind of curious what you plan to do with the keyboard if it did exist) Well, I can see there being *some* fun in turning the PCE into a fully-fledged computer, in just the way that Hudson originally envisioned as a possibility. So having the possibility of a keyboard and mouse, while still having access to the remaining 3 ports on a TurboTap, would be kinda nice. With the TED2 we've got extra RAM, and the SD card as well to act as persistant-storage to replace a floppy, or whatever would have been appropriate for the late 1980s. It's a rather silly fantasy, but I find it a lot more appealing and true-to-the-times than the internet-personalities and Commodore-fans currently pretending that they're creating a what-if successor to the C64 that somehow uses an 8MHz or 16MHz 65816 and 20ns RAM that just didn't commercially exist until the late 1990s. There are many different ways a keyboard can communicate... Back in the days when this machine was out, keyboards on home computers were often a row-column matrix which you would scan row-by-row, in order to determine which key(s) were pressed. I'm assuming that you don't want that kind of overhead... are you looking for: a) a scan code or an ASCII value ? b) notification of press and release, or one-shot ? And lots of ways to interface, not just through the joypad port - since we are talking about custom hardware now. I don't see that there would be much reason to follow the Tsushin keyboard protocol, it's rather ugly IMHO. Just a nice-and-simple PS/2-style keyboard protocol would be fine, buffered to the keyboard by a cheap PIC or ATMEL microcontroller, with bytes sent to the PC Engine using a 2-stage read process just like a 6-button joypad. Heck, there are still unused high-nibble ID values available for such a protocol. How about $Fx for the lo-nibble of a PS/2 keycode, and $8x for the hi-nibble of the PS/2 keycode. That would be instantly recognizable by software on the PCE, since a $8x code is still illegal from a 2-button joypad, and it hasn't been used anywhere else. You could probably even read multiple bytes in a single 1/60 if you wanted, by adding a FIFO chip, although I'm not sure that it would actually be needed in practice.
|
|
|
Post by turboxray on Dec 2, 2019 18:59:03 GMT
Ohh,, I could think of quite a few projects that could use the keyboard haha. Realtime sound/music editor on the system (keyboard would just make it easier), Dungeons of Daggorath style game!, etc.
|
|
|
Post by dshadoff on Dec 2, 2019 19:25:54 GMT
Well, let me see how the PCEmon accelerator goes. It’s not a big leap to make a new version which uses all of the bits and a different polling routine... but one thing at a time.
Actually, when I make more progress on the bluetooth LE mouse, I’ll see if I can do something about keyboard as well at that time. But I might ask somebody else to handle the PCE-side read routine, as I would probably not be thinking about the right use cases.
|
|