Commander X16

On 10/17/2022 at 2:43 PM, codewar65 said:

Something like this, but over a simple TCP/IP packet instead of dialup?



*snip image*

Yeah, like that! Layout wise, that wouldn't be my first choice, but it would work well with the X16's mouse.

On 10/18/2022 at 2:57 PM, Tmp2k said:

Here's a random thought, seeing as the SD card just uses SPI, why not make a device with an SD card interface and a WiFi connection, then you could read/write data to files that are really piped elsewhere. Similar to serial etc in Linux. 



It would be quirky, but you could create a basic system that sends data to and from the X16 via WiFi. 



Might give this a go one rainy day. 

I was thinking along those lines too, where the X16 is just reading/writing files and the network adapter hardware handles the connection with the real world and updating the files. Admittedly, I don't have the expertise/knowledge to make the adapter or the X16 client software, but I could do the Linux server stuff (dime a dozen skills these days really, but that is good actually!).

Realistically, any old computer with a modem and NIC could host all the services, BBS and TCP/IP, and if all the data is stored in the same SQL DB it would be straightforward to make it available to the web page, BBS, and forum clients in their own way. Not real sure about a real time chat server.

Biggest question for me is, could it work with the emulator? If so, it's a project I would consider doing, because I like the Linux side of things and (obviously) I can afford the free emulator. :)

Found some topic related threads.


 

If there's any 8-bit parallel (RTL8029AS), SPI (KSZ8851), I2C (MCP23018) or even DMA assisted port using MII/RMII (any PHY). Then it's a matter of a simple chip to get Ethernet connection. Some chips like S-7600A even handle the TCP/IP. Or one can do the PPP thing over serial port UART.

Then exploit the graphics capabilities. Maybe someone dares to implement TLS..

Point is almost any I/O can be exploited to put an Ethernet interface on the X16.

 

On 10/19/2022 at 6:35 PM, neutrino said:

If there's any 8-bit parallel (RTL8029AS), SPI (KSZ8851), I2C (MCP23018) or even DMA assisted port using MII/RMII (any PHY). Then it's a matter of a simple chip to get Ethernet connection. Some chips like S-7600A even handle the TCP/IP. Or one can do the PPP thing over serial port UART.



Then exploit the graphics capabilities. Maybe someone dares to implement TLS..



Point is almost any I/O can be exploited to put an Ethernet interface on the X16.

There is already I2C. I have a long thread about SPI -- interrupted partway by news that VIA#2 would be on an expansion board, so the status of a motherboard based user port with the free pins from VIA#1 is not yet clear. Without the expansion board, I2C and SPI are the two live options.

Do these SPI or I2C options have DMA by any chance?

 

On 10/19/2022 at 10:32 PM, neutrino said:

Do these SPI or I2C options have DMA by any chance?

By no means.

This is an actual 8-bit system, not a 32bit system emulating an 8bit system. The I2C is reportedly around 150-175kbps. My personal 6502 assembly sketches makes SPI partially bit-banged on 3 VIA GPIO, a handshake pin and a VIA serial shift register at around 300-400kbps.

I think the X16 would benefit hugely from a DMA. Or at least some serial-to/from-parallel register like 74LS598 ( https://www.ti.com/lit/ds/symlink/sn74ls597.pdf ) to offload the heavy serial bit-bang.

 

Search the forum for "DMA" and you'll find a nice long discussion of all the ins and outs of that.

What was the td;lr consensus on DMA ?

 

That it's possible given the then-current slot specification to build an expansion card that could probably get the 65C02 to pause long enough to do DMA. But you'd have to 'be nice' and not hold the bus for too long or other things could start to be a problem, like VIA, I2C for keyboard and mouse, VERA, YM2151... and there's no intention of including any DMA in the system as it is.

Bluntly, neutrino, you're coming WAY late to the party, and offering up ideas and suggestions that are not going to get included because the system is in the final stages of development before manufacture. Things are not likely to change at this point in any major architectural ways. Even the removal of the second 6522 from the motherboard is not going to represent that much of a change in the way the system operates, but will only take away one of the 32-byte I/O channels (and free it up as an I/O channel for other uses) and require some addresses in the KERNAL to be revised.