Developing on your own machine.

[sandras]

Anyone out here developing OSes or bare-metal applications for virtual or physical machines of their own design? I myself haven't got much to show but I'm looking out for ideas.

[mathematician]

I doubt if many programmers would have the know how, or the means, to design a "machine" consisting of much more than a microcontroller and a few flashing LEDs.

[jnc100]

Yeah, I did this once with a 32-bit risc cpu in a fpga with associated peripherals (serial, vga, sd card). Obviously you need to implement an entire toolchain for it too. Others have been mentioned. Searching the forums for 'fpga' or similar will probably get you some results.

Regards,
John.

[jojo]

I doubt if many programmers would have the know how, or the means, to design a "machine" consisting of much more than a microcontroller and a few flashing LEDs.

Honestly, that's a touch rude. And for what it's worth, I'm personally in the middle of building a retro-y system based around a WDC 65816. As far as I'm concerned, digital logic and basic circuit design, while requiring knowledge of some different minor details, pull from fundamentally identical skill sets. One's just more physical and one's more ethereal.

Seriously, if someone has the drive to write an operating system you think they're too dumb to slap a couple of SRAMs and a CPU on some protoboard to make an SBC?

By the way, John, that is dope as all get out.

[Schol-R-LEA]

Too dumb? No. Possessing the requisite knowledge, experience, time, funds, and desire? That's another story.

I agree that Mathematician worded it poorly, but I doubt it was meant the way you think it was. More that it wasn't something most of the people here would bother with, and if they did, they would do a fairly simplified system based on a PLD rather than, say, from TTL components, or MSI circuits, or Tinkertoys, or hand-wrapped electromechanical switches made from stainless steel paper clips (I actually read part of that book sometime around 1981, BTW, having found it in the local public library).

Or, you know, get a copy of a circuit design and simulation program such as Designworks, EAGLE, or Oregano and simulate it. That would be a good idea anyway, and not all that hard (though from the courses I took on computer architecture, which involved implementing about 95% of a MIPS system using Logicworks, I can tell you it isn't the easiest thing to do, either, even working at a fairly abstract level of design).

I think the will, time and money issues are the biggest ones. How many here want to build hardware? I am guessing less than a quarter, even given the heavy DIY focus of this group. How many could afford the material? Few still I will guess. How many would have the time to work on it? Hardly any.

I agree that sneering at someone who builds a SBC from an existing CPU and calls it an original computer design is pretty harsh, though.