Floppy driver board

Nobody gets to berate Rick round here. That’s my job.

😂👍👍👍

If using SWIs and Vectors is to be considered “linking” is it possible to actually run GPL software on RO?

I should point out that when writing that, I had the thought that if the answer to that was no, then how many OS’s out the could you run GPL code on?
Then turn that around and ask whether the different parts of a Linux distribution are actually compatible with each other, and can people legally run a standard installation?
Yep, deep down you all know it’s bull.

That said, and steering away from the bull, how is everybody’s search for decently licenced code and a viable floppy reader for modern RO kit going?

Edity:
In physical build terms, I’d be inclined to use something lower power than the original 74xx series though

That said, and steering away from the bull, how is everybody’s search for decently licenced code and a viable floppy reader for modern RO kit going?

I thought that was said and done a long time ago (like, on page 1 of this thread): GreaseWeazle. All code is MIT licensed. Whether the RISC OS port of Python is up to it remains to be seen – serial port access is needed, and the RISC OS USB Serial driver needs to be compatible with the GreaseWeazle hardware. But this is not the use case I am aiming for, so I won’t be the one trying that out.

In physical build terms, I’d be inclined to use something lower power than the original 74xx series though

If you avoid 74 series stuff, you’ll have to come up with a more complicated and expensive driver to drive the termination resistors – and, yes, they are necessary, unless you’re going to compromise cable length – and, maybe, reliability.

Transmission lines. Been there, done a lot of that. Do them properly, or risk unpredictable consequences.

Do them properly, or risk unpredictable consequences.

Unless you’re talking about SCSI in which case it’s a black art.

Back in the day, this is the setup that worked for me:

A5000 &          Tape [*]       hard
Morley [*] ----- Streamer ----- disc [ ]

* = terminating resistors

Did lots of tests. Each device worked fine when by itself and terminated. But all three, only worked correctly when the middle device was terminated and the end one wasn’t. Otherwise there were all sorts of errors (if I recall “no sense” was the most common).

I know, that defies everything ever written about SCSI. I even managed to get my local(ish) library to order the SCSI spec from a reference library somewhere weird like Belfast ¹ and it was a slog to read (especially as it had to be read in the library) but it said the same thing the Morley user guide did – each end of the chain.

<shrug>

¹ Belfast is “somewhere weird” when your library is Woking!

each end of the chain.

For any bidirectional shared bus, that’s the only sensible configuration.