Can you use a USB floppy drive with RISC OS on a Pi?

This seems to crash the USB stack as I lost keyboard and mouse

The USB stack has never been a strong point in RISC OS. In my experience hot-plugging devices via USB randomly kills the Mouse and Keyboard. I’ve got into the habit of powering the thing down before plugging anything in!

I don’t recall if I’ve tried a CD though, so it might be a special case that’s guaranteed to kill the USB stack.

I think we can conclude that “RISC OS doesn’t currently support USB CD_ROMS” either.

I found that, in the general case, USB CD/DVD/BD drive support in RISC OS mostly works. Never had a scenario like yours (in contrast to card readers and USB sticks, which sometimes block the whole OS until you remove them again).

Of course CDFS is still able to lock itself when it encounters a logical format it does not like, so I nearly always use CDVDBurn to access those drives because I know what it does, and nobody knows what CDFS does.

Could possibly the Linux USB-lib-thingy be ported to RISC OS so that it “just” would be a matter to write drivers for the odd un-supported device? Soon there’ll be USB toothbrushes. or they may even be on shelves as I type..

Could possibly the Linux USB-lib-thingy be ported to RISC OS so that it “just” would be a matter to write drivers for the odd un-supported device?

This would not help in the slightest with using a USB floppy drive to access FileCore format discs.

I have long thought the only way round that is co-opt an Arduino into the gap between the pi and the disc drive.
Keep it simple make it look like a USB serial device, initiate a filecore instance with a matching driver that understands your remote serial sector hex dumper on the Arduino, over complicate it with writing as well if you dare. I have plenty of working original equipment so I have no imperative to actually do this.

Very good idea! Can someone please make this?

Risc OS for Raspberry Pi with USB Floppy drive support and access to FileCore format discs, would be great.

Thank you.

Micky

There are two different approaches pointed too – one what looks like hardware / software that claims to able to read Arcs discs.

The other using a WD37C65C floppy controller.

I wonder if one could use the Pico w like -Colin is using to drive
Bluetooth mouse keyboard.

What is with a Arduino? There are some floppy controllers too. But I think that they are to slow.

https://blog.adafruit.com/2021/06/25/read-floppy-disks-on-a-raspberry-pi-with-bbc-fdc-piday-raspberrypi-raspberry_pi-femtosonic/

https://github.com/picosonic/bbc-fdc

https://github.com/dhansel/ArduinoFDC

Micky

Wonder in how many areas RISC OS has a “Plug & Pray” issue – must be at least a dozen, seeing the development of USB, WiFi, FileCore and other crucial areas has frozen in the early 2000’s. Hence, in no way can RISC OS claim to be a modern operating system, less so for every passing year of development in computer-land.

Of course those smitten with The Rose-Colored Glasses-syndrome will have a field day here. And they are Legion..

Do you actually have anything useful or constructive to say?

Do you think we aren’t aware that things are rather outdated? Do you think we wouldn’t like to see stuff updated? Money doesn’t grow on trees, so opportunities are limited, and we make the best of what we have.

So, what, do you just lurk around here waiting for a problem to appear so you can drop in your £0.02 about how we’re delusional and the solution is so obvious blah blah? Does it amuse you? Because from here it seems like a rather sad and pathetic existence.

Seeing my favorite operating system go to the dogs isn’t amusing at all. And I wonder if it would be in Arm Ltd’s interest to have RISC OS developed up to modern standards – the funding to do so would be pocket change to them, and if you subscribe to the concept of Karma, the absolutely right thing to do.

If developing a commercial fork for use in smartphones the funding could be covered in a pretty short time, and RISC OS for computers – still open source – would be able to utilize all the fancy functionality in modern Arm CPUs/GPUs. Those currently developing RISC OS could cooperate and concentrate on writing a commercial (~£25) emulator for legacy software right after the conversion to C and the development of a new and user friendly SDK, also commercial. This would in sum constitute a true RISC OS revival.

Sveinung, your optimism may not be all that realistic (I’m not in a position to judge that), but certainly admirable.

and I wonder if it would be in Arm Ltd’s interest to have RISC OS developed up to modern standards

Why would it? They already have a number of genuinely modern OSs running on ARM platforms, all of which are developed and paid for by third parties. Why would they want to spend their own money attempting to bring a dinosaur from the 1980s “up to modern standards”, when – were they even to want their own OS – they would be better starting from scratch somewhere else in a high-level language and using modern, accessible UI paradigms?

Here are another interesting project.

http://boginjr.com/it/hw/megafdc/

Micky

We already have inexpensive and proven Greaseweazle and Kryoflux solutions to read ancient discs, especially those that are no longer perfect after 35 years or so. Nobody in their right mind would think about creating a new r/w floppy disc interface for all the four people who demand it. I thought the general idea was to bring the platform forwards not backwards?

And I wonder if it would be in Arm Ltd’s interest to have RISC OS developed up to modern standards

No.

Modern ARM processors already run modern OSs which support all the nice features.

the funding to do so would be pocket change to them

Probably, but you know, I figure ARM pretty much only included the 26 bit modes in the ARM6/7 as a bone thrown in the direction of Acorn. When Acorn went to the wall, the 26 bit was discarded as the relic that it was.

and if you subscribe to the concept of Karma, the absolutely right thing to do.

Why?
When your competitors are RISC-V and various reincarnated MIPS clones (those Realtek chips), and you have your sights set on world domination while trying to figure out who the hell actually owns you these days…
…whatever is happening to RISC OS is the full stop at the end of a footnote on page 327 of a two hundred and fifty page book.

Computing history is littered with the carcasses of former operating systems. EPOC, AmigaOS/Workbench, TOS, Symbian, MacOS (the original), RISC OS, DOS, Windows 3, OS/2…
…some are open source and are still being kept alive by enthusiasts, while others have faded into obscurity.

There’s going to be no great revival for any of them. Once in a while you might come across an article about how “ancient OS has been ported to a toaster” just for the hell of it. But while the toaster might be able to now run Plan 9, chances are the work was done on a Linux box, and it’s just an exercise in doing it because it can be done.

Be thankful that the sources to RISC OS are available, and some people are working on it still, but don’t think that “if we all band together we can change the world”. That didn’t work in 1997 when there were orders of magnitude more people involved.

If developing a commercial fork for use in smartphones the funding could be covered in a pretty short time,

Delusional.

Let me remind you, yet again, that Microsoft poured billions into smartphones and walked away in the end. Plus, a shocking lack of understanding of the security requirements of modern phones. There is probably no better example of an arena within which RISC OS is not suited, unless you’re thinking “what a great OS for a mission-critical flight control system”. ;)

If you disagree, well, the sources are there. Grab a copy and feel free to go off in your own direction. Good luck getting ahold of enough datasheets to do much with a smartphone – even AOSP/LineageOS have to make use of binary blobs to interface the OS to the hardware, because loads of stuff in these devices is neither open nor documented. IPR, trade secrets, blah blah.

Mr. Murray, you appear to have at least two blind spots; user friendliness and productivity. The RISC OS desktop/WIMP user interface is second to none, it’s the development in areas like Filecore, USB and WiFi which has suffered the passage of two decades with little or no development. In these areas it’s an operating system frozen, like a fly in amber.

And RISC OS has what may be considered a micro-kernel, in contrast to the pile of Prism-program trash called “M$ Windows” so it’s way more suited to run on mobile phones, technically speaking – it’d boot in the blink of an eye.. And it’d far from take “billions of dollars” to make this happen – once 64-bit RISC OS could run on anything Arm has made, and written in C, on other CPU’s too if that would prove useful in some setting. The drawback here is the massively shrunken market from Acorn Ltd’s heyday and the resultant small and under-funded companies developing RISC OS today – two very different situations.

To recap, it’s the user interface which is worth saving, and to remain a separate operating system it’s primarily the RISC OS core and relocatable modules which is in dire need of serious development. Add a new SDK and it’ll still take a thousand times less money than when M$ tried to fit a supertanker in a bathtub. With “no money, no people”. It’s been done before.

The drawback here is the massively shrunken market from Acorn Ltd’s heyday

I’d be inclined to say there is no market, just enthusiasts.

it’s the user interface which is worth saving

And yet that has or can be implemented on top of other things.

In the current market, for any device, RISC OS will always be paddling upstream in a canoe, whilst the speedboats wizz past us.

Just as with Windows Phone, you can be as innovative and unique as you like, but without huge support, it won’t go anywhere. I say this having had 2 Windows Phones and finally had to call it quits when apps ceased being available.

Mr. Murray, you appear to have at least two blind spots;

Behind me and around me? Come on, I’m not that podgy.

user friendliness and productivity. The RISC OS desktop/WIMP user interface is second to none

Yes, the Desktop is great. But the desktop could be implemented on top of other things.
And I think you’ll find really quickly that it’s not suitable for a mobile device.
Ever tried using RISC OS with only a touchscreen?

And RISC OS has what may be considered a micro-kernel

Not really. It’s small, yes, but it’s chock full of crap that doesn’t belong in a kernel. OS variables? Conversion SWIs? The VDU driver? The plethora of OS_Byte, CMOS RAM, keyboard gunk… shall I go on?

Strip out all of that crap (like ROLtd mostly did) and then maybe you can call it a microkernel.

in contrast to the pile of Prism-program trash called “M$ Windows"

Advocacy like that is so last millennium.

so it’s way more suited to run on mobile phones

Not even slightly.
It’s so far off the mark it’s like aiming for a target in Stavanger and hitting Milton Keynes instead.

it’d boot in the blink of an eye

And get pwned just as quickly. You want to run your banking app on ROMobile, go right ahead.

And it’d far from take “billions of dollars” to make this happen

Yup, it probably wouldn’t. But the point you’re missing is that Microsoft threw billions at their own system and it wasn’t enough.

If I was making a new mobile phone, I’d be putting some version of Android on it. Tried and tested and, so long as I sucked up to Google, a massive existing infrastructure.

the resultant small and under-funded companies developing RISC OS today

At least they’re around. But expect people to be interested in trying to make what we have work, not in delusions.

it’s the user interface which is worth saving

Perhaps. But it, too, is in need of work. A lot of work. Unicode, for example.

it’ll still take a thousand times less money than when M$ tried to fit a supertanker in a bathtub.

That’s still around 7-9 million (I forget the exact figure and can’t be arsed to look it up).
Which is maybe a thousand times more than we have available.

With “no money, no people”. It’s been done before.

Yes, it has.
But I think you’ll find if you compare the “no resource success stories” against the “no resource abandoned failures” you’ll start to spot a pattern.

I would really prefer our low budget successes to be realistic things.

Thinking that with no resources it’s possible to rewrite the entire OS in a high level language and have it be a successful mobile phone OS is imaginative trolling. Imaginative, certainly. But still trolling.

At the top, you mentioned productivity. Well, a lot of what I use is writing stuff for my blog which would be possible on any word processor really. I just use Zap because I’m used to it (and most of the stuff on Android is advert-laden crap).
Otherwise, it’s creating crappy videos ¹ (in 1080p) to upload to YouTube, or watching stuff on YouTube, Netflix, Prime. Sorry, but my OS of choice for that is something else because no mustard gets cut here.

¹ I’m expecting at the very least fades, flexible cutting and video/audio replacement, overlays (text, image, video), and chromakeying into another video. The free app I use can do all of that, and more.

Any chance that someone involved in the USB floppy issues ‘deep dive’ could summarise the current state of play w.r.t. using such with DOS formatted floppies on RISC OS? If it works, is it limited to HD or can it handle DD as well for those wanting to transfer data back to old Arcs?

I think, if I remember correctly, that it needs a (small?) modification to the stack, possibly to recognise the devices (I know it was shown to work, but don’t remember the specifics).

As for DD/HD, both ought to work as the drives claim to be able to read both, so I’d imagine the fake mass storage device that is presented would be sized according to the media that is inserted.

The RISC OS desktop/WIMP user interface is second to none

First, it’s the least developped part of the system. Untouched since decades.
Secund, it’s of no use for a mobile device.

RISC OS has what may be considered a micro-kernel, in contrast to the pile of Prism-program trash called “M$ Windows”

(Very) Bad example. RISC OS 5 does not rely on a microkernel. Windows 7/8/8.1/10/11 does (OK: almost).

It’s so far off the mark it’s like aiming for a target in Stavanger and hitting Milton Keynes instead.

Yes… and no. The modem part is easier than most think. Just AT commands sent to a serial device. Phone calls are the same. Most 3G, 4G, 5G modems are ‘self working devices’.

Thinking that with no resources it’s possible to rewrite the entire OS in a high level language and have it be a successful mobile phone OS is imaginative trolling. Imaginative, certainly. But still trolling.

Correct. And more: not so imaginative trolling. And I can prove it.

Sorry Sveinung, but you said wrong things several times. I did raise the problem and always asked the same question: did you even try RISC OS 5 on a Pi4? Each time, the discussion stopped here… and started again on another topic. Just trolling, sorry.

So I’ll now go in Stuart (?) mode:
JUST BUY A PI4 AND USE IT.

BTW: thanks Stuart for ArchiEmu :)

it’s creating crappy videos 1 (in 1080p) to upload to YouTube

1/ Where, where, where?
2/ More, more, more.

1/ Where, where, where?

https://www.youtube.com/heyrickblog

God ¹ help you if you waste your time on that crap. ;)

¹ Other mythological beings may be available for a small fee.

The RISC OS desktop being simple does not imply it being unsophisticated nor unproductive. As for phones, of course the UI would have to be simplified (totally redesigned) for touchscreen use (and naturally, a touchscreen driver written). Cleaning up the core by removing everythin not necessary for use in a phone would probably make it a nano-kernel. And not be called RISC OS – maybe ARMOS unless a howling horde of Arm’s predatory lawyers would pounce on that. The phones market is so huge that even a 0.1% market share could sustain the company making it. And it wouldn’t be beholden to no-such-agency like both Apple and Google are, unless your GCHQ are as spy-horny as their trans-Atlantic counterpart.

Right, you’ve had your fun, it’s now time to go find another bridge to live under.