Cyclone 20 - idea for homemade USB-based floppy controller (Catweasel-ish)

[mr.vince]

Here's some technical info on the production boards:

Attachment 25078

[Supamax]

I see that the board supports two floppy drives... why?
I mean: how can it make use of 2 drives at the same time? Perhaps for original--->empty floppy copying without disk swapping?

[Loedown]

Originally Posted by Supamax

I see that the board supports two floppy drives... why?
I mean: how can it make use of 2 drives at the same time? Perhaps for original--->empty floppy copying without disk swapping?

http://www.pctechguide.com/images/tu...loppyCable.jpg

Old floppy cables had provision for two drives, either two 3.5" or two 5.25" or a combination of both, Kryoflux probably has the ability to connect to both at the same time.

[andreas]

Originally Posted by RichAplin

Yes can't complain. Got a bit distracted there for a year, you know how it is.

Delighted to see things are still cooking with Kryoflux over here, props to the SPS folks for taking a rough-proof-of-concept and rebuilding it into a real solid working thing. ;-)

Nice to see you back here...
Just one thing...before you disappear to become a *void() again

Since alexfreed has visibly put lots of work, time and brainpower into the "old" code, what is your point about at least making a recent version available on SVN for the fiddlers?
I think there are people here that cannot be convinced to delve into ASM (let alone C#) and favorising good ol' C, also for reasons of max portability.
For these folks, it would be paradise to have the current C code at their disposal.

vince said, literally, he cannot make any decisions without your (Rich's) permission, so what's your point about this?

- in behalf of many -

[RichAplin]

Hiya,

"Current" (heh) state of Cyclone20 source is basically a pile of testing stuff thrown together to go along with my rants about 20 pages back.

1. There's the C + ASM firmware.
2. The C# windows side (draws graphs, half-assed UI)
3. A tweaked command-line version of Jeff's IPF->RAW thingy that generates track dumps in a raw format that the C# code reads and sends over USB.

It's all about 50% unused cruft, 40% misleading/outdated comments, scribbled notes, global/static variables, incorrect function names, #defines and all the things your mother warned you about.

Most problematic of all, it has a software USB protocol issue on the firmware side which causes sporadic hangups, and this is what I was kicking at last time I got called away with job stuff.

So basically, it's fugly and it's incomplete/alpha/preproduction/broken functionally.
More on releasing this later.

-- Non-technical issues
I think the only issue anyone ever suggested to me against releasing all the code was that putting something out that did perfect writes from IPF files could cause issues with ongoing SPS work, ebay forgeries, etc {see pages back}.

-- My stance on the whole thing
As you know I have the attention span of a goldfish, so really I'm just happy that I was right all along about the "cheap Atmel + a dozen bits of wire" design and the general physics/electrics/gymnastics of it all.

As far as I can see from the SPS postings it's all done+dusted as far as "yes the board can do all the stuff we said it could" ; after some hacking Istvan et al rewrote the Atmel firmware from scratch (rather than picking over the wonky chicken bones I left) and now have that end of it working nice and reliably. All this will be (is being!) revealed by SPS with regular updates and code.

--What about {other features, not in Kryoflux}
Well obviously SPS are building something primarily for their own purposes that they're happy to share with the world.

I understand they're sharing the PC-side as source, and are currently sharing their Atmel firmware as a binary. They have some reasons for this (stated by them elsewhere) which is fine by me, and I look forward to whatever they want to release as src in the future.

Here's the long+short of it; the current C20 source is literally no use to anyone who's not pretty hot on their embedded SOCs, C, ARM and C# coding, not to mention the last 20+ pages of technical blah on this thread.
Alas I don't have time to finish it (at least for forseeable future) or debug the USB issue, or explain/support/tidy the code I threw together.

Bear in mind all the C20 source code I ever wrote is about a week or two of prototyping work thrown shoddily together; Istvan looked at it for a bit, went "this is fugly and kinda broken" and rewrote the whole thing from scratch. He had the right idea. ;-)

Let's be clear, it's mostly about the Atmel firmware, not the PC-side. The PC side (cmd line or GUI) you can write simply in half a dozen possible languages (probably even Visual Basic) ; anything that let you do the basic winusb bindings; performance not really an issue either on a modern PC.

Basically, if SPS want to release their Atmel firmware as src, I'd suggest everyone use that as a base because they've done a lot of work on it.

If SPS release the firmware as binary and then publish the USB command set that's extremely useful too (possibly moreso to many ppl on this list), especially if it writes to disk.


However if someone wants to do (for example) floppy disk emulation with a Ukranian board (and for some reason don't want Jeff's HxC), they might as well start from scratch and put something together from the free "read FAT16 SD card" sample code and some basic timer stuff in C. You can do it readily on a PIC (see Amiga-on-FPGA) so it's pretty easy on an ARM.

blah ramble etc.
Cheers,
Rich

[andreas]

Thank you!

I thought I'd do the "(re)kindling of the info fires", so that the dry mouths be sufficiently watered again (I hope )

Here's the long+short of it; the current C20 source is literally no use to anyone who's not pretty hot on their embedded SOCs, C, ARM and C# coding, not to mention the last 20+ pages of technical blah on this thread.
Alas I don't have time to finish it (at least for forseeable future) or debug the USB issue, or explain/support/tidy the code I threw together.

YET some people are drooling to get that latest code version, obviously (namely e. g. alexfreed). So if you can make 'em happy that way, just go on & do it
Peeps just wanna 'ave sumthin ta play aroun' with...and as the current mood and feelings seem to be, they fear that SPS is going to enCAPSulate [1] the interesting stuff inside a block of "ROM concrete" which said "fiddlers" are so keen on looking at after all...

[1] hah, that's a pun, ain't it!

[Supamax]

Originally Posted by Loedown

http://www.pctechguide.com/images/tu...loppyCable.jpg

Old floppy cables had provision for two drives, either two 3.5" or two 5.25" or a combination of both, Kryoflux probably has the ability to connect to both at the same time.

Yes yes, I know that .

What I would like to know is: are the two drives controlled separately (for example a 3.5" and a 5.25" for different source floppies), or can they be used at the same time, for example for direct floppy copying purposes?

In other words, how will exactly Kryoflux make use of the two drives?

[mr.vince]

@Supamax: You should really ask questions about the KF project in the appropriate thread. I promise you'll get a quick reply.

[Supamax]

Originally Posted by mr.vince

@Supamax: You should really ask questions about the KF project in the appropriate thread. I promise you'll get a quick reply.

Done

[hit]

Originally Posted by RichAplin

Hiya,
...
blah ramble etc.
Cheers,
Rich

- wow, first the SPS thiefs and now Mr Aplin himself, fooling his fellow crowd. *gosh* and i thought this project was doomed.
- thnx for nothing richard - if you are bored again and need some worship - contact SPS directly :P

[Supamax]

Originally Posted by hit

- wow, first the SPS thiefs and now Mr Aplin himself, fooling his fellow crowd. *gosh* and i thought this project was doomed.
- thnx for nothing richard - if you are bored again and need some worship - contact SPS directly :P

???
I probably misunderstood your post, hit... else, are you crazy?

[RichAplin]

Heh. Ok look at it like this.

Imagine that the task of "building Cyclone20" is the same skill as learning to drive a car.

There's lots of people, some of them can drive a car, some of them not.

Everyone wants to travel to different places, and fairly inexpensive cars are readily available.

What to do?

Either wait patiently for someone driving in your direction, or learn to drive for yourself.

If you stand by the roadside having a tantrum, nobody will stop for you or help you learn to drive. ;-)


Cheers,
Rich

[mihcael]

Originally Posted by RichAplin

...
Either wait patiently for someone driving in your direction, or learn to drive for yourself.
...

Or ride a bike! (buy a catweasel)?

[Loedown]

I own a Catweasel, with the stagnant progress of updates and almost non existent support, if SPS' Kryoflux works properly, Catweasel will disappear.

[BuZz]

I think the argument about the ebay forgeries etc, reads just like the sps list of reasons why they don't release the source to their library and really is a non argument (to me at least) - they have a lot of reasons, but they all read like excuses:

* we are all too stupid to understand the source anyway (thanks!)
* people would flood the web with fakes - never heard of providing lists of verified checksums? i dont see mame rom fakes everywhere - noone actually cares to make fakes of something like this!
* we might release the code eventually (when ?)
* you can use the library anyway you don't need the code (sure, i should have to send you some new device and wait for you to build the library for it should i - if that would ever happen ?)

I do not think anyone is going to make a business printing fake boxes, and printing "disks", to try and flog on ebay for £5. It just reads like an excuse to me. Projects aimed at any kind of preservation need to be open, or they are not that useful imho.

the catweasel is reasonably useless too, as no low level documentation or sourcecode is available for the amiga driver, and the other drivers are quite limited in functionality.

The cyclone project to me, did read as - here's a great idea/project and some code in progress.. then sps get involved. everything goes into a private members club with no public code etc. That's my impression at least.

I really wish I didn't feel this way about it. And I do have my fingers cross about their statement about the source that "Yes, absolutely. To not do would be rather flawed preservation. "

I'm not holding my breath mind.

[mr.vince]

Sorry to ask, but why would anyone of us need an excuse for anything? There are reasons for decisions, and sometimes someone tries to illustrate why a particular decision has been made. We make decisions, Rich makes decisions. Some people like them, others don't. You can't please everyone.

We're pretty sure that many people will be happy we started KryoFlux when it was clear that Rich would not return when see the final product and its capabilities.

If you still feel the closed source library for accessing IPFs is a problem for you, then I'd really like to recommend the HxC project (which can ingest and process IPF data) and its sources for further reading.

[gilgamesh]

I think this is getting in the wrong direction. Of course the SPS does not need any excuses, but don't you see the fundamental problem, which keeps people asking?

Preserving software means to grant access even when the media are outdated. But a binary as a tool for access is much more likely to become outdated itself than the more abstract sourcecode, or maybe even a printed format documentation.

[mr.vince]

Yes, I understand your concerns. However, I'd feel more tempted to say something about this matter (again), if this would happen in the correct sub-forums. We should not hijack this thread.

And again, for the time being, the sources for HxC are a good read, for sure.

[Interceptor]

Originally Posted by hit

- wow, first the SPS thiefs and now Mr Aplin himself, fooling his fellow crowd. *gosh* and i thought this project was doomed.
- thnx for nothing richard - if you are bored again and need some worship - contact SPS directly :P

lol, i really hope you gnash your teeth over this until they are ground down to a powder :P

[cosmicfrog]

richard got out of this exactly what he was after

Proof that it could be done

Now don`t go complaing if you did`t get anything out of it

Me I enjoyed the ride