[vcf-midatlantic] How To Geek article on floppy data recovery

Fri May 8 12:04:54 EDT 2020

However, consider that, over time, it may matter less and less as a larger
volume of old media gets archived. Maybe it’s ok that these devices are
somewhat ephemeral. Just food for thought.

On Fri, May 8, 2020 at 11:45 AM Herb Johnson via vcf-midatlantic <
vcf-midatlantic at lists.vcfed.org> wrote:

> > There are a number of newer tools, for imaging of disks, for
> preservation.
> >
> > I've been closely watching this one [called Fluxengine]
> > http://cowlark.com/fluxengine/index.html
> >
> > The author is working on support for a lot of common formats, and
> > additionally, a number of the 80s-90s dedicated word processor formats.
> > It's a simple and rather nifty solution.
> >
> > Also, of course, it's 100% opens source hardware and software, which
> always
> > makes me happy! - Alex
>
> This reply is long, because I'm arguing in opposition, and that means I
> have to make a case about, and explain about. But I'll save some people,
> some time. I"m going to fuss about these microcontrollers becoming
> obsolete. If you don't care about that, save time and stop reading here.
>
> -----------
>
> Well, I read some of the support information on the Fluxengine. Also I
> looked recently, at the Greaseweasel. Both these methodologies use some
> kind of $10 - $20 commodity microcontroller; and that's their problem in
> my opinion. Those interested can look up both and size them up, so I
> won't talk about details and features.
>
> The Greaseweasel uses a STM32 microcontroller chip on a board called "a
> Blue Pill". Problem with that (in my opinion) is, that chip isn't just
> one chip produced identically. There's knockoffs and fakes, that is
> less-featured chip versions remarked or not remarked. And since that
> "product" is more than five years old apparently, it's somewhat out of
> favor (that's my conclusion, it seems hard to find except on eBay).
>
> Bottom line, it's not now produced by bigger named hobby or industrial
> companies; they all seem to come from  ebay/Amazon small producers. At
> literally a few dollars each, the only way to make more money from
> these, is to use cheaper components or reduced feature components. For
> instance: the 3.3V regulators fail because they are undersized for
> power. When I looked around Amazon for these, the reviews per seller
> were all "these are fakes because".
>
> This is not just Herb Johnson's problem. The Greaseweasel support page
> on github, has a section on how to recognize fakes with performance and
> visual tests. It does not have, a section on where to buy these.
> Apparently one has to join a Facebook group to get answers and further
> support. I won't do that, but someone offered to look there for me, to
> see if they discuss suppliers. No word yet on that.
>
> "Why is this a problem?" Specifically: some of the knockoff boards fail,
> apparently, because their 3.3V regulators burn up. Others, lack some
> useful features or even critical features (enough memory for instance).
> And in the longer term, the ability to obtain these items at all, may
> end. In between, the means to (say) assemble the code or download the
> code into the device, may become obsolete or unavailable.
>
> In general, the problem with end-of-production items is this: Because
> you and I are dealing with *vintage computing*. We deal with technology
> that's *decades* old, not last month. When products fall out of
> production, how do you get more? When things break, how do you fix them?
> When you can't get the programming debugging tools from the manufacturer
> (or producers), you have to find old archives of them.
>
> In the real world, this stuff is normal. Things don't persist. You just
> move on to what's current and throw out the old. We in vintage
> computing, can't do that by definition. OK? If you disagree, stop
> reading here, sorry I wasted your time fussing about stuff getting old.
>
> -----------------------
>
> I described at length the problem with Greaseweasel hardware, because
> I'll make the same argument as a *prediction* for  Fluxengine. It uses a
> "Cypress PSoC5LP CY8CKIT-059 development board" for $10. That's
> apparently some ARM processor with FPGA (programmable logic). At this
> point, Cypress (a leading chip producer) is selling them direct. Let's
> check... yep, $10 from cypress.com or $20 from major distributors, plus
> shipping. One expects delays due to Coronavirus disrupting manufacture
> and shipping except from USA stocks.
>
> I have another objection, to programmable logic. All those PLD's are
> unique by model and brand. They simply don't stay in production for many
> years. Worse, the software tools needed to program those specific
> devices, become unsupported in several years. There's no standards for
> such tools and software, other than some data formats, so you are stuck
> with that brand's tool-set.
>
> This is, of course, an intended consequence. Figure that out.
>
> I checked the datasheets for the product: they date from early 2015.
> That's about the same age as the "Blue Pill". But this is still produced
> (or at least offered as current) by Cypress. Also: since this has a very
> particular programmable-logic set-of-features, it may be hard to
> knock-off the chip by an unlicensed company. That's my guess.
>
> So: looking at Amazon, they only sell a book. Looking at eBay, There's a
> couple units from the UK for $28-$33 with shipping. I'd not buy from
> Europe right now...  a few are US sold but essentially it's left-over
> stocks.
>
> So: again in my opinion: while this product is still in manufacturer's
> production, it suffers from being too proprietary - Cypress will lose
> interest someday soon. Also, there's that unique-programming problem,
> figuring out how-to-program those PLA's. If you aren't gonna reprogram
> one of these puppies, I guess that doesn't matter, until there's a
> Fluxengine update you can't update.
>
> And in closing ....
>
> The reason I'm familiar with this kind of "these go obsolete" arguments,
> is because I've seen this show before. Over the last few decades,
> various modern means to read ancient floppy diskettes have been
> produced. They have become in-favor and discussed and distributed. Then
> the developers lose interest; then the users lose interest; and they go
> out of production and use. Over geologic time (several years), these
> things simply come and go.
>
> So the vintage computers remain, but these modern devices do not. That's
> about as simple as I can put the argument.
>
> Regards, Herb Johnson
>
>
> --
> Herbert R. Johnson, New Jersey in the USA
> http://www.retrotechnology.com OR .net
> preserve, recover, restore 1970's computing
> email: hjohnson AT retrotechnology DOT com
> or try later herbjohnson AT retrotechnology DOT info
>