Date : Thu, 16 Nov 2006 20:16:10 +0000 (GMT)
From : debounce@... (Greg Cook)
Subject: OT: DFS on 3.5inch - Is this supposed to work?
On Thu, 16 Nov 2006 09:32:38 +0100, "W.Scholten" <s-whs@...>
wrote:
> Well, I must say I concur with Gareth Babb, maybe it's time to
> unsubscribe:
>
> There's very little of interest/useful responses on the ML for me and
>
> usually there's no interesting discussion.
[...]
> (where
> some guy complained that the ML was [not] interesting enough for him
> anymore;
> as if the ML was there to entertain him!
As if the ML were there to entertain you!
> And it's an example of
> 'those
> who contribute the least, complain the most'),
This is not a complaint?
[...]
> I can give more old examples, but I'll finish with this one:
>
> [ Adam Colley ]
> > Most PCs will read and write it just fine using the correct
> software
> > (IE: Omniflop)
> >
> > I have four PCs here, three desktop machines and a laptop, all four
> > are able to read and write DFS disks just fine with Omniflop
> >
> > I'm not sure where this myth about almost all PC FDC's having
> broken
> > support for DFS etc. came from but it's just that, a myth.
> >
>
>
> Complete and utter [...]. Read the mailing list archives. Read the
>
> docs with FDC. I would not say "Most PCs haven broken SD support"
Neither did Adam.
> but
>
> the facts from people trying e.g. FDC and anadisk (so direct hardware
>
> access, no BIOS or whatever intermediate layer) gathered over many
> years
> on loads of different machines from various eras but esp. up to the
> pentium 2 era (I've seen very few reports after that), are, that SD
> did
> not work.
In other words, there are thousands of MFM only PCs out there, a
significant proportion. This is not disputed.
On the other hand, Adam was able to obtain four FM capable PCs without
especially trying (four out of four). This is one person's testimony
on a backdrop of anecdotal evidence. The dearth of reports, if most
are negative, may mean increasing success as well as loss of interest.
To reconcile these two examples, let's say the actual proportion lies
between 10 and 90 per cent. To get a more accurate answer, write a
bigger cheque.
> I've posted about this a couple of years ago too when
> someone
> said that recent machines seemed to support single density without
> problem. This was not new, that was a *probable* trend (too little
> evidence to be sure) already noted in the docs of the 1998 release of
> FDC.
>
> I tried out a number of PCs with FDC/Anadisk from a 486, to a pentium
>
> pro, k6-233, k6-350, pentium-166 and found that 80 track didn't work,
>
> except with a switchable speed drive on the ppro/p-166. 40 track
> worked
> on the ppro (-> rotation speed) and perhaps also the p166 but none of
>
> the others. Robert Schmidt tried out far more IIRC and he and I got
> lots
> of email over the years of reports from people for whom it didn't
> work,
> and sometimes it did from people for whom it did work.
People are statistically more likely to write when it doesn't work.
> So, that 'most PCs don't support SD' might be false now but I've not
> seen enough evidence to be sure (and your couple of machines are
> statistically insignificant), it looks as if this wasn't so in the
> past.
And what would be a statistically significant number of machines? What
would be a statistically valid number of experimenters? As email
response is an unreliable measure, can you be sure enough different
people have tested enough different machines and reported fairly?
More to the point, is it really worth a flame war over what proportion
of machines supported FM out-of-the-box in any given year? (How much)
is it worth paying for a study?
> This doesn't mean "it's a myth" is correct (look up the meaning of
> the
> word myth). If you think most *current* PCs work with SD, provide
> evidence that your assertion is true, FROM TESTS WITH MORE THAN A
> HANDFUL OF MACHINES/CHIPSETS, and then you can say "is not true (any
> more)".
We can say that already, thank you very much :-) Look up the meaning
of 'can'.
Of course, myths may be true myths. But why is there a myth about the
success or failure of DFS on PCs anyway? Because nobody's put their
foot down and repressed unscientific statements? Or because it's never
been the subject of study? And why would it be studied? The percentage
isn't important for most of us because most of us just need *one* PC
that will do DFS.
> For the integrated chipsets of todays, probably only a
> works/doesn't work per chipset list would be needed. On older systems
>
> things could be broken due to e.g. some pins not being connected
> (multiple chips connected), so a per computer type was needed (see
> also
> the docs with FDC/FDCdemo). Not sure if unconnected pins can cause
> trouble with recent chipsets.
Instead of moaning about ignorance, be the light in the darkness and
share your knowledge! If you want to be scientific about it,
publishing the results (and method) will facilitate the peer review
process (if there's going to be one 9_9)
> > Well I've not found a PC that doesn't work, I'm sure a few exist
> but I
> > suspect they're greatly outnumbered by the working ones, this all
> > sounds like a PEBKAC problem to me...
>
> I had to look PEBKAC up. Aha, the user gets the blame (and what a
> surprise, YAUA(*)). So you Adam Colley, who doesn't know anything
> about the amount of machines this was tested on in the past/recently,
> by whom and with what failure rate,
You hold that against him, yet you're not going to enlighten him.
Nice.
> decide that the user trying it
> is/was the problem?
His criticism is valid or invalid on its own merits, regardless of
whether you look up or down on him. I'm sure this Cerfontaine thinks
you've nothing to say, since you're not a company director yourself.
Even if user error is seldom involved in FM failure, you sir have taken
entirely the wrong attitude. It really didn't deserve a swearword, or
even the suggestion of one.
> Oh yes, read the
> archives for a message from me about disk reading where I mentioned
> 300/360 rpm switchable drives. PEBKAC indeed.
Are you saying that's proof that DFS-on-PC problems aren't your fault?
If anyone can get it to work, you can? That you "know all there is to
know" on the subject? ...Is this flame about someone else having all
the luck with FM? ;-D
Greg
___________________________________________________________
Try the all-new Yahoo! Mail. "The New Version is radically easier to use"
? The Wall Street Journal
http://uk.docs.yahoo.com/nowyoucan.html