18 Feb
2019
18 Feb
'19
4:26 a.m.
On Sat, 16 Feb 2019 at 22:53, Jeffrey S. Worley via cctalk
<cctalk at classiccmp.org> wrote:
Of all machines I've used, the beloved Atari 8-bit is most vocal. It
has the feature of 'i/o noise' by default. It can be disabled with a
Poke, but every kind of io has distinctive sounds and actually
represents the data being sent/received. If you disable it and crank
the volume on your TV, you can STILL hear it, but very muted. I think
this feature was created to conceal this fact...
It isn't just the Atari8 that has this 'feature' in its muted version,
all of the RF-TV-type machines from the 80's produce it.
Interesting, yes. I noticed that with my ZX Spectrum machines and yes,
it was interesting and occasionally useful for debugging. "No, it's
not crashed, it's deep in a loop, and hang on, I can hear that every
few seconds it's briefly going through the outer loop, so it's still
working... give it time..."
In theory, I
think you could snoop the actual data, Tempest-like, using some radio
gear.
I think it's been done.
In the 90's I got the hpfs386 driver out of a warp
server pack and hung
it on my warp 4 client. I LOVED hearing it hit the drive at boot. Boy
howdy what a performance increase that gave.
Yes, exactly! :-D
There were busmastering drivers for both Win95/98 and for NT4.
On Win9x they made no particular difference, because the "kernel" was
single-threaded and so couldn't overlap anything else with disk I/O.
You got a very small benefit from DMA being a smidgen faster than PIO
and that was it.
But on NT, even on a uniprocessor, when it offloaded disk I/O onto the
Intel Triton 84230 PIIX disk controller, the kernel could get on with
doing other stuff while waiting for disk accesses to complete. Boot up
got tens of seconds quicker, and more to the point, the machine
remained responsive when under heavy disk load. E.g. even if it was
thrashing you had a chance of killing the offending process, which you
didn't without the DMA drivers.
But no system builder installed them by default. Possibly because then
the disk wouldn't boot on different IDE hardware -- bear in mind this
was before Windows activation, so an install could just be duped onto
another machine. But if the other machine didn't have Triton chipset,
it wouldn't boot.
--
Liam Proven - Profile: https://about.me/liamproven
Email: lproven at cix.co.uk - Google Mail/Hangouts/Plus: lproven at gmail.com
Twitter/Facebook/Flickr: lproven - Skype/LinkedIn: liamproven
UK: +44 7939-087884 - ?R (+ WhatsApp/Telegram/Signal): +420 702 829 053