On Jul 17, 2014, at 3:51 PM, Kevin Keith <krfkeith at gmail.com> wrote:
Ha!
In all seriousness, what I was referring to was what I had read about
it awhile ago on Wikipedia,
"[Futurebus] intended to replace all local bus connections in a
computer, including the CPU, memory, plug-in cards and even, to some
extent, LAN links between machines."
How accurate is that statement? There doesn't appear to be any citation.
Wow.
I never heard any such thing. If I had heard the part about LAN links I would have
laughed the speaker out of the room. In my (not very intense) involvement with
Futurebus+, it felt like an I/O bus, like BI and XMI etc. but based on an international
standard rather than an internal design.
We did some interesting things for this, because (from what we were told) it was important
to avoid memory reads if possible. So we constructed some novel host <-> device
interface data structures that used writes almost exclusively, much more so than, say, a
classic descriptor ring (as found in Lance).
paul