>>>>> "Victor" == Victor Yodaiken <[EMAIL PROTECTED]> writes:
Victor> On Sun, Jul 11, 1999 at 02:14:26PM -0400, Richard Skaggs
Victor> wrote:
>> Hello, Victor, Dave and all, I read your comments below, and sorry
>> to ask such a lo tech question, but am planning to do a machine
>> control RTL project (with EMC if you should know it). With respect
>> to hardware, I will be building a 366 or 400 Mhz MB Intel
>> architecture machine. What's an appropriate bus if not ISA for
>> RTL? Or,
Victor> PCI is not wonderful, but it is far better than isa. Isa is
Victor> simply slow and very poor at handling contention.
One interesting thing about buses is that few if any have well-defined
real time properties. Usually if you ask a bus designer about latency
bounds you merely get a blank stare. They can talk marketing best
case numbers all day, and perhaps even "typical" or "average" case
numbers, but the notion that anyone might be interested in a worst
case bounds seems to be foreign. If you push hard, you may get a
number. Often, that number will come with units of milliseconds, not
microseconds.
I find it somewhat ironic to have people get into religious debates
about the non-real-time nature of Ethernet when the underlying system
buses aren't real time in the first place.
paul
--- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
----
For more information on Real-Time Linux see:
http://www.rtlinux.org/~rtlinux/
