Message-ID: <1459@emory.UUCP>
Date: Tue, 8-Jan-85 23:36:43 EST
Article-I.D.: emory.1459
Posted: Tue Jan  8 23:36:43 1985
Date-Received: Fri, 11-Jan-85 22:31:46 EST
Distribution: net
Organization: Math & Computer Science, Emory University, Atlanta
Lines: 12
Xref: watmath net.unix-wizards:11428 net.micro.pc:3085

Are there any rumors of 4.2 ports to the PC/AT? Is there
a reason why it would not be practical?

-- 
Ken Mandelberg
Emory University
Dept of Math and CS
Atlanta, Ga 30322

{akgua,sb1,gatech,decvax}!emory!km   USENET
km@emory                      CSNET
km.emory@csnet-relay          ARPANET

Message-ID: <342@rlgvax.UUCP>
Date: Fri, 11-Jan-85 01:41:19 EST
Article-I.D.: rlgvax.342
Posted: Fri Jan 11 01:41:19 1985
Date-Received: Sun, 13-Jan-85 07:46:11 EST
References: <1459@emory.UUCP>
Distribution: net
Organization: CCI Office Systems Group, Reston, VA
Lines: 23
Xref: watmath net.unix-wizards:11480 net.micro.pc:3100

 >  Are there any rumors of 4.2 ports to the PC/AT? Is there
 >  a reason why it would not be practical?

How good is the 80286 at 32-bit integers?  If the answer is "very poor",
forget it - I'd estimate several programmer-months of work converting
4.2BSD to use "long" for everything that must be 32 bits (if it's even
practical at all).  If it can add and subtract them reasonably well,
that's probably enough - the 68010 can't do better, and there are several
4.2 ports to it.

How good is the 80286 at paged virtual memory?  Can the memory mapper handle
an address space made up of fixed-size pages, not all of which need be
resident?  If not, you'd have to hack a lot to make a segmented system
out of 4.2.

And now for the $64,000 question: how good is the 80286 at 32-bit pointers?
If the answer is "very poor", forget it.

I've not heard of any ports; I suspect the answer to at least one of the
above questions is "not good enough".

	Guy Harris
	{seismo,ihnp4,allegra}!rlgvax!guy