Newsgroups: comp.sys.amiga.misc
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!mintaka!geech.gnu.ai.mit.edu!rjc
From: rjc@geech.gnu.ai.mit.edu (Ray Cromwell)
Subject: Re: A new Amiga needed...and soon!
Message-ID: <1991Jun30.172242.5422@mintaka.lcs.mit.edu>
Sender: news@mintaka.lcs.mit.edu
Organization: The Internet
References: <1991Jun29.194017.10138@apollo.hp.com> <1991Jun29.204141.22773@mintaka.lcs.mit.edu> <16029@exodus.Eng.Sun.COM>
Date: Sun, 30 Jun 91 17:22:42 GMT
Lines: 67

In article <16029@exodus.Eng.Sun.COM> cmcmanis@stpeter.Eng.Sun.COM (Chuck McManis) writes:
>You forgot the smiley face. You should check out OSF's latest fooforah
>something called "ANDF" which sounds like your P-CODE idea. Only it
>is a bit different, they actually want to "compile" ANDF code into the
>native machine code and then run that compiled executable. Sounds like
>a support nightmare to me (how do you know if a bug in the application
>you just sold your customer is in your code or the ANDF compiler? And
>if it is in the ANDF compiler how do you convince the customer? Virii
>are also a sticky issue if your compiler gets infected) 
>
>Several different p-code like systems have of course been tried, the
>most common being UCSD Pascal and Xerox's SmallTalk system (which was
>a virtual machine). Unfortunately such systems always seem to be at
>a disadvantage to something like IBM PC clones which just forego the 
>various architectures and all use the same one.

   I was sorta of thinking about a loader that takes the second pass
output from a compiler. (sort of like GCC's RTL) It would contain
instructions like (move memory into register, move register to stack, etc)
Of course there would be hinting in the file to have loaders optimize
at load time. Of course, the loader program could always be saved in
its native machine code for faster loading (it would still have the
p-code hunks in it though)

>>  I can't wait for the day when we all have fiber-optic network lines
>>to the home just like phones. ...
>
>I'd just like PacBell to get off their butt and get ISDN service to
>my house. Even for a zillion dollars they don't seem to inclined to
>offer this, much less direct fiber. 

  I'd like direct fiber to the home so I can run Ethernet speed
modem connections (or >Ethernet speed), video transmissions (HDTV
channels?) , multiple phone lines (ISDN). I've heard that the cost
of fiber is approaching that of copper now. I'd gladly pay to have
a fiber line to my house.
(atdt 5551212
CONNECT 10485760)

>>  I believe on (one?) day there will be a processor independent, platform
>>independent, extensive OS. I could care less if it run MSDOS programs or not,
>>I'd rather discard that excess baggage for state of the art.
>
>Well SmallTalk or a decent Lisp environment might cover it then. Did you
>want an "extensive" or "extensible" OS ? Either way it will take a lot
>of memory :-) but heck, they're showing of 64MBit DRAMS at ISSCC so 
>those 4MB simms ought to be getting cheap enough for the home market.

Extensible. One that never needs a complete replacement. After all,
with a massive standard like this, we don't want it to become obsolete.

All in good time. With 14gb 128mb/sec optical cd's on the way (via
Optex,inc and their modulated light encoding method) and the really big
drams, memory and disk storage won't be the main problem.

>--
>--Chuck McManis						    Sun Microsystems
>uucp: {anywhere}!sun!cmcmanis   BIX: <none>   Internet: cmcmanis@Eng.Sun.COM
>These opinions are my own and no one elses, but you knew that didn't you.
>"I tell you this parrot is bleeding deceased!"


--
/ INET:rjc@gnu.ai.mit.edu     *   // The opinions expressed here do not      \
| INET:r_cromwe@upr2.clu.net  | \X/  in any way reflect the views of my self.|
\ UUCP:uunet!tnc!m0023        *                                              /

