Why are Janet addresses backwards?
Seth Grimes
file paths. Here's my next question:
Why is the "@node.domain" portion of "Internet" addresses on Janet (Joint
Academic Network), the UK portion of the Internet, backwards? That is, for
internal Janet use, an address is given as bur...@uk.ac.mi5 instead of
the order an uninformed-in-this-instance person like me would expect. Yet I
can give my mailer the regular RFC-822 Internet address, bur...@mi5.ac.uk,
and the mail will get there.
Seth, gri...@oecd.fr
Bernie Cosell
} Why is the "@node.domain" portion of "Internet" addresses on Janet (Joint
} Academic Network), the UK portion of the Internet, backwards? That is, for
} internal Janet use, an address is given as bur...@uk.ac.mi5 instead of
} the order an uninformed-in-this-instance person like me would expect. Yet I
} can give my mailer the regular RFC-822 Internet address, bur...@mi5.ac.uk,
} and the mail will get there.
The perversity of the British, what more can i say?? :-) when the
network addressing standard was being sorted out, the US [via its
RFC mechanism] chose to do right-to-left domain parsing, and the UK
chose to do left-to-right. This has been a nuisance since the day
it first happened [which was many years ago], but its crockiness
has largely been contained by smart-enough mail gateways
surrounding Britain which flip the order of the fields as the mail
goes through. It _is_ possible to confuse the gateways and
get a message botched [try making a site in the US whose name is 'uk',
soyou get a US domain address that looks like "y...@uk.yourcompany.com",
then send a message to someone in England and see if they can reply
to you :-)], but overall I've been amazed [for years!] that it works
as well and as generally invisibly as it does.
/Bernie\
Werenfried Spit
Brits will just take any opportunity to do things differently then the
rest of the world. But given that the internet addressing standard was
made by Americans maybe theyare not to blame in this case.
------------------------------------------------------------------------
Werenfried Spit tel: 96-386 4550
Departament de Física Teòrica e-mail: sp...@vm.ci.uv.es
Universitat de València sp...@evalun11.bitnet
Kees Goossens
>It _is_ possible to confuse the gateways and
>get a message botched [try making a site in the US whose name is 'uk',
>soyou get a US domain address that looks like "y...@uk.yourcompany.com",
>then send a message to someone in England and see if they can reply
>to you :-)], but overall I've been amazed [for years!] that it works
>as well and as generally invisibly as it does.
Palindromic addresses are a problem too.
--
Kees Goossens Keep in Touch with the Dutch:
LFCS, Dept. of Computer Science JANET: k...@dcs.ed.ac.uk
University of Edinburgh, Scotland UUCP: ..!mcsun!uknet!dcs!kgg
Wiskunde is bouwen in de geest. --- Luitzen Egbertus Jan Brouwer.
Alan Cox
>
>Why is the "@node.domain" portion of "Internet" addresses on Janet (Joint
>Academic Network), the UK portion of the Internet, backwards? That is, for
>internal Janet use, an address is given as bur...@uk.ac.mi5 instead of
>the order an uninformed-in-this-instance person like me would expect. Yet I
>can give my mailer the regular RFC-822 Internet address, bur...@mi5.ac.uk,
>and the mail will get there.
>
Its a long and fairly complex story. When JANET was created it was using
4800 baud and 9600 baud links and the protocol chosen wasn't tcp. At the
time TCP/IP was a new curio of the arpanet and the US DOD rather than
tested technology. JANET used X.25 with X.29 as the upper layer. X.29 is
basically little more than X.25 with the ability to negotiate X.25 parameters
and a few crude service selection facilities. On the other hand its
performance for interactive use over a slow point-point link blew the
tcp/ip of the time out of the water (CSLIP now addresses much of this).
Thus when the JNT (who run JANET) specified the addressing there wasn't that
much of a standard to follow (other than written letter order - which the
internet uses). In their wisdom they chose the other way.
When we had a gateway between the uk and the internet for mail - initially
by uucp from ukc (kent) to uunet(I belive), software was written to swap
addresses around as they crossed the gateway. As time went on this caused
havoc especially with addresses ending in valid domain names (In fact
it still does). When we joined the internet for real the logical thing
was done - the JNT chose to make uk ip addresses the reverse of the uk
janet addresses.
Alan
Liam R. E. Quin
> Why is the "@node.domain" portion of "Internet" addresses on Janet (Joint
> Academic Network), the UK portion of the Internet, backwards?
JANET is now connected to the Internet, but it is in no way the UK portion
of the Internet. They are separate networks, with their own histories.
cos...@world.std.com (Bernie Cosell) writes:
> The perversity of the British, what more can i say?? :-) when the
> network addressing standard was being sorted out, the US [via its
> RFC mechanism] chose to do right-to-left domain parsing, and the UK
> chose to do left-to-right.
In fact, the JANET addresses predate the Internet altogether, and the RFC
mechanism. I am not sure what state ARPANET and MILNET were in when JANET
was designed; I _think_ it was only point-to-point, with no domains, then.
When the Internet people chose to use right-to-left domains, they presumably
did so without bothering to look at what other countries were using.
Isolationism in networking :-)
The JANET format was chosen because it saved a few bytes in the parsing
routines. We're talking _old_ computers :-)
Lee
--
l...@sq.com (Liam Quin) the barefoot programmer; SoftQuad Inc +1 416 239 4801
OPEN LOOK UI FAQ; Metafont list; HexSweeper NeWS game; lq-text text retrieval
`Take care of Widows, to mingle them, as Violets,
amongst Virgins and Martyrs.' Nicolas Caussin, ~1624 (trans. ~1663)
Garrett Wollman
>In fact, the JANET addresses predate the Internet altogether, and the RFC
>mechanism.
This is simply not true.
Network Working Group 4689
RFC-3 April 1969
Steve Crocker
UCLA
> I am not sure what state ARPANET and MILNET were in when JANET
>was designed; I _think_ it was only point-to-point, with no domains, then.
Point-to-point network connections and domain-style naming are
completely separate issues. The original ARPANET ran a protocol
called NCP, which used node numbers to identify hosts. When this was
replaced by TCP and IP starting in 1979, host addresses were assigned
as 10.host.0.imp, where imp was the number of the IMP (see the other
thread), and host was the port number on the IMP that the host was
connected to. These hosts were identified by names like
`HOPKINS-EECS-ALPHA' and `MIT-ML' and `WSMR-SIMTEL20', which were
listed in the HOSTS.TXT file distributed weekly by SRI-NIC. The
Domain Name System was introduced some years later, after the
precedent of using `us...@host.network' was established by early
versions of sendmail (and probably other MTAs of that era).
As someone else in this thread pointed out, when JANET was starting
up, the ARPANET has just made the final switch over from NCP to
TCP/IP. TCP was still quite new, and there was not the huge installed
base that exists now, so they went off and developed something
completely different, based on a connection-oriented network system,
and developed file transfer and mail protocols on top of that.
(Terminal access was already provided by Triple-X.)
>When the Internet people chose to use right-to-left domains, they presumably
>did so without bothering to look at what other countries were using.
You should probably ask Eric Allman, the author of sendmail, how this
came about. If he did not originate this, he was probably around when
whoever it was, did.
>The JANET format was chosen because it saved a few bytes in the parsing
>routines. We're talking _old_ computers :-)
Not significantly older than the ARPANET computers...
-GAWollman
--
Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ...
wol...@emba.uvm.edu | Shashish is the bonding of hearts in spite of distance.
uvm-gen!wollman | It is a bond more powerful than absence. We like people
UVM disagrees. | who like Shashish. - Claude McKenzie + Florent Vollant
Dik T. Winter
> In article <1pc95c$3...@access.digex.com>, Seth Grimes (gri...@access.digex.com) wrote:
> } Why is the "@node.domain" portion of "Internet" addresses on Janet (Joint
> } Academic Network), the UK portion of the Internet, backwards?
>
Right. They do not even drive on the right side of the road.
--
dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland
home: bovenover 215, 1025 jn amsterdam, nederland; e-mail: d...@cwi.nl
Dik T. Winter
> Palindromic addresses are a problem too.
Why? Requiring unlimite number of cycles because when the gateway has
reversed it decides the address needs reversing again?
Sam Wilson
> In article <C4sI0...@world.std.com> cos...@world.std.com (Bernie Cosell) writes:
> > The perversity of the British, what more can i say?? :-)
>
> Right. They do not even drive on the right side of the road.
However we do, of course, drive on the *correct* side of the road... :-)
Sam
Sam Wilson
> mechanism. I am not sure what state ARPANET and MILNET were in when JANET
> was designed; I _think_ it was only point-to-point, with no domains, then.
Others have pointed out the confusion about "point-to-point, with no
domains". JANET came into being about 1982/3, transmogrifying from an
earlier research network, SERCNET. ARPANET started in ?1969. The first
DNS RFCs date from about 1983 but 'm not sure quite when the UK NRS
order (bigendian) was defined.
> When the Internet people chose to use right-to-left domains, they presumably
> did so without bothering to look at what other countries were using.
> Isolationism in networking :-)
I thought they chose it because it was the order you write on letters
and follows the convention that the least significant part of a mail
address (the bit before the '@') comes first. From that point of view
it's very logical. It's rather a shame that it is the opposite way
round to almost every other computing hierarchy. The later (forced)
invention of the in-addr.arpa domain where internet addresses have to be
reversed to fit the model rather bears that view out. Another triumph
or market forces over technical excellence. (For those in the know I'm
DEFINITELY NOT saying that the NRS is technically excellent, just that
the choice of DNS name order isn't).
> The JANET format was chosen because it saved a few bytes in the parsing
> routines. We're talking _old_ computers :-)
Can't comment on that, but have you ever tried to do a sensible sort on
a list of DNS-order names?
Sam Wilson
Network Services Division
Computing Services, The University of Edinburgh
Edinburgh, Scotland, UK
Roger D Binns
does remain are places whose computer science subdomain is cs. If I try to
send mail to cs.berkeley.edu, it will try to send it to Czechoslovakia.
Thus, to mail any cs departments in America, you have to stick another @ on
the front.
eg to mail cs.xxx.edu, you can mail cs.xx...@berkeley.edu.
This is why most UK cs departments now have the subdomain as dcs.
Roger
--
+=============================================================================+
| cs8...@brunel.ac.uk Roger Binns Brunel University - UK |
|:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
| 'When can we hope to leave?' 'You can hope anytime Mr Taylor ...' |
| - Beneath the Planet of the Apes |
+=============================================================================+
Michael Covington
>Well, they also drive on the wrong side of the road don't they? I suppose
>Brits will just take any opportunity to do things differently then the
>rest of the world. But given that the internet addressing standard was
>made by Americans maybe theyare not to blame in this case.
Actually, it's the Americans (and French, etc.) who drive on the "wrong"
side -- due I believe to a misinterpretation of the way Napoleon made his
troops march. Driving on the left is the older practice.
>Werenfried Spit tel: 96-386 4550
> Departament de Física Teòrica e-mail: sp...@vm.ci.uv.es
> Universitat de València sp...@evalun11.bitnet
^
Hmmm, Catalan goes beyond the bounds of ASCII...
--
:- Michael A. Covington internet mcov...@ai.uga.edu : *****
:- Artificial Intelligence Programs phone 706 542-0358 : *********
:- The University of Georgia fax 706 542-0349 : * * *
:- Athens, Georgia 30602-7415 U.S.A. amateur radio N4TMI : ** *** **
Ian Phillipps
>> Seth Grimes (gri...@access.digex.com) wrote:
>> Why is the "@node.domain" portion of "Internet" addresses on Janet (Joint
>> Academic Network), the UK portion of the Internet, backwards?
>In fact, the JANET addresses predate the Internet altogether, and the RFC
>mechanism. I am not sure what state ARPANET and MILNET were in when JANET
>was designed; I _think_ it was only point-to-point, with no domains, then.
>When the Internet people chose to use right-to-left domains, they presumably
>did so without bothering to look at what other countries were using.
>Isolationism in networking :-)
I'd go further:
Internet (IP) addresses are big-endian
Internet port numbers are big-endian
In fact, Every binary (or indeed ASCII) number transmitted as part of
TCP/IP is big-endian
Janet symbolic addresses are big-endian
UUCP mail addresses are big-endian (sort of)
DNS domain names are little-endian.
Spot the odd one out? It's not JANET, is it?
Ian
--
Ian Phillipps, Unipalm Ltd, 216 Science Park, Phone +44 223 420002
Milton Road, Cambridge, CB4 4WA, England. Phax +44 223 426868
PIPEX is a division of Unipalm Ltd. - phone 0223 424616.
Ian Phillipps
>Although the reverse addresses are not usually a problem, the big one that
>does remain are places whose computer science subdomain is cs.
Don't forget such oddities as:
and as for bitnet.uucp ...
Ross Smith
Right. The right side is the wrong side. The left side is the right side.
Perfectly simple.
--
... Ross Smith (Wanganui, NZ) ............ al...@acheron.amigans.gen.nz ...
"But there I go, powering up the law-and-logic-defying Extrapolation Drive
and careering wildly off at lightspeed into the realms of iffy skiffy..."
(Iain Banks)
--
Michael Jennings
>cs8...@brunel.ac.uk (Roger D Binns) writes:
>
>>Although the reverse addresses are not usually a problem, the big one that
>>does remain are places whose computer science subdomain is cs.
>
>Don't forget such oddities as:
>
> uk.sun.com
> uk.tele.nokia.fi
>
Not to mention uk.ac.cam.cl (the Computer lab here and/or Chile)
uk.ac.cam.ch (chemistry and/or Switzerland)
uk.ac.cam.cs (Computer Science and/or Czechoslovakia)
(BTW, do the new Czech and Slovak republics have new country abbreviations?
Does anyone know what they are?)
Michael.
Peter Gutmann
>I'd go further:
>Internet (IP) addresses are big-endian
>Internet port numbers are big-endian
>In fact, Every binary (or indeed ASCII) number transmitted as part of
> TCP/IP is big-endian
>Janet symbolic addresses are big-endian
>UUCP mail addresses are big-endian (sort of)
>DNS domain names are little-endian.
>Spot the odd one out? It's not JANET, is it?
I'd say it is. Everything but Janet is consistently big or little-endian.
Janet is sort of semi-middle-endian. If it were truly big-endian
then Janet addresses would be com.bar@foo, not f...@com.bar.
f...@bar.com is little-endian.
com.bar@foo is big-endian.
f...@com.bar is weird-endian.
Peter.
David Lesher
Think of the problems this solves ;_}
--
A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu
& no one will talk to a host that's close............[301] 56-LINUX
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433
Alec Muffett - Sun IS - System Administrator
>Don't forget such oddities as:
>
> uk.sun.com
> uk.tele.nokia.fi
Actually, this is less bad than you might think, since for NRS
addresses, almost all of them fall into these categories:
uk.ac.*
uk.co.*
uk.mod.*
uk.<some govt funded science institute I cant remember, sorry>
uk.jet # joint european torus.
- so I got round this by writing a little Perl script which parses all
mail addresses in an outgoing message, flips big->little endian, and
invokes sendmail in the usual way.
Wire this into your ~/.mailrc and kiss your problems goodbye.
- alec
(alec.m...@uk.sun.com, alec.m...@uk.co.sun, ...)
---
Alec Muffett (alec.m...@sun.co.uk) Sun Microsystems IR, Bagshot, Surrey, UK
#include <stddisclaimer.h>
Phill Hallam-Baker
|>In article <1993Apr4.0...@unipalm.co.uk> i...@unipalm.co.uk (Ian Phillipps) writes:
|>>cs8...@brunel.ac.uk (Roger D Binns) writes:
|>>
|>>>Although the reverse addresses are not usually a problem, the big one that
|>>>does remain are places whose computer science subdomain is cs.
|>>
|>>Don't forget such oddities as:
|>>
|>> uk.sun.com
|>> uk.tele.nokia.fi
|>>
|>
|> Not to mention uk.ac.cam.cl (the Computer lab here and/or Chile)
|> uk.ac.cam.ch (chemistry and/or Switzerland)
|> uk.ac.cam.cs (Computer Science and/or Czechoslovakia)
|>
This is why when I had to use JANET in the UK I always posted over DECNET
into a European site and used the IN% protocol from there.
The proper solution for janet would have been to have declared a domain :-
in.round.other.the.expect then there would have been no confusion.
Alternatively since the messages should know which network they were on it
would hardly have taxed the mind for
de.desy.zws012 to be flipped by the bridge into zws012.dest.de. It was the
allowing people to access addresses from inside janet using the TCP/IP
format that was dumb and stopped the thing from working properly.
Fortunately most of the people I contacted tended to be on SPAN/HEPnet
anyway which never had that problem.
Phill Hallam-Baker
Guy Dawson
In article <C4y0y...@athena.cs.uga.edu>, mcov...@aisun3.ai.uga.edu (Michael Covington) writes:
> In article <93091.11...@EVALUN11.BITNET> Werenfried Spit <SP...@EVALUN11.BITNET> writes:
> >Well, they also drive on the wrong side of the road don't they? I suppose
> >Brits will just take any opportunity to do things differently then the
> >rest of the world. But given that the internet addressing standard was
> >made by Americans maybe theyare not to blame in this case.
>
> Actually, it's the Americans (and French, etc.) who drive on the "wrong"
> side -- due I believe to a misinterpretation of the way Napoleon made his
> troops march. Driving on the left is the older practice.
Knights of old rode on the left while jousting...
>
> >Werenfried Spit tel: 96-386 4550
> :- Michael A. Covington internet mcov...@ai.uga.edu : *****
Guy
--
-- -----------------------------------------------------------------------------
Guy Dawson - Hoskyns Group Plc.
gu...@hoskyns.co.uk Tel Hoskyns UK - 71 251 2128
gu...@austin.ibm.com Tel IBM Austin USA - 512 838 3377
Ian Phillipps
>>Don't forget such oddities as:
>> uk.sun.com
>> uk.tele.nokia.fi
> Not to mention uk.ac.cam.cl (the Computer lab here and/or Chile)
> uk.ac.cam.ch (chemistry and/or Switzerland)
> uk.ac.cam.cs (Computer Science and/or Czechoslovakia)
Gotcha! Those are Janet addresses; the ones I quoted are Internet addresses.
[See also followup by Sun's man in Bagshot, proud owner of address no.1]
BTW -
The route to the uk.tele.nokia.fi is quite interesting, as it goes the
pretty way to a site about a quarter of a mile away. Not quite as
pretty as the route to Sun, even closer, which takes in California on
the way.
>(BTW, do the new Czech and Slovak republics have new country abbreviations?
>Does anyone know what they are?)
That's in the faq in comp.mail.misc (?). Not yet is the current
answer, at least as far as the net's concerned
Phil Brownfield
>Knights of old rode on the left while jousting...
Talk about maintaining upward compatibility with obsolete standards... :-)
--
Phil Brownfield
home: ph...@blackice.uucp work: phil_br...@email.sps.mot.com
The road goes on forever, and the party never ends - Robert Earl Keen
Werenfried Spit
>
>Actually, it's the Americans (and French, etc.) who drive on the "wrong"
>side -- due I believe to a misinterpretation of the way Napoleon made his
>troops march. Driving on the left is the older practice.
days?
>> Universitat de Valhncia sp...@evalun11.bitnet
> ^
> Hmmm, Catalan goes beyond the bounds of ASCII...
------------------------------------------------------------------------
Werenfried Spit tel: 96-386 4550
Universitat de Val\čncia sp...@evalun11.bitnet
Seth Grimes
>In article <C4y0y...@athena.cs.uga.edu>, mcov...@aisun3.ai.uga.edu (Michael
>Covington) says:
>> Hmmm, Catalan goes beyond the bounds of ASCII...
>Sure. As any civilized language does.
Folks, please post what the acronym ASCII stands for, including
explanations of any further acronyms/abbreviations encountered. Then it
will be easier to see why the Catalan foot doesn't fit the ASCII glass
slipper.
Seth, gri...@oecd.fr
B.M. Buck
should be obvious from this that Catalan (which isn't any language I've
heard of in use here in America. But then, I don't know which languages
besides Spanish and Portugues are in use below the Rio Grande...)
doesn't fit the mold of ASCII.
> Seth, gri...@oecd.fr
--
-----
Buddha Buck bmb...@ultb.isc.rit.edu
(insert-file ".disclaimer")
"I'm not an actor, but I play one on TV."
Rodney Wines
> In article <C52yw...@austin.ibm.com> gu...@austin.ibm.com (Guy Dawson) writes:
> >Knights of old rode on the left while jousting...
>
> Talk about maintaining upward compatibility with obsolete standards... :-)
And swordsmen kept to the left so their sword arm would be towards a
potential oncoming enemy. I can't remember why Americans switched, bit I
think it had someting to do with the way the early wagons worked. In any
event, this all predates computers, even in folklore.
===================================================================
| Rodney Wines, Alcatel STR AG, | Phone: +41 1 465-2205 |
| Friesenbergstr. 75, CH-8055 Zurich | FAX: +41 1 465-2411 |
-------------------------------------------------------------------
| Internet: rodney...@alcatel.ch | "I always wanted roots, |
| X.400: c=CH a=arCom p=Alcatel | but if I can't have roots |
| s=Wines g=Rodney | I'll have wings." |
===================================================================
Mark C. Lawrence
R_WINES@saphir (Rodney Wines) writes:
>In <1993Apr8.0...@blackice.uucp> ph...@blackice.uucp writes:
>> In article <C52yw...@austin.ibm.com> gu...@austin.ibm.com (Guy Dawson) writes:
>> >Knights of old rode on the left while jousting...
>> Talk about maintaining upward compatibility with obsolete standards... :-)
>
>And swordsmen kept to the left so their sword arm would be towards a
>potential oncoming enemy. I can't remember why Americans switched, bit I
>think it had someting to do with the way the early wagons worked.
The way I heard it was, when two wagons (drawn by ox teams) met on
the road, the team drivers wanted to be standing to the left of their
teams, so as to (1) have their right hand to control the team, and (2)
place the drivers between the teams to avoid conflict between the oxen
of the two teams.
Mark C. Lawrence
Systems Programmer Internet: M.Law...@Forsythe.Stanford.edu
Stanford Data Center Bitnet: M.Lawrence@STANFORD
Stanford, CA 94305-4136 Tel: (415) 723-4976
P. J. Remner
In article <C52yw...@austin.ibm.com> gu...@austin.ibm.com (Guy Dawson) writes:
>Knights of old rode on the left while jousting...
Amurrica has no knights.
(And let's not have any nasty puns about "the sun never sets...")
--
/* aj...@cleveland.Freenet.Edu (Windows Freak)
* I want my... I want my... I want my Win/NT...... Or OS/2 2.1. Or a Mac
* IIvx. Whichever one of these appears on my doorstep first is fine.
*/You are not expected to understand this.
P. J. Remner
In a previous article, GA....@forsythe.stanford.edu (Mark C. Lawrence) says:
>
>The way I heard it was, when two wagons (drawn by ox teams) met on
>the road, the team drivers wanted to be standing to the left of their
>teams, so as to (1) have their right hand to control the team, and (2)
>place the drivers between the teams to avoid conflict between the oxen
>of the two teams.
>
(2) can alternatively be: so the driver could make sure that the wheel hubs
wouldn't "conflict". :-)
Phill Hallam-Baker
|>
|>In a previous article, GA....@forsythe.stanford.edu (Mark C. Lawrence) says:
|>
|>>
|>>The way I heard it was, when two wagons (drawn by ox teams) met on
|>>the road, the team drivers wanted to be standing to the left of their
|>>teams, so as to (1) have their right hand to control the team, and (2)
|>>place the drivers between the teams to avoid conflict between the oxen
|>>of the two teams.
|>>
|>
|>
|>(2) can alternatively be: so the driver could make sure that the wheel hubs
|>wouldn't "conflict". :-)
Isn't it odd that Internet chose this addressing scheme when it is the
reverse of the big-endian philosophy?
telnet zws006.desy.de
Trying...131.169.46.46
We have reversed the sense of the bytes here. logicaly the corespondance
would be
de 131
desy 169
zws016 46.46
so to write the address in internet byte order you would get
de.desy.zws016.
Another reason why this is better is that when routing you get the most
significant part of the address first. This makes the extension to
variable length addresses much easier since if I was to have a collection
of nodes all hiding behind zws006 they would logically be :-
131.169.46.46.1.2.3.4.5.6
^
If you only do 4 byte routing stop here.
Each subnet can agree locally how many bytes it needs, there is no need
for special setups...
If you want a single network addressing scheme even the 14 byte OSI addresses
look a bit sparse. Consider an MIMD based on a 256 by 256 grid of processors,
each processor might have up to several thousand processes all wanting to talk
to the outside world - if we were to use TCP/IP with the present address
scheme we have just taken up an entire class B network.
One per desk and things start to look really fried. If we allow that the
current 4+2 byte addressing scheme is insufficient and beef it up to 6+6
as being more realistic we only have two bytes left!
Phill Hallam-Baker
Phill Hallam-Baker