Re: [vox] Overheard...
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [vox] Overheard...
On Wed, 15 May 2002, Jeff Newmiller wrote:
> On Wed, 15 May 2002, ME wrote:
> > On Wed, 15 May 2002, Peter Jay Salzman wrote:
> > > a network gateway, i think, is a computer that has two IP addresses and
> > > acts as an interface between two networks. i'm pretty sure this doesn't
> > > describe (any? most?) DSL modems.~
> > Unless you are MS. Then, a gateway is, "an application layer redirector or
> > proxy server". This was according to their MSNE from a few years back and
> > their NT 4.0 testing.
> > Checking nearly any RFC shows that the rest of the world thought a gateway
> > was a layer 3/4 (mostly 3) router.
> Those uses are deprecated.
You are entirely correct. :-)
Yep, that is why reference is "past tense" as in "thought", "was" and then
"(mostly 3) router." :-/ (sorry. my fault for not being more explicit.)
> Look at the definition of "gateway" in RFC1392:
> The term "router" is now used in place of the original definition
> of "gateway". Currently, a gateway is a communications
> device/program which passes data between networks having similar
> functions but dissimilar implementations. This should not be
> confused with a protocol converter. By this definition, a router
> is a layer 3 (network layer) gateway, and a mail gateway is a
> layer 7 (application layer) gateway. See also: mail gateway,
> router, protocol converter.
> I think "router" is a much more appropriate term for a Layer 3 function.
> The software that interfaces functionally similar but different layer 7
> protocols (like, for example, an I2C interface to motherboard health
> monitoring sensors to an SNMP TCP service) is more of the kind of function
> I think of that "opens gates" between different domains.
> The term "default gateway" is unfortunately probably never going to go
> away, though.
I prefer "proxy" or "relay" or "Application layer redirector" when working
with something at the application layer.
I will agree that "router" is a better word choice for the layer3/4 device
that "routes" packets. (Very specific, and not polluted with a history
as much as router.)
My complaint was their attempt to force a new definition to such a word
that has such an ambiguous history. (You know, 'there is only one right
"gateway" has older RFC (pre 1000) that use gateway still and no new
revisions to those RFC have been made to remove this terminology.
Also, re-use of this word leads to ambiguity with one example being the
one you stated with default gateway and the "route" command for
specifying a "gateway" for a subnet.
And yet, even though MS requires a diff def for "gateway" look in the MS
Windows "Networking control panel", and they call the IP address for the
router on your subnet the "default gateway" and then if you have multiple
subnets on the same NIC, the advanced tab shows "gateway" for the router
IP addresses on the specific subnet. (If you are going to tout a new
definition to a word, then you should ensure you only use the word in the
same you you have defined it to be used. Aother complaint about them.)
Then there is "BGP" which works to provide information between routers and
routing info as well as "EGP" with the "G" as "gateway". (Plenty of others
that also use gateway in the router sense.) Changing them to BRP, ERP
(etc) would be good. (RIP is ahead of its time? heh-heh)
Too much legacy loaded in "gateway" to force test-takers to all agree that
"gateway" is necessarily, and absolutley "an application layer redirector
or proxy server."
I contest their assertion in use of "gateway", but I do not say we should
exclusively use "gateway" over "router".
-----BEGIN GEEK CODE BLOCK-----
GCS/CM$/IT$/LS$/S/O$ !d--(++) !s !a+++(-----) C++$(++++) U++++$(+$) P+$>+++
L+++$(++) E W+++$(+) N+ o K w+$>++>+++ O-@ M+$ V-$>- !PS !PE Y+ !PGP
t@-(++) 5+@ X@ R- tv- b++ DI+++ D+ G--@ e+>++>++++ h(++)>+ r*>? z?
------END GEEK CODE BLOCK------
decode: http://www.ebb.org/ungeek/ about: http://www.geekcode.com/geek.html
vox mailing list