In the CLHS, the entry for pathname-host says it returns a string or
list of strings (and some other stuff). I can understand a string,
but what systems return a list of strings? What did that mean?
How does one reconcile that with the fact that host-namestring only
returns a string?
Just curious about some history,
Ray
Raymond Toy <···@rtp.ericsson.se> writes:
> In the CLHS, the entry for pathname-host says it returns a string or
> list of strings (and some other stuff). I can understand a string,
> but what systems return a list of strings? What did that mean?
VAX VMS had a protocol (over DNA? not sure...) that used to let you
say:
foo::bar::baz::
which meant like the old uucp "!" notation when sending mail.
Also, in principle, other operating systems could do this. On ITS,
you could only have one host/device but they were often merged.
e.g., the "AI" device was something that would talk to the AI machine
[back before domain names], the "ML" device got you the ML machine,
and same for DM and MC, but you could use devices like AIMLMC: to route
uselessly through a lot of machines wasting a lot of resources that
were better not wasted. There were originally only 4 ITS machines,
incidentally, and they were all in direct contact with one another.
> How does one reconcile that with the fact that host-namestring only
> returns a string?
(host-namestring '("FOO" "BAR" "BAZ")) => "FOO::BAR::BAZ::" ?? I dunno.
> Just curious about some history,
Not unreasonable.
>>>>> "Kent" == Kent M Pitman <······@world.std.com> writes:
Kent> Raymond Toy <···@rtp.ericsson.se> writes:
>> In the CLHS, the entry for pathname-host says it returns a string or
>> list of strings (and some other stuff). I can understand a string,
>> but what systems return a list of strings? What did that mean?
Kent> VAX VMS had a protocol (over DNA? not sure...) that used to let you
Kent> say:
Kent> foo::bar::baz::
Kent> which meant like the old uucp "!" notation when sending mail.
And how many people remember what that old uucp notation means? :-)
I do, vaguely, because I had to use that way back when, but even then
it was almost completely replaced.
[snip]
Thanks for the history!
Ray
Raymond Toy wrote:
> >>>>>"Kent" == Kent M Pitman writes:
>
>
> Kent> Raymond Toy writes:
> >> In the CLHS, the entry for pathname-host says it returns a
> string or
> >> list of strings (and some other stuff). I can understand a string,
> >> but what systems return a list of strings? What did that mean?
>
> Kent> VAX VMS had a protocol (over DNA? not sure...) that used to
> let you
> Kent> say:
>
> Kent> foo::bar::baz::
>
> Kent> which meant like the old uucp "!" notation when sending mail.
>
> And how many people remember what that old uucp notation means? :-)
> I do, vaguely, because I had to use that way back when, but even then
> it was almost completely replaced.
I have vague memories of UUCP notation.
However, one can conceivably say that
cl-prompt> (defvar *google* (parse-namestring "http://www.google.org"))
==> *GOOGLE*
cl-prompt> (pathname-host *google*)
==> ("www" "google" "org")
cl-prompt> (pathname-device *google*)
"http"
Of course I am armed with a large carpet under which I am sweeping a lot
of unwanted dust, but that is another matter altogether :)
Cheers
--
Marco Antoniotti
From: Christopher C. Stacy
Subject: Re: pathname-host results?
Date:
Message-ID: <uu1drn4lo.fsf@dtpq.com>
>>>>> On 24 Mar 2003 18:14:53 -0500, Kent M Pitman ("Kent") writes:
Kent> There were originally only 4 ITS machines, incidentally,
Kent> and they were all in direct contact with one another.
By which he means "through the ARPANET, and later the CHAOSNET".