Re: Web server name?

Martijn Koster (m.koster@nexor.co.uk)
Fri, 20 Jan 1995 19:57:21 +0100


> Nick Arnett writes in <ab44baf2060210049877@[192.187.143.12]>:
[ that he wants to extract a long server name from a Web server: ]
> >I'm thinking that perhaps it should
> >be an element in the header of the default document. Ideas?

I don't think it should be a header element of a specific doument,
that's confusing (what are the semantics of this tag elsewhere?), and
would take ages to get through an HTML WG, and the last thing we want
is yet another browser-specific new tag.

> Since from my viewpoint this is an attribute of the server rather than the
> documents served, it should be an HTTP header, something like
> "Server-Longname: Thomson Consumer Electronics Marketing & Sales".

This is a slippery slope, you can end up including far too much info
in what is a only a transport protocol after all. The majority of web
applications would ignore it, and it's quite redundant to send it
every time.

I do think this sort of information would be neat, especially if it's
not limited to "Server-Longname", but could include email addresses of
the webmaster, offical host names etc. Hey, you could provide lots of
useful info, like descriptions of services, contact numbers etc.

<soapbox type="aliweb">

One way to do it is of course to use a description on a known URL (say
/site.idx) in a known format (say based on the IAFA templates). E.g:

http://web.nexor.co.uk/site.idx:

Template-Type: SITEINFO
URI: /
Host-Name: web.nexor.co.uk
Admin-Handle: webmaster@nexor.co.uk
Owner-Organization-Name: NEXOR
Description: The WWW Server at NEXOR, an OSI X.400/X.500 software company.
Keywords: NEXOR, OSI, X.400, x400, x-400, X.500, x500, x-500

This is implementable today; reasoneably standardised (in that the
IAFA templates have been around for a while and are an ID, and there
is currently no competing alternative), and can be quite efficient if
the client gets it when needed, and caches it for future use.

</soapbox>

But then, I would say that :-) The disadvantage is that such an index
file in turn might give you far more than you want to know. I think
this is preferable to hacking individual requirements into HTML or
HTTP.

-- Martijn
__________
Internet: m.koster@nexor.co.uk
X-400: C=GB; A= ; P=Nexor; O=Nexor; S=koster; I=M
X-500: c=GB@o=NEXOR Ltd@cn=Martijn Koster
WWW: http://web.nexor.co.uk/mak/mak.html