discuss: Which Licenses Should LDP Recommend? GFDL
Subject:
Re: Which Licenses Should LDP Recommend? GFDL
From:
Randy Kramer ####@####.####
Date:
27 Apr 2001 13:36:50 -0000
Message-Id: <3AE97507.5510@fast.net>
Andreas Thienemann wrote:
> > Would this really work, though?
> Well, we would either need to provide a robots.txt file, which prohibits
> robots from accessing the ldp-mirror.
> Or we could modify the doc-book template to include some meta-tags in the
> html-header that says something about "robots: noaccess, nofollow".
>
> That would suffice for not having the mirrors show up in search engines.
(Because, AFAIK, most major search engines honor the robots.txt and
relevant meta-tags.)
The points Andreas made here are important!
>
> But in that case the whole idea of providing mirrors is moot.
> The current ldp-site is not designed to persuade people to select a
> mirror, they (probably) access the howto documents on the main site and
> never visit a mirror.
>
> Thus I'd suggest to modify the main index-page of the ldp-site to redirect
> the visitor to a geographically close mirror automatically.
Would that be an automatic redirection once he chose a document for
download, or would he be redirected and then have to search for and
choose the document to be downloaded?
Sounds good, can that be done? Is it truly geographic? Can it be
"Internet geographic" (i.e., based on nearness to major trunks, etc.)?
>
> That way a mirror gets selected automagically and the mirrors themselves
> don't turn up in searchengines.
Just for the sake of discussion, I envisioned something slightly
differerent (because I didn't consider the possibility of an automatic
redirection of the download after choosing a document on the LDP site):
I envisioned indexing only the main LDP site, but making a list of
mirrors easily accessible. Then we try to encourage people to learn
(and bookmark) the name of a reliable mirror near them (Internet
"geograpically"). They use the index to find the document they want,
then download it from their local mirror.
Perhaps there are some ways to do more than encourage people to use a
local mirror:
Limit download access from the LDP site by password, connections, or
user limits, and prominent reminders to consider the use of a mirror
site.
Set up user accounts on LDP (or a cookie) that would maintain the user's
choice of local mirror. He could ask for the download on the main LDP
site, but it would really come from his chosen local mirror.
All of the above is on the assumption we are addressing three
objectives:
1. Avoid the distribution of stale documents.
2. Minimize the "pollution" of Internet search engine results with hits
on mirror sites.
3. Avoid overload of the main LDP site. (This might not be a problem
today, but would become one if everyone attempted to download from the
main site instead of the mirrors.)
IMHO, the concept of indexing only the main LDP site sounds promising.
We probably need more discussion on the exact implementation before
proceeding.
Randy Kramer
