From: Aquarion (email@example.com)
Date: Tue Sep 10 2002 - 21:48:38 BST
On Tue, Sep 10, 2002 at 07:35:30PM +0000, Ian Hickson wrote:
> On Tue, 10 Sep 2002, Aquarion wrote:
> > >
> > > I still see pingback as being, on the client side, just a blogging tool,
> > > but I don't see why, on the server side, I shouldn't be able to catch
> > > links to pages other than my blog entries.
> > Because everything you get with a pingback - that someone has followed a
> > link from $reference to $page - is already (and automatically) filed
> > away in your server log without there being any other protocal involved.
> In practice, this isn't so. Referrer logs are noisy, they don't point to
> permalinks, and they contain lots of links from search engines and
> aggregation services like weblogs.com.
All of which are filterable.
> > The /only/ benefit pingback has over scanning log files is the permalink
> > that it gives back to the refering article (that is, When I link from
> > index.php, the pingback tells you the link comes index.php?id=404,
> > because that's the place you'll /always/ be able to get it)
> That's a non-trivial advantage.
For HTML documents, I'd agree. For documents that cannot display the
resulting information (text files, videos, basically most things that
arn't X/HTML) having the information recorded seperatly isn't worth it.
> > which can be used to create a list of "further reading" links. Doing
> > this for anthing other than x/html isn't as useful, because part of
> > the point of the system is you have a list of links of people who
> > talked about this.
> It would be useful using Simon's idea of getting the links out of the
> system by XML-RPC -- a client side "related links" system which instead of
> using a centralised server like Alexia depends on the site itself.
This is an interesting idea :-)
Message sent over the Blogite mailing list.
This archive was generated by hypermail 2.1.5 : Tue Sep 10 2002 - 23:05:00 BST