From: Aquarion (nicholas@aquarionics.com)
Date: Sat Sep 07 2002 - 15:32:52 BST
On Sat, Sep 07, 2002 at 12:13:57PM +0100, Simon Willison wrote:
>
> At 03:36 06/09/2002 +0200, mort wrote:
> >Possible uses:
> >
> >- Servers that exists in the logs but not in the repository can be
> >double-checked and added
> >- Warnings about unreachable servers.
> >- Statistics of most pingedpack stories, etc...
> >
> >This way information flows both from repository to blogware *and* from
> >blogware to repository.
> >
> >A XML syntax for the logs could go
> >
> ><post permalink="">
> ><timestamp> </timestamp>
> ><links>
> ><link url="">
> ><PingbackServer>
> ><status>200 OK</status>
> ><host></host>
> ><path></path>
> ><response>Thanks for pingingback!</response>
> ></PingbackServer>
> ></link>
> ></links>
> ></post>
> >
> >Comments?
I think we're veering into overengineering territory here.
I'm not sure there is a need for a central repository at all. Remember,
when you are pinging a blog, you are only pinging it /once/ for each
post you make for it. Yeah, if we were pinging every single time someone
follows a pingback, or something, we might want somewhere to remember
who does what. The centralised reasouce would have to be talked to every
time (You mentioned keeping a local copy. Bad idea, something like this
could get /huge/ if it takes off, why do you think blo.gs restricts
grabs to one per hour? even at that rate I wouldn't want their bandwidth
bill.
Also, this is an "instant blogroll" system that's currently being piled
in with a trackback replacement system, and I would *highly* recommend
they get spilit before the eventual pB spec becomes loaded with stuff
not really in it's remit.
Message sent over the Blogite mailing list.
Archives: http://www.aquarionics.com/misc/archives/blogite/
Instructions: http://www.aquarionics.com/misc/blogite/
This archive was generated by hypermail 2.1.5 : Sat Sep 07 2002 - 16:05:00 BST