Henrik Nordstrom wrote:
>
> Wojtek Sylwestrzak wrote:
>
> > In practice: do you know of many replicas not being mirrors,
> > acoording to this definition?
>
> Not really, a replica is most of the time also a mirror. The exception
> is when you have two way replication of information and neither of the
> replicas is the authoritive origin, but I haven't yet seen this model
> actively used in web services.
>
> For the taxonomy draft it may be sufficient to define mirror in the
> broader concept.
>
> When discussing the actual transfer of information the situaction gets a
> bit more complicated.
>
Right, but then we'd have to involve things like distributed file
systems, and there's already a lot of theory behind them, so we probably
shouldn't reinvent it anyway.
also, i think it's better not to broaden the scope of the taxonomy draft
beyond the web if possible, for obvious reasons.
--w
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:26 MST