RE: Bogus UID xx000000

[Kevin Hawkins]

> -----Original Message-----
> From: Stuart Booth [mailto:<a
href="/group/xAP_developer/post?postID=R78rRaaKOsfbpPefiCGm59O5fSlFVf2knr9lg_cZWWC1csInprNqZSDIi5syvYFdH0tJFF8j7TtTdCKC4Fzi-A">lists@x...</a>]
> Sent: 15 October 2003 10:05
> To: <a
href="/group/xAP_developer/post?postID=pGz1_16fZdrkBxJnbIsNtMFVruwvSLIDxZog57SNq2fMQxzqrNgT_pFb050S70FHNN9y1eisERuckeuOqCuVZS9hXanrcnr-">xAP_developer@xxxxxxx</a>
> Subject: Re: [xAP_developer] Bogus UID xx000000
>
> On Wed, 15 Oct 2003 02:03:15 +0100, "Kevin Hawkins"
> <<a
href="/group/xAP_developer/post?postID=CiMIIgc5lB-4jEFyKlJ9pcgqDWiP0XzTKe4MK6HjGl0dQuw_Oj7gzFdIksdIGqZqi2c83ICmWP82D8BKxHnp">lists@u...</a>>
wrote:
>
> >The first two digits I think should still represent the network -
so a
> >normal UID would be FF000000 - or actually because this is a
remote
> network
> >(bridged) maybe it should be FE000000 or something. The network ID
may
> have
> >to be different to route messages.
>
> If we imagine a future world where there is another similar system
> that we want to bridge to, would we want to be able to identify what
> type of system the message came from?
>
> Ie should a certain network type ID be reserved for xPL (FE000000),
> and another for 'SDI' (because it's a neat TLA) e.g. FD000000?
>

The first two digits are not so much a network type as a network name -
meaning that if you connected 5 different xAP networks together through
bridges then each one would have a different first two digits in the UID.
This is used by the routers (and bridges) to decide what to do with a
particular message. For example in a targeted message to a.b.c.d that is
known to have a UID of F3123400 then a local router would look at the F3
and
say - ahh the best way for me to get this to network F3 is via network F7 -
so I'll throw this out on network F7. Just like the internet does . Now
there are some issues here, any routing implies that the full broadcast
nature of xAP is compromised (as not all messages appear on every network).
This may be acceptable in scenarios where you link two remote networks or
where one network is deliberately filtered , perhaps because it is a slow
serial network.

So that sort of puts an obstacle in the way - however - it's an interesting
point as you are sort of alluding to the fact that I have done this for
xPL.
What maybe the point is here - should we reserve a few network ID's for
special non xAP networks in a way that would be useful. Rather than just
"something that isn't xAP" . xPL is a sort of special case
because it has
partial compatibility in the way it can interact with xAP messages - there
may be others - any thoughts... ?

K