SAMP v1.0 WD published

[Mark Taylor]

On Mon, 7 Jul 2008, Alasdair Allan wrote:

> In your list of changes you have
>
> * call(), callAndWait() - now returning msg-id
>
> I'm presuming that should be "call(), callAll()"...?

thank you, yes I wrote the wrong thing and "call(), callAll()" is 
what I should have said.

> In which case, how can callAll( ) return a single msg-id?
>
> Surely each message generated by the Hub intended for each client will have a 
> separate msg-id. Or are you arguing here for a single msg-id for all of the 
> copies of the message passed on from the client by the Hub. In other words, 
> messages going to different (end) clients (from the originating client, via 
> the hub) all with the same id...? In which case I'd argue that's a bad 
> thing...

Yes, messages going to different end clients as a consequence of
the same callAll() call use the same msg-id.

> Confused... can you tell me what's intended here? Previously this wasn't a 
> constrain in the specification. Is this an oversight, an unintended 
> consequence, or did you actually mean to do this?

it's an unintended, but not unforseen, consequence.  I agree that it
is a constraint in that it means the hub re-uses message IDs.
I believe that it's harmless; it doesn't lead to unavoidable 
confusion, since any later reference to the message ID will also 
have access to the identity of the recipient, so can identify 
which of the sends with the same msg-id is being talked about.

Can you think of any particular use cases in which this behaviour
would cause problems?

Mark

-- 
Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-928-8776 http://www.star.bris.ac.uk/~mbt/