Subject: Re: Gate endpoint constraints
From: Rick Reed TSE (rickreed#tseng.co.uk)
Date: Tue Mar 06 2001 - 20:11:36 GMT
Eckhardt Holz at holz#informatik.hu-berlin.de wrote on 26/02/2001 14:18:
> Hi,
> It seems that the clarification for the introduction of implicit
> channels (TDD14r1.doc ETSI-meeting) does not explicitly take into
> account the endpoint constraints of gates (gate G out to atleast BType).
> There should be atleast a note mentioning the additional requirements.
Eckhardt is correct - for <block definition> and <block reference> and
<block diagram> - BUT NOT for <block reference area> because in the latter
case the gate is represented by a <gate property area> and in this case a
constraint is not allowed. Is it allowed for there to be a constraint on the
gate on the block diagram, but to have no constraint corresponding gate on a
reference to the block? Processes are the same as blocks.
However, it seems to me that TDD14r1 correctly emulates TDB19 which also
ignored the constraints.
I have revised TDD14r1 to TDE14 with changes to take account of constraints.
> It seems also to me that a gate with such a constraint must always be
> directly connected to an agent/state conforming to the constraint, i.e.
> both must be directly contained in the same agent.
> I donīt know whether this is intended, atleast it is easier to verify.
>
> Is the sentence:
> "The <identifier> of <textual endpoint constraint> must denote a type
> definition of the same entity kind as the type definition in which the
> gate is defined" (Z.100 Sect. 8.1.6)
> a leftover or is there any reason not to allow an block agent having a
> gate
> with a state-type-endpoint-constraint (or a process endpoint). For state
> (types)
> it actually does not make any sense to have a constraint of the same
> entity
> kind (there are no channles between states).
>
> I think the sentence can be removed.
I agree.
-- Rick Reed - rickreed#tseng.co.uk Tel:+44 1455 55 96 55 Fax:+44 1455 55 96 58 Mob.:+44 7970 50 96 50
This archive was generated by hypermail 2a23 : Mon May 05 2008 - 20:30:55 GMT