TOPIC:

In a TTCN-3 program, connections are between ports. Yet,
tciSendConnectedBC sends to "all components being connected to this
port". Why not the ports? Specifically, why is only a component, and
not a port passed to tciEnqueueMsgConnected, even though the component
identifies the port?

To me, this suggests that the connection mapping needs to maintained in
two different places, i.e. the TE (for deducing the port from the
component in tciEnqueueMsgConnected), and in the CH (for calling
tciEnqueueMsgConnected on the correct component). Is this really
intended?

--
Regards,
Mike

Hello Mike,

It is not allowed to connect a port of a component A to two different
ports of component B. If component A is the sender and component B the
receiver as in your example, when enqueuing the message at component B
it is sufficient to know the sending port to determine the corresponding
receiving port. The relation between ports is known from the tciConnect
operations.

The restriction mentioned above is case 7 in figure 7 in edition 3.4.

Best regards

Thomas


>

---

Original Message

---

>From: active_ttcn3 : mts stf133 ttcn version 3 - active
>members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] On Behalf Of ext
>Michael Sperber
>Sent: Tuesday, 23. December 2008 11:35
>To: This email address is being protected from spambots. You need JavaScript enabled to view it.
>Subject: Another TRI question
>
>In a TTCN-3 program, connections are between ports. Yet,
>tciSendConnectedBC sends to "all components being connected to this
>port". Why not the ports? Specifically, why is only a component, and
>not a port passed to tciEnqueueMsgConnected, even though the component
>identifies the port?
>
>To me, this suggests that the connection mapping needs to maintained in
>two different places, i.e. the TE (for deducing the port from the
>component in tciEnqueueMsgConnected), and in the CH (for calling
>tciEnqueueMsgConnected on the correct component). Is this really
>intended?
>
>--
>Regards,
>Mike
>

"Deiss, Thomas (NSN - DE/Duesseldorf)" <This email address is being protected from spambots. You need JavaScript enabled to view it.> writes:

> It is not allowed to connect a port of a component A to two different
> ports of component B. If component A is the sender and component B the
> receiver as in your example, when enqueuing the message at component B
> it is sufficient to know the sending port to determine the corresponding
> receiving port. The relation between ports is known from the tciConnect
> operations.

My question was about where the mapping is maintained at run time: My
reading suggests that it needs to be maintained in both the TE and the
CH, which seems redundant. Is this indeed the intention?

--
Regards,
Mike

Hello Mike,

It should be sufficient to maintain the mapping at the TEs. The CH knows
the receiving component, hence it knows the corresponding TE. This TE is
then able to determine the receiving port from the receiving component
and sending port.

Best regards

Thomas

>

---

Original Message

---

>From: active_ttcn3 : mts stf133 ttcn version 3 - active
>members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] On Behalf Of ext
>Michael Sperber
>Sent: Tuesday, 06. January 2009 09:02
>To: This email address is being protected from spambots. You need JavaScript enabled to view it.
>Subject: Re: Another TRI question
>
>"Deiss, Thomas (NSN - DE/Duesseldorf)" <This email address is being protected from spambots. You need JavaScript enabled to view it.> writes:
>
>> It is not allowed to connect a port of a component A to two different
>> ports of component B. If component A is the sender and
>component B the
>> receiver as in your example, when enqueuing the message at
>component B
>> it is sufficient to know the sending port to determine the
>corresponding
>> receiving port. The relation between ports is known from the
>tciConnect
>> operations.
>
>My question was about where the mapping is maintained at run time: My
>reading suggests that it needs to be maintained in both the TE and the
>CH, which seems redundant. Is this indeed the intention?
>
>--
>Regards,
>Mike
>

"Deiss, Thomas (NSN - DE/Duesseldorf)" <This email address is being protected from spambots. You need JavaScript enabled to view it.> writes:

> It should be sufficient to maintain the mapping at the TEs. The CH knows
> the receiving component, hence it knows the corresponding TE. This TE is
> then able to determine the receiving port from the receiving component
> and sending port.

But how does tciSendConnectedBC, say, know what components to call
tciEnqueueMsgConnected on, if the CH doesn't know the mapping?

--
Regards,
Mike

Hello Mike,

I focused just on the case with unicast sending.
For broadcast I have to say that I am not really sure, it would be good
if somebody else could answer this question.

Best regards

Thomas


>

---

Original Message

---

>From: active_ttcn3 : mts stf133 ttcn version 3 - active
>members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] On Behalf Of ext
>Michael Sperber
>Sent: Tuesday, 06. January 2009 14:56
>To: This email address is being protected from spambots. You need JavaScript enabled to view it.
>Subject: Re: Another TRI question
>
>"Deiss, Thomas (NSN - DE/Duesseldorf)" <This email address is being protected from spambots. You need JavaScript enabled to view it.> writes:
>
>> It should be sufficient to maintain the mapping at the TEs.
>The CH knows
>> the receiving component, hence it knows the corresponding
>TE. This TE is
>> then able to determine the receiving port from the receiving
>component
>> and sending port.
>
>But how does tciSendConnectedBC, say, know what components to call
>tciEnqueueMsgConnected on, if the CH doesn't know the mapping?
>
>--
>Regards,
>Mike
>