Hello!
> I was just wondering if someone could help me in the understanding of
> the TCI standard.
>
> My first question is on the operation tciGetModuleParameters which is
> called by the TM and implemented by TE. The text states"The TE provides
> to the management a list of module parameters of the identified module.
> If no module parameters exist, an empty module parameter list is
> returned. If the TE cannot provide a list, the distinct null value shall
> be returned."
> So my question here is how does the TE return (value) information about
> module parameters which have no default value specified?
> A null value is not enough since then the TM has no way to figure out
> the type of the module parameter - only its name.
> It would be nice if the text could also be more elaborate on this issue.
I guess there is an omission even in the TTCN-3 edition 3 standard here,
you cannot get that information through the TCI-TM API. I guess this
requires yuo to know the types a priori during the compile time.
We have solved this problem by giving our users a function to get
that information. This funcion is:
TciType tciGetModuleParType
(TciModuleParameterIdType parameterId);
> My second question is about verdict values. Currently a verdict value is
> a (Tci)Value which stores an integer value which returned as part of the
> tciTestCaseTerminated
> operation. At least in the c-mapping only a reference to the object is
> returned. Here I wonder if anybody has ever considered that the TE may
> delete that object before the TM has any change to access the verdict?
> Or is there anything else in the standard that garuantees that this will
> not happen?
If the specification indeed has forgotten to mention about the lifetime
of the objects, I would think that vendors need to provide this detail.
A resonable implementation would give the user the duration of the
function call time to copy any necessary information to be used
after the function call such as tciTestCaseTerminated returns.
> So my question is why this has not been simply mapped to an integer
> (alias) parameter in the operation definition .. and instead to an
> object?
I guess in European specs very simple solutions are not possible. :)
Best regards,
Vesa-Matti
--
Vesa-Matti Puro
OpenTTCN Oy
E-mail:
This email address is being protected from spambots. You need JavaScript enabled to view it. * Tel.: +358 440396461 * Web:
www.openttcn.com