Subject: SDL-News: Re: signal delay
From: Rick Reed TSE (rickreed#tseng.co.uk)
Date: Sat Oct 27 2001 - 12:00:49 GMT
Become an SDL Forum Society member <http://www.sdl-forum.org/Society/members.htm>
The originator of this message is responsible for its content.
-----From Rick Reed TSE <rickreed#tseng.co.uk> to sdlnews -----
In response to: pa.Ke#t-online.de (Patrik Kessler)
> I have a question with respect to the signal-delays in SDL.
> SDL provides signal-delays on channels.
>
> However, the SDL-specification does not contain concrete statements describing
> these time durations.
You are quite correct that the amount of delay on channels cannot
(currently) be specified in SDL. The issue is currently being studied by
ITU-T under Q.23/17 (previously Q.7/10) and I suggest you contact either the
rapporteur for the study: Daniel Vincent
<daniel.vincent#rd.francetelecom.fr>
or
Dieter Hogrefe <hogrefe#itm.mu-luebeck.de>
who is the project leader of the associated European Interval project.
The issue is complicated because in some cases the delay may not be constant
(for example if the channel itself represents an underlying communication
path with error checking and correction and retransmission). There needs to
be some specification of the distribution of the delay (and perhaps there
also needs to be the concept of channels that may lose or even corrupt
signals).
In the meantime each tool needs to have a mechanism for giving the delay
some value. It seems that the Telelogic tool normally treats the value as
"insignificant" - which arguably is within the language semantics (But not
very useful for modelling purposes).
It would be possible to introduce some extra elements in the SDL model to
force the delay: for example, a process on the channel that explicitly
delays the signals by a fixed duration.
> Which tool has implemented this feature without supplement? The tool I know
> (Telelogic tau) does not generate signal-delays on channels. This means,
> events can take place through Timer Signals at unequivocally times (unique
> time points). And this is a contradiction to SDL. In my current project, I
> require solid event-times.
I think that Cinderella may have a different approach to this, and I am not
sure that it is always the case that Telelogic Tau does not model the delay
at all. However, I expect the tool vendors to give a more helpful response
on this.
From your text above it is not exactly clear what you mean by:
1. "events can take place through Timer Signals at unequivocally times
(unique time points)";
and
2. "require solid event-times".
I believe the semantics of Timer Signals to be well defined and unambiguous
in SDL. If you can indicate more precisely what is wrong with the language
definition I would like to know. I think the problem that you have is
related only to the way that SDT models the passing of time.
As for as "solid event-times" is concerned, you may do better to examine the
value of NOW directly - but again you will have to depend on a good tool
implementation for modelling.
> With the tool, this modelling would be possible. But, with SDL, this is
> incorrect.
>
> Are there any experiences with such kind of problems?
The short answer is "yes", and I expect others to provide more explicit
answers on their experiences.
-- Rick Reed - rickreed#tseng.co.uk Tel:+44 1455 55 96 55 Fax:+44 1455 55 96 58 Mob.:+44 7970 50 96 50--End text from Rick Reed TSE <rickreed#tseng.co.uk> to sdlnews --- For extra SDL Forum Society benefits join at <http://www.sdl-forum.org/Society/members.htm>
This archive was generated by hypermail 2a23 : Thu May 09 2013 - 16:05:49 GMT