Subject: RE: Z108 or Z100
From: Thomas Weigert (thomas.weigert#motorola.com)
Date: Thu Nov 22 2001 - 21:59:16 GMT
I don't think there is a right or wrong here. We collectively have to decide
which of your concepts should be part of SDL and which should be additions
in Z.108. This situation is similar to the Z.109 case. This started out with
a number of additional concepts to SDL which later were included in Z.100.
The same may be the case with Z.108. However, I would begin with producing a
document that is completely additive to SDL. In other words, produce Z.108
as a standalone document that extends Z.100. We may then discover that some
of the concepts introduced in Z.108 should be moved into Z.100 proper. Th.
> -----Original Message-----
> From: owner-meeting#sdl-forum.org [mailto:owner-meeting#sdl-forum.org]On
> We are now about to present a coherent proposal for "SDL for real
> time". But
> where to put it ?
>
> Most of the proposals for SDL are really to be considered for core SDL :
> "generalized timer type", generalized "clock type", "urgencies"
> and even a
> new scheduling mechanism
> (choice of transition to be triggered) that takes urgencies in
> account for
> real time
> (nevertheless compatible with the "old fashion" scheduler : real time
> extensions will be
> transparent for those who do not need it) A new SDL construct will be
> proposed for each
> of these concepts.
>
> Other extensions are not realy core SDL, and could just be implemented as
> annotations (task delays or channel delays or channel reliability).
>
> Finally, a third set of extensions (assumptions, performance requirements,
> resources)are realy not core SDL.
>
> Our question is : what are the rules (if any) to put something
> rather as an
> extension to Z100 than in a separate recommandation?
> Does it become mandatory ?
>
>
This archive was generated by hypermail 2a23 : Mon May 05 2008 - 20:30:55 GMT