Re: exception handling
"Cezary Zielinski" <[email protected]>
| Newsgroups | gmane.science.robotics.orocos.user |
|---|---|
| Message-ID | <[email protected]> |
The classification of errors is a good idea. In MRROC++ we distinguish 3 types of errors: - non-fatal errors (computational errors, e.g. associated with kinematic computations - violation of joint limits) - fatal errors (malfunction of a robot, e.g. someone cuts off power) - system errors (degradation of distributed control system capabilities, e.g. someone turns off one of the nodes of the computer network) Perhaps a finer granularity is needed. Cezary ----- Original Message ----- From: "Jorge Cabrera Gámez" <[email protected]> To: "Herman Bruyninckx" <[email protected]> Sent: Tuesday, March 05, 2002 6:11 PM Subject: Re: [Orocos] exception handling > Hi, > > Herman Bruyninckx wrote: > > > On Tue, 5 Mar 2002, Cezary Zielinski wrote: > > > > > Exception handling should be synonymous to handling error situations. We > > > should not try to use that in situations that are not disastrous to the > > > system. If a disaster occurs the time restrictions are not valid any more, > > > so the indeterminism that you are speaking about is no longer an issue. I > > > use exception handling extensively in MRROC++ and don't have problems with > > > that. > > > > There are different levels of error situations! Some are indeed a > > `disaster', but others reflect nothing more than a change in the > > currently valid assumptions in the robot control. The latter ones can > > and should be dealt with in real-time, in some cases. > > > > The terminology `exception' is indeed a bit overloaded :-) I think > > Anders used it in the C++ interpretation of the word, which is the one > > I also use in the previous paragraph. > > I agree with Herman. I think we need to define a taxonomy or typology of > "errors", ranging from catastrophic failures to simple "warnings". I think we > don't need to specify all possible errors, but the general clases. Then, I will > try to enforce the adoption of an uniform handling method for each error > typology that should be part of the Orocos' component specification. > > Regards. > > Jorge > > =========================================================== > Jorge Cabrera Gamez > Facultad de Informatica > Universidad de Las Palmas de Gran Canaria > Campus Universitario de Tafira > 35017 Las Palmas de Gran Canaria. > SPAIN > Tel: +34 928 458758 > Fax: +34 928 458711 > E-mail: [email protected] > ============================================================ > > > > > _______________________________________________ > Orocos mailing list > [email protected] > http://mail.mech.kuleuven.ac.be/mailman/listinfo/orocos