I promise that this will be short... I find myself agreeing with almost all
that John Winans says. Some brief notes:
1. I mentioned DRAMA in order to be even-handed. I agree with John's comments.
2. As a matter of interest, why is the time stamp not the 11th quark? Is it
never required along with graphic or control information?
3. The severity quark is fine for VALID / INVALID. The status quark has the
problem that it is 16 bits and not easily extensible, hence much of the
discussion to date. A message string quark would need to be more than a
canned translation of the status
4. There is, I think, a distinction between a status which comes back to a
client in a transaction response and a status which is part of a record's
state.
5. I would happily throw out the requirement that status codes be converted
to message text. It has been interest in upwards compatibility which has
kept that alive. All along we have agreed that it is text which flows on
the wire.
6. Small subsystem-specific error codes still don't address the "package A
calls routine in package B which calls routine in package C which fails"
problem unless B converts all C codes to B codes before forwarding them
to A (which is a bit of a pain for B; I've several times written systems
which were low-level and returned things like NULL to indicate failure,
only to have to invent error codes at the next level up and to wish I'd
done the job properly in the first place).
William
- Navigate by Date:
- Prev:
Re: EPICS status codes proposal winans
- Next:
Re: EPICS status codes proposal watson
- Index:
1994
<1995>
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
- Navigate by Thread:
- Prev:
Re: EPICS status codes proposal watson
- Next:
Re: EPICS status codes proposal Jim B. Kowalkowski
- Index:
1994
<1995>
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
|