Redman, Russell O. wrote:
...
I have never looked at sscanRecord before, but just reading the
documentation it looks like a superset of the scanRecord with
> somewhat clearer documentation. Would an sscanRecord be a direct
> substitute for a scanRecord (i.e. backwards compatible)?
No. (That was the reason for the name change.) There are two
principal sources of incompatibility:
1) The scan record depended on an external agent to detect when
operations it started were completed. The sscan record uses
putCallback for this purpose. For some kinds of scan databases,
putCallback completion is much easier to implement (you don't
have to do anything) and it's certainly more bulletproof.
However, for scans in which some of the processing occurs as
the result of a ca_link, or a ca_monitor, EPICS putNotify can't
trace the processing, and putCallback therefore doesn't correctly
detect completion. In cases like this, the database designer must
arrange for some record, whose processing *will* be traced by
putNotify, to refrain from firing its forward link until all the
processing started by the sscan record is done.
This case is kind of similar to the way the scan record operated,
but now the external completion detector can be built into the
database being driven by the sscan record, and instead of processing
the scan record, it just fires its forward link to tell EPICS
everything is done. (This is good, because now the database will
also signal completion to any other CA client that drives it.)
2) The detector names are different: the scan record had 15 detectors
numbered 1...F, while the sscan record has 70, numbered 01...70.
I am puzzled, however, because the documentation explicitly states
> that it uses recDynLink which is normally distributed as part of the
> package. If the purpose of the suggested changes is to escape from
recDynLink, how would we be better off using sscanRecord?
The plan is to convert the sscan record so that it doesn't use
recDynLink. I haven't done this yet.
-----Original Message-----
From: Marty Kraimer [mailto:[email protected]]
Sent: Friday, 2003 December 05 1:02 PM
To: [email protected]
Subject: scanRecord for 3.14
The next 3.14 release will support put callbacks for database
channel access
links. This is being done so that synapps can remove it's
dependence on recDynLink.
3.13 and early versions of 3.14 has support for scanRecord.
This was removed
from 3.14 because it required recDynLink. Now that put
callback links will be
supported the scanRecord could be easily modified so that it
no longer requires
recDynLink.
The question I have is does anyone use scanRecord?
Synapps provides sscanRecord which provides more
functionality than scanRecord
so is scanRecord still needed?
Marty Kraimer
--
Tim Mooney ([email protected]) (630)252-5417
Beamline Controls & Data Acquisition Group
Advanced Photon Source, Argonne National Lab
- References:
- RE: scanRecord for 3.14 Redman, Russell O.
- Navigate by Date:
- Prev:
Re: scanRecord for 3.14 John Maclean
- Next:
Re: scanRecord for 3.14 Marty Kraimer
- 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: scanRecord for 3.14 Redman, Russell O.
- Next:
Re: scanRecord for 3.14 Marty Kraimer
- 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
|