Experimental Physics and Industrial Control System
|
On 03/12/2014 20:44, Mooney, Tim M. wrote:
I agree with a lot of this.
The more common pattern is to have <module> export <module>.dbd for
inclusion by applications, and I would lean towards proposing synApps
make a step in that direction.
I've come to agree with this, mostly because people's first guess seems often enough to be that "module.dbd" is what they should include. However, it's not the sort of change you can creep up on, as far as I can tell; you either go there or you don't, and I'm wary of causing configuration trouble for users who want to upgrade existing applications.
Upgrade paths are always ... what about this.
If you decide to e.g. name your example and test dbd files
<module>Example.dbd or <module>Test.dbd, you can have <module>.dbd and
<module>Support.dbd as identical files (or file plus link) in the
"outer" support module for as long as you want to stay compatible.
User applications were not supposed to include <module>.dbd files from
synApps in the first place.
It was tolerated in 3.14, it will fail in 3.15, and thus has to be
changed anyway. Making <module>.dbd the correct file to be included
might in the end save users from trouble: If their application was
already set up correctly, the "new" <module>.dbd will just work. If
their application was not set up correctly, 3.15 requires it to be
changed anyway. Why not change it to import resources under the final
<module>.dbd name?
At some point (before you remove the <module>Support.dbd files/links),
your synApps version update script for user applications can just
replace <module>Support.dbd with <module>.dbd - at that point this
should work, as these two files will have been the same for some time.
Tests and examples should rather be moved into embedded self-contained TOP structures - 3.15 provides
convenient Makefile rules to do that.
That sounds pretty good. I'll look into it.
It keeps the module, its tests and examples nicely in the same source
structure (= code repository), which is good as they are usually
developed in parallel.
At the same time it completely modularizes the source code and
de-clutters the generated libraries, resources, and binaries, making it
very obvious which belong to which part.
The base rules for this are new and basic. If you find anything not
working or not working as expected ... happy to improve the mechanism.
Cheers,
~Ralph
- Replies:
- RE: FW: EPICS 3.15.1 Mark Rivers
- RE: FW: EPICS 3.15.1 Mooney, Tim M.
- References:
- EPICS 3.15.1 Dudley, David
- FW: EPICS 3.15.1 Dudley, David
- RE: FW: EPICS 3.15.1 Dudley, David
- Re: FW: EPICS 3.15.1 Michael Davidsaver
- RE: FW: EPICS 3.15.1 Mooney, Tim M.
- Re: FW: EPICS 3.15.1 Ralph Lange
- RE: FW: EPICS 3.15.1 Mooney, Tim M.
- Navigate by Date:
- Prev:
RE: connecting EPICS to a Beckhoff PLC ronaldo.mercado
- Next:
RE: 3.15.1 DB Build Issue Dudley, David
- 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: FW: EPICS 3.15.1 Mooney, Tim M.
- Next:
RE: FW: EPICS 3.15.1 Mark Rivers
- 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
|
ANJ, 17 Dec 2015 |
·
Home
·
News
·
About
·
Base
·
Modules
·
Extensions
·
Distributions
·
Download
·
·
Search
·
EPICS V4
·
IRMIS
·
Talk
·
Bugs
·
Documents
·
Links
·
Licensing
·
|