Experimental Physics and
Industrial Control System

Michael Davidsaver via Tech-talk <tech-talk at aps.anl.gov> · Wed, 15 Dec 2021 09:52:10 -0800

On 12/15/21 4:35 AM, Ralph Lange via Tech-talk wrote:
> On Wed, 15 Dec 2021 at 07:52, Alfio Rizzo via Tech-talk <tech-talk at aps.anl.gov <mailto:tech-talk at aps.anl.gov>> wrote:
> 
>     So my conclusion is that giving this limitation of 256 chars of maclib parser,
>     It is wise then to never use EPICS_DB_INCLUDE_PATH in a snippet to add
>     local db path, since if you have a long list of require you may hit into trouble.
> 
>  
> So, indirectly, this is related to ESS's E3 development environment that keeps all dependency modules in separate locations (with long paths).

It might be reasonable feature for E3 to assemble $EPICS_DB_INCLUDE_PATH
at the same place (I think) it assembles $LD_LIBRARY_PATH.

> Using other environments and/or deployment mechanisms that don't allow parallel use of multiple versions of modules, shared things get installed into a single location, so your EPICS_DB_INCLUDE_PATH stays as short as "$(TOP)/db:$(SHARED)/db".

I sometimes begin st.cmd with a variation on the following
to avoid a need for st.cmd to change directory.

> epicsEnvSet("EPICS_DB_INCLUDE_PATH", "$(TOP)/db")

or

> epicsEnvSet("EPICS_DB_INCLUDE_PATH", "$(PWD)/../../db")

or

> epicsEnvSet("EPICS_DB_INCLUDE_PATH", ".:../../db")

(I sometimes put instance specific .db files under "iocBoot/*/" )

And I also prefer to make use of DB_INSTALLS+= in my App/db/Makefile [1]
to copy in external db files and fragments at build time, so that the
resulting IOC (if statically linked) is self-contained.

(I mentions this because both $EPICS_DB_INCLUDE_PATH and DB_INSTALLS
seem to be fairly obscure)

[1] https://github.com/mdavidsaver/pyDevSup/blob/ea5f0d62412085dfc941e8fe758695abdfdab548/devsupApp/Db/Makefile#L17-L23

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System