Re: Fwd: Control System Data Access API

[Marty Kraimer <mrk@aps.anl.gov>]

Ned,

Please send such message to core-talk

Marty

nda@aps.anl.gov wrote:

> Yep. But a new ca feature might be "send me all commands that this aliased pv understands", which would include the fields of the device and strings it supports. I guess this would also be defined in the alias table.
>
> One could then have a *generic* device record that supported state oriented devices, and the generic field names (CMND1, CMND2, STAT1, STAT2) could be aliased to device specific names via Ron's scheme. One would not need a new device record for each device type (which is the nightmare of a device model). 
>
> This would provide alot of device capability that would fit well into the current record oriented architecture of EPICS.
>
> Simple, but functional.
>
>  Ned
>
>  
>
> > If I read this correctly it is just asking for the CA server to support 
> > aliases.
> >
> > Marty
> >
> > nda@aps.anl.gov wrote:
> >
> >    
> >
> > > Hmmm, a clever and simple idea.
> > >
> > >  Ned
> > >
> > >
> > > ------------------------------------------------------------------------
> > >
> > > Subject:
> > > Control System Data Access API
> > > From:
> > > "Chestnut, Ronald P." <ronc@slac.stanford.edu>
> > > Date:
> > > Wed, 19 Oct 2005 17:36:18 -0700
> > > To:
> > > gtkacik@princeton.edu
> > >
> > > To:
> > > gtkacik@princeton.edu
> > > CC:
> > > Matthias Clausen <Matthias.Clausen@desy.de>, "Dalesio, Leo `Bob`" 
> > > <dalesio@slac.stanford.edu>
> > >
> > >
> > > Gasper et al.
> > >
> > > There has been some talk of adding device orientation in a robust, complete, object-oriented(?) way. Enclosed is a more easily achievable approach.
> > >
> > > Ron Chestnut
> > >
> > >
> > >      
>
>