Experimental Physics and
Industrial Control System

Marty Kraimer <[email protected]> · Wed, 15 Aug 2001 09:26:56 -0500

Brian McAllister wrote:
> 
> >>> On 8/15/2001 at 8:49:13 CDT, Marty Kraimer wrote:
> 
>   > I am not aware of any plans to change the current mappings which are:
> 
>   > epicsShareDef unsigned short dbDBRnewToDBRold[newDBR_ENUM+1] = {
>   >       0, /*DBR_STRING to DBR_STRING*/
>   >       4, /*DBR_CHAR to DBR_CHAR*/
>   >       4, /*DBR_UCHAR to DBR_CHAR*/
>   >       1, /*DBR_SHORT to DBR_SHORT*/
>   >       5, /*DBR_USHORT to DBR_LONG*/
>   >       5, /*DBR_LONG to DBR_LONG*/
>   >       6, /*DBR_ULONG to DBR_DOUBLE*/
>   >       2, /*DBR_FLOAT to DBR_FLOAT*/
>   >       6, /*DBR_DOUBLE to DBR_DOUBLE*/
>   >       3, /*DBR_ENUM to DBR_ENUM*/
>   > };
> 
> Why is USHORT promoted to LONG, but UCHAR mapped to CHAR (rather than
> SHORT) ?

These decisions were made many years ago so the following is a best guess at the
thinking.

Except for UCHAR => CHAR the other mappings preserve precision, i.e. all values
from the source can be represented in the destination. As I recall the exception
was made because it was thought that 8 bit waveform digitizers would be widely
used. Thus it was not a good idea to map UCHAR to SHORT because of the increased
memory required.

Again in hindsight it would have been better to use uniform semantics but I
think it is too late to change.

Marty Kraimer

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System