![]() |
![]() ![]()
Experimental Physics and
| ||||||||||||||
|
Hi Mark,
On Tue, Jun 5, 2018 at 3:42 PM Mark Rivers <[email protected]> wrote:
Not too much, actually. Well, clarity, of course. They are dedicated string records, so their value is an array of char (waveform has configurable basic field type). They are dedicated input or output records, so they have an INP or OUT link (waveform uses the INP link for both directions). In terms of processing and Device Support connection, they pretty much behave like a waveform record of char basic type. - If I do caget on an lsi record that contains a long string do I need to use the -S option as I do for a waveform record holding a long string? Unfortunately, yes. For Channel Access, it is still an array of char, and a generic CA client like 'caget' needs a hint on how to interpret that array. Specific clients that have knowledge about the EPICS database could ask for the record type (using DBR_CLASS_NAME) and take an lsi/lso reply as a hint to display that array of char as a string. - Should support for the lsi and lso records be added to devAsynOctet? Probably. Personally I haven't used lsi/lso, but they do seem appropriate for long strings. Cheers, ~Ralph ps. While we're at it: in 3.16 the mbbiDirect/mbboDirect records will be 32 bits wide. Will that require changes in ASYN?
| ||||||||||||||
ANJ, 06 Jun 2018 |
![]() · Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing · |