General comments

• The SETDAT protocol is transmitted as the data of an ACNET packet.  SETDAT protocol structures use the ACNET network representation.
• The Front-end process which services SETDAT requests connects to ACNET with the rad-50 handle SETDAT .
• In the description of the packet formats below all fields in the packets abut.

SETDAT request list packet

 

Length(bytes)	Field Description	Type
2	Flags	2-byte unsigned integer
2	Number of entries in request list	2-byte unsigned integer
not fixed	Request structure 1	structure see below
---	---	---
not fixed	Request structure N	structure see below

 
• Currently there is only one flag defined, DB.   DB is the least-significant bit and indicates whether the settings are to be forwarded to SYBSET.  The settings are forwarded if DB is set.  SYBSET goes by the ACNET rad-50 handle "DBM...".   The front-end process which does the forwarding goes by the ACNET rad-50 handle "WRITER".   The protocol used for echoing settings is documented  elsewhere .

Request structure

 

Length(bytes)	Field Description	Type
4	Property-device index pair	4-byte unsigned integer
8	SSDN	array of 8 1-byte unsigned integers
2	Number of bytes requested from property data (ILEN)	2-byte unsigned integer
2	Offset in bytes to requested property data (usually)	2-byte unsigned integer
ILEN + ILEN mod 2	Setting data	unknown to SETDAT

 
• The property index in a property-device index pair is placed in the most-significant byte of the integer.
• The device index in a property-device index pair is contained in the least-significant 20 bits.  The most-significant 4 bits may be used for flags of various sorts.
• Please refer to the  Property Notes  for information about property indices and there meanings.
• The ILEN(n) mod 2 term in the length of the data insures that everything is word aligned.  The data in the packet should be left-justified, i.e., should be in the bytes with the lowest offsets from the start of the packet.   However, since the data in the packet is in network representation the data will not be left-justified on the wire.   On the wire the data will be word-wise byte-swapped, which causes among other things the padding byte to be the next to last byte on the wire.   Note that Charlie's Acnet implementation does not perform the word-wise byte-swap so the user's methods must use the network represenation justification when assigning values to the memory area passed to the method by MOOC.

SETDAT reply packet

 

Length(bytes)	Field Description	Type
2	Error status for request list entry 1	2-byte signed integer
---	---	---
2	Error status for request list entry N	2-byte signed integer

 
• The error status contains an error byte and a facility byte.  The facility byte is an unsigned 1-byte integer.  The error byte is a signed 1-byte integer.  The error byte is in the most-significant byte of the error status.   A zero error status indicates success.  A negative error indicates failure, a positive error indicates conditional success, e.g., a pending.   The facility indicates what returned the error.
• Design note 22 does not document the SETDAT reply, because SETDAT replies were added after the note was written.