Open SperkLabs opened 11 years ago
I also support the above proposal. I am currently using controllers that uses propriety format, in which same PGN is used for multiple messages. The messages are differentiated by the priority and sender. In addition to signals within a message can be further identified based upon the byte1 and byte2 values.
I tried implementing the J1939 based database, but it classify multiple messages as one type, even when they have different priority and command bytes.
I would rate this major improvement since it is important if you work with dynamic node addresses that are part of the CAN ID. Please have add this feature.
Some protocols have fields in either the CAN ID or the message data that determine message type. Also, some protocols have fields in the ID that do not determine message type, such as priorities and addressing.
For J1939, part of this is handled by using a J1939 database. However, some proprietary J1939 messages use the first 1 or 2 bytes to further specify the message type, and then different message types need to be parsed differently.
I propose that for databases, masks (care / don't care) can be set up for both the message ID and portions of the data.