|
Key
This line was removed.
This word was removed. This word was added.
This line was added.
|
Changes (1)
View Page History... |
| The bits in the field prefix indicate the ordinal is included, with no name, and String data described by 1 width byte prefix. Although the fields are given in numerical order in this example, the encoding specification allows for arbitrary ordering. Taxonomies must only contain field names that are less than 256 characters in length, so will always use a 1 byte length encoding. Ordinal values that are not defined by the taxonomy, or that would map to the empty string must not be encoded. |
| Note that there are no special indicators to distinguish a message in this form from any other message or require that it be processed any differently to other messages as part of the encoding system. Nor does the presence of such a message imply any taxonomy details of subsequent or preceding messages in a stream. It is the responsibility of a decoding application to interpret the message according to any [serialisation conventions|Serialization Proposal] Framework] it is sharing with the message origin which may support the dynamic update of a taxonomy library using messages encoded in this manner. |
| Note also that other encodings are possible but would rely on the list, array, or map serialisation conventions which imply a fixed ordering of data within the message. The specification does not require the ordering of uniquely named or numbered fields to be preserved within a message allowing applications more flexibility in that they can select appropriate arbitrary orderings that might allow more efficient algorithms to be used. |