BFlat is intended to provide performance similar to FIX or NVFIX for binary data. It's a simple, non-hierarchical binary message format. BFlat is designed to be compact, self-describing, and very efficient to serialize and parse.
BFlat provides efficient, built-in support for arrays and direct serialization of numeric types. (In fact, one of the most common uses of BFlat is to efficiently serialize and deserialize large arrays of longs or doubles without changing the representation or precision).
BFlat uses a simple tag/value format that is equivalent to FIX or NVFIX. However, BFlat does not require restrictions on the character set for tags or values. Data values in BFlat are always strongly typed, which speeds processing and reduces or eliminates loss of precision when transmitting data.
AMPS fully supports BFlat, and libraries for application-side BFlat processing are available through an MIT-style license from the BFlat project site at http://bflat.io.
keywords: bflat, message types, binary, numeric types