Hi Jason, On Sun, Jan 1, 2017 at 4:44 PM, Jason E. Aten <j.e.a...@gmail.com> wrote:
> On Monday, December 26, 2016 at 4:18:45 PM UTC-5, Jason E. Aten wrote: >> >> Announcing: next generation serialization in Go: >> >> https://github.com/glycerine/zebrapack >> > > Update: In addition to being friendly to many languages, by the latest > measurements ZebraPack is the fastest serialization format available, > beating out all competitors. See the benchmarks on the page above. > that's quite impressive! do you have benchmarks against colfer? it is the fastest library, according to: https://github.com/alecthomas/go_serialization_benchmarks#results I used the same strategy than colfer for my WIP binary I/O library: https://github.com/go-hep/brio (except that, as you, the schema is in Go, not some external DSL like protobuf) being able to transfer on the wire, store and load types' schemas is really great. do you think the schema part could be extracted as a standalone package, to mimic what "dtype" from numpy python module managed to do (a type/schema universal interface to describe types and their binary layouts, b/w packages and programming languages) ? -s -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
