You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
At RelationalAI, @Drvi and I have undertaken doing a native-julia protoc implementation (called protojl). It parses proto files and can generate the appropriate modules/structs/serialize/deserialize for full protobuf integration. It solves some of the long-standing issues people have had with protobuf in Julia: weight of the protoc binary dependency on runtime applications, performance/allocations of serialize/deserialize, and improves coverage over the entire protobuf format specification.
We've started using it internally at RAI, but our goal since the beginning was to contribute back to the community, so robustness of the code can be accelerated and everyone can benefit. In lieu of starting a 2nd package, we'd rather just overhaul this existing package, as long as current maintainers agree. We're happy to assume maintainership of the new code going forward and are eager to support other applications/industry use.
If there are major concerns, let's discuss here; we're planing on putting up a PR sometime next week.
At RelationalAI, @Drvi and I have undertaken doing a native-julia protoc implementation (called
protojl
). It parses proto files and can generate the appropriate modules/structs/serialize/deserialize for full protobuf integration. It solves some of the long-standing issues people have had with protobuf in Julia: weight of the protoc binary dependency on runtime applications, performance/allocations of serialize/deserialize, and improves coverage over the entire protobuf format specification.We've started using it internally at RAI, but our goal since the beginning was to contribute back to the community, so robustness of the code can be accelerated and everyone can benefit. In lieu of starting a 2nd package, we'd rather just overhaul this existing package, as long as current maintainers agree. We're happy to assume maintainership of the new code going forward and are eager to support other applications/industry use.
If there are major concerns, let's discuss here; we're planing on putting up a PR sometime next week.
cc: @tanmaykm @aviks
The text was updated successfully, but these errors were encountered: