(.tsl -> .proto) = gRPC? #361
Replies: 3 comments 1 reply
-
There's one more thing on my mind. I have no idea how communication is implemented in GE, so excuse me if I'm missing the point. Would gRPC be a good candidate for client-server and server-server communication? If so, it could reduce maintenance costs by simply relying on gRPC for communication. Why reimplement security features if there are ready-made solutions? Time and effort saved due to reusing a communication protocol could be invested in core GE's functionality. |
Beta Was this translation helpful? Give feedback.
-
FYI - the GE Communications architecture, design, and implementation is written in C++ and works over low-level TCP and HTTP/RESTful. It is extensible so that you can implement gRPC. The OOB communications implementation is one of the best I've ever seen in over 30+ years of distributed client/server programming and cut my teeth on IBM DCE/RPC & CORBA back in the day. The GE communications model is lightweight, fast, efficient, and supports symmetric network computing. |
Beta Was this translation helpful? Give feedback.
-
Please forgive my newcomer's questions, I've only learned about GE's existence yesterday. I promise I'll learn about TSL and never underestimate it again 😉 |
Beta Was this translation helpful? Give feedback.
-
First of all, thanks for keeping this project alive. It seems like a fascinating piece of technology and it would be a shame if it just disappeared.
I'm currently getting to know GE and while looking at TSL it's hard not to draw similarities to Protocol Buffers. I'm wondering if it would be a good (and doable) idea to create an extension (think GraphEngine.Protobuf) to automatically translate
.tsl
into.proto
in order to take advantage of the awesome support of gRPC in .NET. It would make GE easily accessible for any gRPC-enabled client software together with any security-related shenanigans. I feel like this could help bring more attention to GE.There's also another, more drastic possibility, to simply drop TSL in favor of Protocol Buffers, which is a widely recognized and supported language for specifying data and operations contracts. What do you think? Is TSL worth keeping alive or is it going to be dead-weight? Maybe I'm missing something and there are some features of TSL impossible to implement in Protocol Buffers?
Beta Was this translation helpful? Give feedback.
All reactions