The Story of Why We Migrate to gRPC and How We Go About It - Matthias Grüter, Spotify

[@yoloswaggins2161]

Why didn't they open source hermes so an ecosystem could grow around it?

[@yoloswaggins2161]

@Ken Nguyen I don't think this is very plausible. No coder at spotify would hard-code credentials into their code if that code would ever be checked into a repo. Beyond that even if they did it's a problem that's so easily solved that I doubt it would be a blocking issue for long.

[@willinton06]

Yolo Swaggins you’ll be surprised at things engineers do when they feel safe

[@lordmushroom723]

@Ken Nguyen lmao troll

[@k42p3r]

Interesting story. It sounds like a great technology for Spotify to try out due to the drawbacks of Hermes and their need for high performance. That said, I don’t see gRPC as a competitor to existing REST APIs since they IMO have a higher level of compatability and are easier to work with. I’m looking forward to see how this hot technology will evolve the next few years.

[@jimmiedoe1114]

kzbin.info/www/bejne/iKC7hZKIoMxrgqM

[@jkuang]

Every time when we finally reach a CLEAN API SPEC LIKE REST, some stupid people would come up with something else claiming "it could solve some problems for you." Complicated API SPEC is the reason that plagued us for decades. Keep the API SPEC simple. Save yourself YET ANOTHER TRIP back to memory lane of complication!

[@blackstratmx]

Well... nothing is perfect, and also is how technology works, I use gRPC every day at work and It save us a lot of problems by the use of strong types. With a large setup of micro services you'll find errors with REST API's.

[@jkuang]

@@blackstratmx I heard this kind of talks many times over the years. At the end of the day, it gets so complicated and we just abandon and move forward with REST. And someone will hold onto this fringe solution "that promised to solve your problems" and suffer silently for the rest of their career at the job. Repeat the same tragic story all the time, always have a rosy beginning and always ends up in a dark room silently suffering for as long as the application is running.

[@blackstratmx]

@@jkuang Well I don't think you should suffer from your job, or stay on dark rooms, It's just a job. On the other hand is really bad to see one solution as the only one valid. REST serve its purpose, gRPC serves another one, being flexible is what makes this industry so great! Don't get in love with technologies or your code :-), get in love with the Money.

[@johanw2267]

Well programmers love to solve problems. So every tech you use is an iteration of an older tech in order so solve problems from the previous version. Does graphQL need to exist? It evolved from REST and some find it better, some do not. That is just how it goes in this industry. If things work you do not need to change it, but obviously for Spotify they hit a roadblock with their DIY solution and had to change. Use the right tool for the job.

[@jkuang]

@@johanw2267 In many ways, people came up with something with the vague pretend that they are solving some edge case problems. But it is just yet another attempt to re-invent and feel relevant. I think invention is really needed to clearly solve a real problem but sooner or later you will realize that in most cases, they just highlight "some" problems that could have be solved with an existing CLEAN SOLUTION, and chose to come up with something else.

[@timzeynalov3537]

so loooowww

[@markcuello5]

HELP

[@FudgeBondu]

Jackson Robert Robinson Deborah Jones Donna