Nice-to-have Current features.

[dimacurrentai]

This issue is mostly to track some brave ideas of early 2024 that might be worth investing into.

Nothing even remotely resembling planning. I just want to have several things laid out, in case we do revisit Current development.

I have quite a few things in mind.

• Logging.
  • We need some basic logger.
  • Singleton.
  • A factory, for custom log types / files.
  • With build-in rotation.
• Better CORS support for HTTP APIs.
  • Some HTTP headers are custom-ish.
  • Probably best to configure per HTTP port.
  • Also, we need some RespondWithJSON into Request, so that the argument is a std::string, but all the return headers are set as if it is a value JSON, expected to be parsed as such.
• "Current as a Service".
  • Once logging is in place, this is effectively about the "graceful restart".
  • Unified health checks and telemetry, of course.
  • cron-friendly.
  • First that the currently running binary on the port the "new" one is expected to take.
  • If "self" is up and running, do nothing.
  • If not, signal it to shut down gracefully, then replace it.
  • Quite a common pattern, we've built this before.
  • May even leverage our nginx integration.
• Dynamically loaded .so-s.
  • A nice addition to the "Current Service".
  • If new code was added, and the new .so-s were successfully built, have the currently-up-and-running service pick them up.
  • In-code support, via smart pointers or Owned/Borrowed, so that some pieces of code get updated "on the fly", while the "main logic" is intact.
• Protobuf <=> CURRENT_STRUCT integration, as well we gRPC <=> Current integration. Ideal end state:
  • Build with Current.
  • Everything is exposed via HTTP.
  • With just a few macros here and there, those HTTP endpoints are also high-throughput low-latency gRPC endpoints.
  • That's for the server. For the client ... well, we'll need a client too, Current-friendly, as well as tests and benchmarks.
  • Last but not least: if we have HTTP and gRPC, might as well have binary TCP sockets, and compare all three!

Not sure how much time I will have in the coming months, but I'd like to have this written down somewhere, and a GitHub issue sounds good. We'll close/resolve it as soon as any real work begins, and/or if we choose to re-prioritize. Also, nothing prevents us from editing this message, although I'd prefer to keep it as is due to my personal preference of tracking things fairly.

Last but not least: the Educational Designs section of the SysDesign Meetup README page outlines a few principles which Current may well follow, since, who knows, some of those designs I keep planning to materialize could be Current-first.

Thanks,
Dima

[dimacurrentai]

@mzhurovich WDYT?