emacs-request VS emacs-dumb-japanese

That is quite normal thing to do. Have you not seen Emacs Async? Take, a look, it is a useful thing. Or Emacs Request. Since Emacs does not have proper thread scheduler, that is the best next thing you can do.

Exciting! I've been using request.el for my own projects mostly out of habit. Could you outline some of the relative advantages of plz?

Instead of shelling out to curl, use url.el or request.el.

You mean, John Wigley's async package? Maybe it isn't used so often, however async processes are used in Emacs. Check for example functions 'native-async-compile' or 'async-byte-compile-file'. There is another package, request.el that uses async processes to do the network I/O (via curl).

This? https://github.com/tkf/emacs-request

You can already use emacs-request, it offers very nice asynchronous API and uses curl by default if present and falls back on Emacs url if curl is not found.

Have you tried (package! emacs-request), since that’s the name of the source repository (https://github.com/tkf/emacs-request)?

It's possible that I'm brain dead but I'm struggling to wrap my head around the emacs-request code found here: https://github.com/tkf/emacs-request

I have been using plz.el to write Japanese. For dealing with a simple GET endpoint, it's more concise than url-http.

I am developing a Japanese input method that tracks user inputs as text properties. In theory, it is possible to extract romaji/hiragana from kanji transliterated with it, but the project has less priority for me.