dependencies cleanup

[falkenhawk]

Would you consider cleaning the project dependencies a bit?
https://bundlephobia.com/result?p=conveyor-mq@0.0.36

• all @types/ should go to devDependencies
• consider ditching lodash (I suppose all of the used helper functions can be easily refactored into pure typescript)
• maybe also find a way to reduce the dependency tree by replacing moment (it's actually the biggest dependency) with native Date and small helpers

[jasrusable]

Hey @falkenhawk 👋

Thanks for the great suggestions!

all @types/ should go to devDependencies

This has been merged now in #69

consider ditching lodash (I suppose all of the used helper functions can be easily refactored into pure typescript)

maybe also find a way to reduce the dependency tree by replacing moment (it's actually the biggest dependency) with native Date and small helpers

Both good ideas, I will add these to my backlog. moment has actually mostly been removed from the codebase, there are just one or two instances of it left before I can remove the dep entirely.

[jasrusable]

Released 0.0.37 with all typings moved to devDeps (As well as other dep updates).

[falkenhawk]

heads up: momentjs is dead done
https://momentjs.com/docs/#/-project-status/

[jasrusable]

@falkenhawk Thanks for the heads up on this, I saw the announcement yesterday and evaluated some alternatives and Ill most probably be switching the codebase over to date-fns or luxon in the near future.

[dougg0k]

A lightweight option https://day.js.org/ 2kb and immutable. For it's utility, it's worth.

If the project is not dead, is it?

[dougg0k]

I have opened a PR with all these changes and more. But the project seems to be abandoned. I came here from a comment in bee-queue, which just seems the best option out there up to this point. But is also no maintained anymore, and I prefer to not go back to using other projects like bull.

It would be cool if this were still being maintained and the roadmap going forward, it would probably be the best option available that does not make use of MQ brokers, if not already.

Project are just not that known. Otherwise it would get more interest from others for sure.

If there were benchmarks + feature comparison with other options. It would be a plus too.

Accordingly to BullMQ https://github.com/taskforcesh/bullmq#feature-comparison, their pro version is the most complete one, though bee-queue might be the fastest currently. If there were one option with all the features + faster than them. It would be the go to option.

[jasrusable]

Tracking these suggestions ^ in separate issues now: #168 and #167.