Closed jrburke closed 8 years ago
I've traced hubot and create a bot with some adapter and integrate with several plugins for test purpose: https://github.com/gasolin/huhu
But during the experiment, I got several critics:
As we planned some features that can't be fulfilled and not easy to patch the hubot without breaking its compatibility. Since the hubot core is compact and can be converted in ~1 week, we decide to port it to es6, learn more around chatbot framework then build new features.
What's in my mind to make improvement from hubot:
OK, it just seemed secondary to proving out webbybot. And while I am not a fan of coffeescript, the experiment I did was fine to just let that core part be in coffeescript but writing my bot specific code in the JS of my choice. It seemed more beneficial to help extend hubot directly, given its larger community vs creating something that may diverge at some point.
But I can see the technical interest in getting to know how the bot stuff worked on a lower level. Closing as not a bug ticket, just a discussion ticket.
thanks for rising the question, it still worth to add more references for record:
By the way it might be possible to load hubot .coffee modules through require.extensions
magic, so webbybot might still benefit hubot directly at its early stage.
We've ported hubot to ES6, compatible with all coffeescript plugins, and learned how chatbot works
related post here http://blog.gasolin.idv.tw/2016/03/how-we-ported-hubot-from-coffeescript.html
I saw a note somewhere about "porting Hubot from coffeescript to es6". What are the motivations for doing so vs just building on top of Hubot as-is?