I think it’s great to see what kinds of things you need to put together to get something like Meteor. It would be interesting to see a more in-depth comparison - are the source maps good? does it work on Windows? does it work in IE9? does it have meteor shell?
There have been a lot of resolved bugs and PRs in Meteor for nitpicky problems that we don’t even remember we solved. That’s the benefit of using something that’s 3 years old and actively maintained by a dedicated team.
This project clearly has some advantages - code splitting, newer dependencies, etc. We’re on the path to making those improvements this year.
Curious if someone tries using this, let me know how it is - I’m pretty busy these days so don’t have time to try it.
It’s easy to follow the beaten path and forget about the edge cases you guys have covered to a T.
Edit: Also, look how insanely popular that thread is. Can you imagine how Meteor will explode when 1.3 lands with proper npm support and graphql? Meteor will become the defacto way to build web applications.
This is just another proof of the sadomasochistic nature of the average javascript developer, who rather inflicts the pain upon himself of using 10+ different projects thrown together than using 1 coherent system.
I wanted to add that this project’s creator originally pushed this out before the 1.3 features were even announced.
Back when Webpack and HMR was still a brand new thing for most Meteor devs, this stack provided features that would significantly enhance developer experience.
Since then, 1.3 has been announced as well as Mantra and the overall direction moving towards GraphQL, Redux and the Facebook stack. It’s unfortunate that this project didn’t get as much attention back then (maybe because it was the holiday season), but as of now, it is less compelling in light of the new developments.
The guy did a lot of analysis and work. I imagine how cool it would be for Meteor community if @mattkrick would add all these goodies as Atmosphere.js packages with Meteor core becoming more flexible and allowing for such choices and plugins over time. Meteor is already very attractive with choice between ‘bare’ Meteor or Space/Mantra, Blaze or React/Angular. More individual choices would be better for everyone.
And, actually, now I understand why all these cool things were possible. Meteor is constrained by need to compile everything into monolithic build just because there is no other way to deploy it to iOS or Android. So, SaaS part is cut by simplicity of Play/AppStore, whereas in reality it is more flexible by default.