This is very similar to Akshell (http://toughbyte.com/akshell/) which we worked on more than six years ago with @korenyushkin and that made the front page of HN on a number of occasions.
Despite being promoted by Fog Creek, I suspect that Gomix will run into the same problem that we did: professional developers won't use it as they already have the tools they prefer, whereas hobbyists won't be willing to pay. Without a viable business model, a service like this won't grow.
PS. If you're looking for something similar, but simpler and geared towards backend apps and micro services (think AWS Lambda), you should check out https://unitcluster.com/
Don't people who are starting out find these systems to be too difficult? You have to know much more to use this, then the "view source" days: Http codes/headers, javascript, npm?, css... Those weren't things you dealt with when changing Html on your local browser. I'm sure I don't have a good perspective on what new coders can handle, but this seems like too much.
Good point. AppJet and Akshell had their own built in libraries and guides, whereas with Gomix you're expected to already know about Node, NPM and figure out which libraries to use yourself. This is likely to make this too complex for people just starting out, making the model even less viable.
On a somewhat related note, one thing I'd like to see is cheap Heroku like hosting for Node. Most Node web apps could get by with 64MB of memory and processes could be started quickly in response to incoming requests, making it possible to price this at 1 USD per app per month.
I think they are relying on the instant forking of a working app to save users. The idea being that even if you don't understand all the concepts, most people can muddle through and modify and enhance based on seeing what's going on inside an already working code base.
It's true that starting to code from a working example is a tremendous advantage. However it's also true you have to eventually grasp the concepts and node has never been the easiest backend to grasp.
Despite being promoted by Fog Creek, I suspect that Gomix will run into the same problem that we did: professional developers won't use it as they already have the tools they prefer, whereas hobbyists won't be willing to pay. Without a viable business model, a service like this won't grow.
I hope I'm wrong though! AppJet, a similar service from a YC team (https://en.wikipedia.org/wiki/AppJet) was the reason I got into JavaScript.
PS. If you're looking for something similar, but simpler and geared towards backend apps and micro services (think AWS Lambda), you should check out https://unitcluster.com/