Adam Eberlin

@ajb Please restart this job when you get a chance, issue doesn't look related to the changes. https://travis-ci.org/chrismccord/sync/jobs/108574596

@ajb, With regards to the coffee initialization process, I think I can simplify it by requiring that sync dependencies be loaded explicitly, without auto-loading themselves. If we can require a...

Overall reasoning was because of a race condition with dependencies I experienced locally. Not a huge deal, but it made me look at things a bit differently.

@ajb I still need to root out some issues with the `debug_flag` and `stompjs`, but do you see anything lingering still that needs to be addressed?

@ajb, The race condition was due to order of dependencies in sprockets. I took a stab last night at removing the auto-initialization and found that it was more complicated to...

I think that https://github.com/chrismccord/sync/pull/212 will actually make this easier to wrap minds around. If we refine the STOMP/websocket configuration vars and generalize those in Pusher, the bulk of the work...

This would definitely be a breaking change. I would suggest building out a small DSL for things to be passed to the sync render context, possibly by leveraging a library...

Without monkey patching or pestering the hell out of Pete, not really. I've overridden CSS to hide items, but all out altering them is a bit tricky.

Hey @brousalis, could you add me as a collaborator here? I would like to make some compatibility and documentation updates, and it's been pretty hard to reach you lately. Sorry...