EmberData BETA vs EPF... Any experiences to share

So I started working on a project and I have been stubbing out with the EmberData beta (and EmberAppKit) and was interested before I start on the API whether I should reconsider EPF for the Client Side persistence layer.

Does anyone have any thoughts on the comparison between the two?

Does anyone have a real world example using EPF? I have looked at epf-example but would like to see more things using EPF.

I’m a bit biased here since I’ve worked on Ember Data, but the Ember Data project is backed by the core team and will be recommended as the default going forward. So Ember Data is my choice. However, we do intend to support other data frameworks so if you have a strong reason for using EPF, that’s still an option.

I moved to EPF before ember-data beta was released to facilitate live editing (saving happens transparently as the user does stuff in the app). This was proving difficult to do with ember data, and was fairly easy with EPF using the idea of child sessions that autuomatically merge the version of a model that the server returns from a PUT with the current version of the model in the parent session. EPF has worked well for me, and was a much better experience than ember-data pre-beta. I did have to patch it to add more flexibility for custom model endpoints; hopefully those pull requests eventually get accepted.

One current disappointment is that it can’t be used with ember-validations out of the box – but I think that is also still true of ember-data.