Would be great if the features that go into engines/slices are abstracted from real apps though as suggested by Guoliang above. Refactoring platforms like Insoshi, Lovdbyless, Community Engine, Spree, Substruct would be one way to “eat your own dog food” if their maintainers are interested.

Your public API approach, is also appreciated a lot.

I would like to extend your futuristic vision with a little suggestion from my side, since mountable applications has much more to offer than what we can think right now.

Rails 3 should have at least one full fledged ” working website demo ” in near future, that shows all the advanced capabilities and features of Rails 3. It would be more helpful to have 2 working demos of Rails 3. (1) Default Way ( Rails Way and other (2) Custom way.( Merb way ) This would help and please Rubyists and Merbists as well.

It would be great to start with a simple working demo and let is grow more and more, along with additions which are parallel with your progress. This working site would work as an Inspiration, Tutorial as well as an Solution to common and frequent questions on “How and Why” unexplained elsewhere.

Let me thank once again for your great efforts, making Rails powerful, day by day.

I enjoyed the Russian Dolls slides as well, but I felt they were a bit of a tease! :P

I recall you saying a while ago that your primary motivation for checking out Merb back in the day was mailers. Well, mine was first class support for mountable apps through slices, so this talk was of particular interest to me. Definitely looking forward to API/DSL stabilization of this feature in Rails.

The overall mix of talks at the conference was right. I actually think looking at an unfinished product is quite interesting. If anything I would encourage more of these 90/10 “unicorn” presentations. It gives the community a chance to provide feedback (most of which will be constructive.)

I didn’t attend the Russian Doll session, but I’m looking forward to browsing the notes.

Great job and thanks for sharing the knowledge.

I have one worry though. I understand you guys are great and have a ton of experiences with building applications. However, when aiming to do the next big thing and implementing a big idea like mountable apps, if the implementation does not come from a widely accepted prototype or extracted from a successful app, it tends to become something either too generic or too rigid, something that solves no-one’s problem, and turn people away because of frustration.

I would suggest when you are building rails 3, try to eat your own dog’s food and create some practical app with it. If there are conflicts between the app’s need and what framework provides, bend the framework to make it work. If you do this, you’ll create a better framework.

This is just my 2 cents as a pessimist and I believe you guys have thought about and have a way to around the hole.

Thanks for making the slides available, particularly from those of us who weren’t able to attend.

Just a note, the Russian Dolls slides appear to have the speaker notes as ‘stickies’ which is great, but in some cases these appear to block useful content. In all but a few of these cases the note seems to be just a cue as to whether you or Carl talked about that slide sequence. Any chance of a minor edit?