Hey, nice work. We're adopting publify for a fairly significant project and one thing I'm personally having difficulty with is our approach to upstreaming changes. The idea that you take a snapshot of publify at a particular point in time and make your modifications as you see fit.
Occasionally you'll need to make a modification that doesn't sit well with the 'theme' way of thinking and the tendency is to try and shoe-horn something in where it doesn't make sense.
I think if publify was self-contained within an engine it would be easier to install/maintain. Any thoughts?
Making Publify an engine would be great. Publify as it is distributed today would then simply become a ready-to-run installation of that engine.
That said, we don't have any concrete plans to do this, but it's definitely on our (at least my) mind.
If you have any guidance to offer on how to get this done, that would be awesome.
I feel the same way and I think that it's the best approach for starting complex projects.
In my experience, creating an engine from a previous app, it's harder than it seems, so @mvz, I would start extracting some smaller pieces into new reusable projects, like the trackback system.
I'm actually going the other route: I'm mass-extracting all of Publify into an engine in the same repository as a first step.
The extraction to several engines has been implemented in master. Next step is to actually have gems.
We have gems!