I’ve promised to write more about persistence ignorance in the Entity Framework, and while waiting for builds during some late night coding binges for a deadline and various other distractions, I’ve been giving it a lot of thought. My conclusion, though, is that I’m not ready to treat this topic fairly. While I think I understand and have some useful things to say about how the Entity Framework does things (both in Orcas and in our vision for future releases), I don’t think I understand the viewpoint of the folks asking for full persistence ignorance well enough yet. As you may have heard in the blogs of some other folks like Scott Bellware, Jeffrey Palermo and David Laribee, we had some great discussions during the MVP summit about these issues (thanks for the feedback all of you who were there!).
As you might take away from the blog postings about those discussions, it was a good exchange of ideas, and I think folks left with a sense that the long term vision is not so far off as they preivously thought, but everyone agrees that Orcas won’t be a perfect fit for what some folks in the agile/domain-driven-design camp are looking for. While the amount of change that can happen between now and the release of Orcas is very limited, I want to work hard to understand this way of looking at the world deeply both so that we can get on track to converge in a later release and so that we can make certain nothing goes into Orcas that will prevent delivery of that vision later.
So, I’m taking my own little crash course on the subject which will by no means get me to the depth of understanding that some of you have after living and breathing this approach for multiple projects, but I’m hoping it will get me a bit farther along the curve before I respond on this subject and end up demonstrating my ignorance.