There's an interesting post by XJM that came up on Drupal Planet this morning. In it, she discusses the state of critical bug levels in D8. The good news is that the core devs are holding their ground now and hopefully, the count will start to gradually reduce.
But what I found interesting was a remark she made in passing:
I'm going to have to say "When it's ready"... and here's why. I'm not going to promise something that people are going to make business decisions around, and then be wrong about it. We think it's going to be coming out in 2014. Whether it's early 2014 or mid-2014 depends on how much help we get and things that we don't know yet.
When Cameron Weagans and Tim Plunkett visited a couple months ago and talked to us about D8, Tim mentioned that he thought it would release in the February time frame. I thought he was way too optimistic. I have been thinking that if they're really going to get the criticals to zero (and not just decide that a lot of them aren't actually critical), then it was going to be mid to late 2014 before D8 releases.
I'm not naturally a pessimistic person, but I have seen this before. Critical problems are hard to fix and take a lot of time to get right, probably a number of fix cycles to get all the edges worked out.
Anyway, to see a core maintainer saying out loud that this could stretch into the summer has validated my hunch for me. And that means that it will be well into 2015 before anyone has to really start thinking about building a project site in D8.
There's a constant tension between the enterprise level dev companies and we smaller minnows. For us, taking the time to learn a new release to get to the point where we're competent to use it is significant, and when that first big project comes up that requires the new release, we will spend a lot of time researching possibilities and learning the ropes. This is time when we are not being productive and getting paid. A big shop can afford to have someone go off in a corner and do this, and then return to train their team. It's much more efficient. We smaller fry would prefer to get more projects out of our existing D7 expertise, and our customers probably don't understand why they need to upgrade their perfectly adequate sites every few years.
I know that this is a part of the rationale for Backdrop. I personally disagree with the thinking behind Backdrop, but believe that they will meet with significant interest, at least in the short term. For me, each time Drupal has made significant changes, it has made the project a lot more flexible and useful. Stripping out layers of abstraction seems like a mistake to me, even if it does make sites perform better on shared hosting. As they used to say in the 90s, "What Intel giveth, Microsoft taketh away". Moore's Law hasn't yet been revoked.
It will be difficult for the D8 leaders to resist releasing D8 as early as they can justify it. But I think that the D7 experience should temper that. Releasing D7 with the entity system only half fleshed out was a mistake. Everyone still had to wait for Entity API and other such things anyway, and it would have been better if it had been complete and more of the core data objects had been fully converted to Entities (Files, for example).
IMHO, even with the perceived pressure to release, and the possibility of losing some of the community to Backdrop, the D8 leadership would be well advised to hold out for releasing D8 when it is really ready™, and not just when they can justify its readiness to themselves. We saw that with D7, where you couldn't really build a commercial quality site for about 9 months after release. D8 is promising a shorter time to usefulness with things like CTools and Views in core, but that won't happen if it's another system that is neither fish nor fowl. (Files, for example.)
We will still be here, and the features and internal advantages will still push Drupal ahead a great deal. And when it's really ready™, it will take off.