"The technical winners (Panels, complex ODBC, entity fields…) were the opposite of what we were rooting for. I’d rather configure my apps with code rather than layers of complex and limitating UI."
I'm not sure what you mean by "complex ODBC" since ODBC is not used in Drupal. The opposite of what you were rooting for is for the end-user to have less power and the coder more power. Get this, Drupal is not taking that power away from you, it is giving some of it to the end-user. It is 2014 and there are still products out there that you have to modify using in-line code in order to get a simple feature such as hiding the author of a post. This is where Drupal gets it right, by letting the end user have more power, they get the site they want, not the site that a coder hands them. Don't like the way posts are displayed on the home page, then create a new view in a UI, not in a php while loop.
There are some beautiful things in Drupal and Drupal 8 is going to be even better. Drupal is an excellent example of software getting better by rebuilding, using best practices rather than just keeping cruft from 10 years ago.
Looking at the numbers from Google Trends, I'd argue that Drupal may have gone too far in its quest to make everything configurable through a UI. It makes, in a lot of cases, for very complex UX.
The bottom line for us was that the compromise Drupal was offering (you can semi-code in the UI, but still need to rely on some custom coding for some glue) wasn't serving us well anymore. The Developer Experience was getting worse at each iteration.
I also don't feel it was ultimately serving our customers; most features are overwhelming, and it usually took a lot of training on our end to make the end-user not too unhappy.
The transition from Drupal 6 to 7 was the end of Drupal being a useful scaffold for open ended development: meaning the prior versions offered useful features one could use and then continue forward and build something custom. But Drupal 7 is too encompassing, too database intensive, and simply too complicated. Drupal 8 will have inertia, but will feel like the Titanic for the majority of lessor skilled web site builders who can't develop themselves.
What I mean is that at this stage, the direction of the project is pretty much set, and there's not a lot of effective pull made in other directions. A lot of the folks who were bringing some balance have simply left.
There are indeed still tremendously smart and innovative folks in this community. Not a lot of people questioning what "the Drupal way" should be though.
I'm not sure what you mean by "complex ODBC" since ODBC is not used in Drupal. The opposite of what you were rooting for is for the end-user to have less power and the coder more power. Get this, Drupal is not taking that power away from you, it is giving some of it to the end-user. It is 2014 and there are still products out there that you have to modify using in-line code in order to get a simple feature such as hiding the author of a post. This is where Drupal gets it right, by letting the end user have more power, they get the site they want, not the site that a coder hands them. Don't like the way posts are displayed on the home page, then create a new view in a UI, not in a php while loop.
There are some beautiful things in Drupal and Drupal 8 is going to be even better. Drupal is an excellent example of software getting better by rebuilding, using best practices rather than just keeping cruft from 10 years ago.