I received a personal response to this concerning the ability to still
use v1 API calls for the missing functionality. And that I was making
this seem like a big deal when in fact it isn't because Tumblr have
not marked v1 as obselete.

If I were maintaining a website I'd have no problem with this as I
would have immediate access to the raw code should I require any
changes. To maintain two API codebase for a deplplyed app requres I
maintain, support and debug two request/response mechanisms, two
serialization/deserialization mechanisms, two error handling
mechanisms and two sets of tests. With the kind of speed available to
mobile app developers at the moment in terms of version release
through an app store it's also a problem that every time the API does
get closer to feature complete I'm going to be releasing new updates
with no perceivable difference to the end user, but vastly improved
stablity. Plus the more code that is in one API allows any performance
increase or robustness in that codebase to be applied quickly across
the board.

Doubling the risk of failure to the app, or possibly worst, making the
app to appear at fault because some of the api calls are working and
some not, is a risk for an app maker, and not one I'm happy having to
make due to a new API release not being feature complete.

If being freature complete is not expected, this should not have been
marked as new in the engineering blog and it certainly shouldn't have
been marked as the de facto API in the documentation. Developers
expect like for like functionality with full backwards compatibility
or at least a list of breaking changes. I don't believe that's
unacceptable. I don't allow the code I release to do this in my
professional or personal endevours and I hold others to the same
level.

If I'm seen as overly vocal or critical it's because as a user and a
developer I want Tumblr to be the harness for a lot of great
functionality and I really think the new API solves a lot of problems
I've had so far. I just think the risk is too great at the moment.