First of all thanks @rdoeffinger for keeping this project alive!
I find that the round materialish buttons you recently introduced in the app are useless and actually hinder usability instead of improving it.
Useless because all their actions were already 0 or 1 clicks away in the old interface:
- Searching is the default action when you enter a dictionary. The cursor is automatically in the search box and keyboard is opened. Thus we need zero clicks to get to it. Even if we clicked somewhere else, searching again required only 1 click in the search box, which is a big target that you can easily hit with your fingers.
- Clearing the current search word took 1 click in the
X
button to the right of the search box. Again this is a big target that no one misses.
- Switching languages also took a single click in the flag, again a big target. This one was somewhat hidden. It is hard for a new user to guess that clicking the flag changes languages. The straightforward solution would be to place (watermark) the two arrows symbol on top of the flag.
Summing up: these new buttons solve no usability issue.
Now, even worse, is the fact that they actually are a usability hindrance, because they hide important text underneath. Checkout what happens in my device:
![quickdic-material-buttons](https://cloud.githubusercontent.com/assets/1716156/19929022/2edb302a-a0f9-11e6-853d-371e41c666b1.png)
90% of the time we use Quickdic with the keyboard opened. These new buttons take a lot of the precious little space we have left. With them, I am forced to hit the back button to hide the keyboard in order the read the translations. That's something I didn't had to do before. Thus they actually increased the number of clicks for me.
Taking this into consideration, I would urge you to remove these buttons, or at least add a preference's option to do so.
Tips for future UI improvements
The original UI by Thad Hughes is actually really good. All major actions are easy to do with it. I would exercise caution before adding modifications.
I see that you are in the mood to update it to a full blown Material UI. If so, please consider starting such work in a new branch. Besides avoiding breaking the UI on the production app, a new branch would allow you to start from scratch. It is rarely a good move to work iteratively on top of an old UI design. Better start anew and see whether applying Material solutions to this problem renders a better UI than the original Holo theme.