As we've continued with our beta (http://bit.ly/w4E2zn) its becoming abundantly clear from feedback and our own research that we need to get better about search. Search is the primary way that people start their interaction with Wikipedia and we currently do it poor at best on mobile.
We've taken the first stabs at it
* Change the search input box to stretch to full width (horizontal & landscape) * Lower the amount of search results and increase the font size of what we do show * Include spaces in between results to better segment each item * Add a "+" to refine search term by term (our search data set will need to get better to make this more awesome)
Thats already way better then what we've had in the past but we still need to do more.
We need to make some choices about where we take search next. One current problem is that were not really using our screen real estate as effectively as we could on mobile. Whenever I see users using search their consistently moving forward. Their not interacting with the content thats on the article that their on and instead they want something new. If thats the basic case then why should we be wastin screen real estate on a tiny device that no on is really using. This is a basic fact of mobile that you need to focus the user on what's happening not distracting the with noise.
I took a look at a lot of mobile sites and one stuck out to me as really interesting for search.
bing.com (yes i know its microsoft .. lets pay attention to the design rather then the ideology)
Go to their site on a mobile (on a mobile) and start typing in text. Notice how its now stolen your whole screen for search? Its loud but thats the point. People are searching and moving forward. Not interacting with the content thats on the screen.
Other good examples
http://www.google.com/m (clean) http://m.yelp.com/ (really good use of images)
There are other design hurdles to tackle but I think we can do a lot of good by just tweaking search before we do a big UI redesign.
Lets discuss!
--tomasz