It's really complicated to have more than one search box, but we currently do that for two reasons:
One search is performed against work metadata, or collections, while the other is performed against pages based on text
Search results for works or collections are totally different from search results for pages, which include syntax highlighting and text snippets.
I've never figured out how the UI for combined search results could work, but James Creel from TAMU just suggested a "bento box" approach to me -- segmenting the SERP into a box listing works or collections, and another box listing full-text results.
I's not ideal, but it's the first approach I've ever heard which might work.
@anabeabas would you research this idea of a "bento box" search results page, including any examples you can find or articles about the UI of complicated search results?
It's really complicated to have more than one search box, but we currently do that for two reasons:
I've never figured out how the UI for combined search results could work, but James Creel from TAMU just suggested a "bento box" approach to me -- segmenting the SERP into a box listing works or collections, and another box listing full-text results.
I's not ideal, but it's the first approach I've ever heard which might work.