I agree that the search interface deserves some reconditioning.
The search interface is the first entry point for many users. Thus, it
needs to be designed as simple as possible. I guess my grandma would be
completely overextended by the current one with its dozens of options.
I propose to completely forget thinking about namespaces. The vast
majority of users doesn't even know what namespaces are. The typical
user comes to Wikipedia and hits the search button, because he's just
interested in ns 0, so ns 0 should be the default.
On the other hand, some users, mostly our contributors, need other
search options. Although they have an understanding about namespaces,
they visit the search interface with the intention to find something
either on a talk page, an image description page, a project page or
something similar. Most of them do not intend to search the talk
namespace, the image namespace, or the Wikipedia namespace. It's
the search interface that forces people to transform their search
destinations into namespace names.
I admit that namespace names are quite self-explanatory, so we don't
have a serious problem here. Though, a good user interface doesn't force
users into an unnatural way of thinking. Namespaces are MediaWikis
technical way to devide pages up, they are not the way how ordinary
people think. The appropriate way of handling this would of course
be MediaWiki internally mapping common search options to predefined
sets of namespaces.
In conclusion, we should have a search interface containing three basic
- a big search box -- Google has a simple front page with a centralised
search box for a reason
- some common options, like those proposed by FT2:
On 31.10.2008 13:13:13, FT2 wrote:
* Articles and article discussion
* Articles, discussion, project space, and project talk
* Project and project talk spaces
* Images, categories, templates and their talk pages
* All namespaces
- and finally a "Go" button.
Leon Weber, leon(a)leonweber.de 0x8E04D7FC
jabber: leon(a)jabber.ccc.de (icq: 261067046)
Ein dreieckiges Rad ist gegenueber dem viereckigen Rad ein Fortschritt,
da es pro Umdrehung einen Hopser weniger gibt.