I know I'm new here, so my opinion won't count much, but I too feel that non-wildcard matching should be the default option. Most people, nowdays, have a general idea of globbing (if not RE), so they'll know what a * does. Plus, it would follow a similar syntax to the bourne shell. Consistency is important!
- Vishesh Handa On Wed, Mar 10, 2010 at 9:07 PM, todd rme <[email protected]> wrote: > On Wed, Mar 10, 2010 at 10:11 AM, Sebastian Trueg <[email protected]> wrote: > > Currently the Nepomuk query API will add a wildcard '*' to each full > > text search term so terms like "Nepo" will also match "Nepomuk". > > Recently some concern has been raised[1] about this practice and I tend > > to agree with it. > > Basically the competing idea is to not add a wildcard automatically but > > let the user specify it like so: > > > > "nepo*" > > > > The advantage is that 1. we would support both wildcard matching and > > non-wildcard matching and 2. simple full-text queries would be speed up > > a little bit. The disadvantage is that it would be a behavioral change > > in the API (the old behavior is not documented though). > > > > So I need some opinions here. > > I think it is essential to support non-wildcard matching. I am not > clear on which should be the default, but in the absence of suggestion > of a good way to force non-wildcard matching I think that this is > really the only option. > > -Todd > > >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to > unsubscribe << >
_______________________________________________ Nepomuk mailing list [email protected] https://mail.kde.org/mailman/listinfo/nepomuk
