Hi, Thanks for your answer rupak but I tried your fix and it did not give the expected result. On the contrary, now I think that the plug in is no longer using cache, even for non-empty results...
On 31 août, 13:58, rupak mandal <rupakn...@gmail.com> wrote: > Hi , > you have to make a little changes in autocomplete.js > > Search for "*if (data && data.length*)" condition which is inside "request" > function. Add this code before the *if * condition > > if(data) > data.length=0; > > Thanks > Rupak > > On Mon, Aug 31, 2009 at 3:36 PM, mathieuuu <mathie...@gmail.com> wrote: > > > Hello, > > > I am currently using the jquery autocomplete ($Id: > > jquery.autocomplete.js 5785 2008-07-12 10:37:33Z) with remote data. > > > Obviously I did not disabled the cache and "matchSubset" option (all > > matches of "foot" are a subset of all matches for "foo"), which > > reduces the number of queries made to the server. > > > This is working well except in one case : if "foo" returns an empty > > result set, continue by typing "foot" will still generate another > > query to the server ... > > From my point of view this shouldn't be useful (if "foo" doesn't > > return results, "foot" won't either) and I am looking for a way to > > have the same "cache usage" with empty result as with not-empty > > results. > > > Does anyone have an idea of how this could be done or knows the reason > > why there is this difference of behaviour ? > > > Thank you a lot for your help.
