Jim , don't feel too bad, I guess the fact that english is not my primary language helps me get past the inconsitencies in the manuals (I just don't understand what they really say there (bg)) IAC it is slower because it creates a cursor - wheter it will be actually written to disk or not , that is another q. as you've found out Arnon > >That is as you and others have pointed out. >I gues I sometimes (often?) simply read things differently, placing the >emphasis on the wrong thing. >In this case I read the warning about slowing things down because it will >create a file to disk as the "main point' regarding the sub-clause. In >other words I read it as 'you will get a useable cursor BUT things will go >slower because there *WILL* be a table made to disk'. > >I guess I need a remedial reading course in addition to all my other >shortcomings! > >Cheers, >Jim N > >>Jim >>the whole point of NOFILTER is to garentee a cursor (rather than a >filter) >>wheter it will reside on disk or in the memory is an internal descision >by >>VFP >> >> >>Arnon >> >>>That's an interesting point of view, especially when the *WHOLE POINT* >of >>>NOFILTER appears to be to explicitly cause wiritng to disk. >>> >>>regards, >>>Jim N >>> >>>>1) This is more a bug in the documentation. It should not explicitly >>>>state that a file is created on disk. >>>> >>>>2) A Solution (well, a workaround) >>>> >>>>CREATE CURSOR {whatever} >>>>SELECT INTO ARRAY >>>>INSERT INTO {whatever} FROM ARRAY >>>> >>>>> Yeah, I'm classifying it as a bug too. >>>>> >>> >>> > >