>Hi All,
>
>Can anyone explain the following behaviour:
>
>Create a table (test1) and add some records
>Create an identical table (Copy structure to Test2)
>Select * From Test1 Into Cursor curTest1 NOFILTER
>
>weirdness number 1: The function DBF() shows a blablabla.tmp in the windows temp directory. The file is not there (explorer doesn't show it) while NOFILTER should create it?
>
>weirdness number 2: A Select out of DBF('curTest1') triggers a file blablabla.tmp not found error while an Append from DBF('curTest1') works fine?

blablabla.tmp is there despite explorer doesn't show it. To demonstrate it's there you could execute :

handle=fcreate(getenv("temp")+"blablabla.tmp")
? handle  && -1
? ferror() && Access denied

Before "nofilter" what we were doing is to add an extra nonexisting field expression or .t. to where :

select * from mytable where .t. into cursor myCursor
use dbf("myCursor") in 0 again alias myEditable
use in "myCursor"
select myEditable

Except .t. usage in where same goes on with nofilter. Now there is a true cursor that you can use in subsequent operations or make it editable like above.
In second "weirdness" you would directly select from curTest1 not dbf("curtest1").
PS: Also file("blablabla.tmp") returns .f. but do not thrust it (as it also returns .t. for a nonexistant remote connection file after timeout period).
Cetin