>>Cetin,
>>It's great speak with you. I've understood perfectly your reply even your 'contribution' 'with 2 cents'.
>>
>>Well, first of all:
>>
>>>alter table myNewCursor add column test c(10)
>>
>>Does this works? With the alias instead of the dbf name?
>>
>>I'll check it.
>>
>>And about the 'philosofical' question:
>>
>>what seems amazing for me is that file(dbf("mycursor")) !!! returns .f. !!!
>>No matter which is the name of the cursor and where he is located this function must returns .t.
>>
>>Isn't it? tell me something about this to calm my mind
>>Best regards,
>>Carlos
>Carlos,
>First of all these are my ideas that I believe to be true and never mean someone else is wrong.
>-alter table alias works.
>-I rely on "use again", "dbf()", "alter table" etc but not on file().
>As I said before I can show you file() returning .t. for a nonexistent file ! So should I expect it always return .t. for an existing one ?
>Here is a little test for you :
>

create table \temp\test (small l) && Smallest possible size I know
>select * from \temp\test into cursor mycursor nofilter
>cFileName = dbf("mycursor")
>? cFileName, file(cFileName) && file() returns .f. so assume it's not on disk
>handle=fcreate(cFileName)    && So let's create one
>? handle                     && Ooops. -1 Couldn't create - check why
>? ferror()                   && 5 - Access denied ??? Guess why

If you still think to use it (I do) use alias.
>Cetin

Cetin,
I've tested and works fine. Of course i'll use your solution.
About file() and dbf() i've send to you another message i'll wait.
Carlos