Well you could just open a new cursor do what is needed there and your
original has never left the base row. Just go back to the original cursor.
Jim
Mahanimann wrote:
> In SQL you don't need an index you can request sets of records in any
> order with any filter. Having an index speeds things up. Generally
Yes I know. I must have expressed myself wrongly.
What I mean is when you f.ex. use a generic sort-function, you need to :GoTo(nTheActiveRecnoBeforeTheSort) after the sort. But, the :Recno() numbering change with the sort order, so one can't use :GoTo(). When you can't use :GoTo(), you need another unique value to seek on. When the environment is generic, you need to know where to find that unique value within the oCursor.
That's where my question about :aPrimaryKey comes in. If I could access oCursor:aPrimarKey in the generic sort-function, my "problem" is solved. Of course it is simple to overcome this, but why code extra for something that is allready there. I do define :aPrimaryKey in all cursors.
That is one part of my question, oCursor:aPrimaryKey is there, so why not use it? (and why not give access to it?)
The other part is: How do seasoned SqlExpress'ers handles situations where they need to go back to a specific row/record (after f.ex. a change in a browse sort order). The function is generic, so it has no clue about which table it is sorting. The sort-function only recives oCursor as a parameter.
I'll be more specific in my phrasing from now on :-)
Thanks for any reply,
Mahanimann