Mindaugas Kavaliauskas wrote:
The only ODBC application I've ever tried to write in my life is contrib/rddsql/sddodbc/tests/test1.prg, so, I have no idea that the answer is. I would expect to be it a little slower on client part (I mean result set manipulation).
I found some quite old "research". The figures indicate very little speed difference.
One more SDD driver I am thinking about is ADO SDD. But I'm not sure if it is worth to invest time. Using IDispatch* interface will make result set manipulation even slower (character message names should be resolved to do IDispatch call). Using more low level interface (I guess it is called OLE DB) requires using of many interfaces : IADOConnection, IRowset, IColumnsInfo, ... The full list is on http://msdn.microsoft.com/en-us/library/ms709709(VS.85).aspx and has 81 (!) interface in specification. I do not want to put my hands on dirty things like this.
Yes, I think OLE DB would be a world of pain. FWIW I have used ADO directly from Harbour (simply as an object), and it is acceptable for drawing down a wildcard list of 2 or 300 debtors and details (< 1s). This must involve the same resolving of character names to IDispatch calls that you mention.
Regards Alex _______________________________________________ Harbour mailing list Harbour@harbour-project.org http://lists.harbour-project.org/mailman/listinfo/harbour
