On 23-Feb-2004 Jeff Urlwin wrote: <excessively snipped> Not that I am following this DBI discussion here closely but I remember one more issue with DBD::ODBC and MS SQL Server.
> That's a little harder for me, but I'll try to summarize what I did and try > to remember the reasons why. It got nasty there for a while, especially > with SQL*Server stored procs (just look at all the 45_xx releases! I do > remember, at first, it was a simple: if you called for more results, I'd > check then. But there were a few nasty instances where, inside a procedure, > you'd have: a procedure returning output parameters > update > select a, b, c > update > update > select d, e, f > SQL*Server returned, respectively: > empty result set > a, b, c > empty > empty > d, e, f output parameters returned on the last call to SQLMoreResults. > What DBD::ODBC does now is that the end-user will only see (I think! It's > been a while!): > a, b, c > d, e, f e.g. SQL = {? = call fred(arguments) } the ? output parameter is not available until the last SQLMoreResults returns SQL_NO_MORE_DATA. Don't know if this makes any difference to the discussion - just thought I'd mention it. Martin -- Martin J. Evans Easysoft Ltd, UK Development