On Wed, 2 Jan 2008 17:49:36 -0600, "Jay Sprenkle" <[EMAIL PROTECTED]> wrote:
>On Jan 2, 2008 5:31 AM, Kees Nuyt <[EMAIL PROTECTED]> wrote: > >I found a solution that seems workable. I ended up rewriting my query >class so it assumes a bunch of things which are valid for my application >(but not for general usage). The columns in the result set can always be >correctly identified given a single key, that key is the first column of >every query, and the code does not generate it's own SQL to do updates >or deletes. >I have to craft the sql for each query/update/delete. It's disappointing >since it could have been much more elegant and generic (and a lot less >work for me!). > >> >> I know what the answer will be: "This is not a planned feature. >> Adding this would slow down the code for vast majority of people >> who do not need it." > >I noticed the column metadata routines are not included by the default >compile settings >so Dr. Hipp isn't averse to putting in things that aren't commonly used. Ok, interesting. Thanks for your feedback! -- ( Kees Nuyt ) c[_] ----------------------------------------------------------------------------- To unsubscribe, send email to [EMAIL PROTECTED] -----------------------------------------------------------------------------