Hi, Thanks for the feedback.
I voted "no" because for very little improvement (which really isn't worth > it) > It would help me to understand why you think this. It's difficult to generate queries with the best execution plans if the type information isn't good. Floating point values are part of SQL. Is your thought that they aren't used enough, that this implementation is missing something, or something else? * users foot-gunning themselves due to misuse of float values (yes, they > will put floats where real is needed) > Did you see the latest rev of the RFC? It should address this point. If you pass a string and label it PDO::PARAM_FLT, only emulated prepares and pdo_sqlite will cast it to a float. And that's only for consistency with how they handle other types. Does that make you more comfortable? Or is there another use case you're concerned about? That said, see https://wiki.php.net/rfc/voting#resurrecting_rejected_propos > als about resurrecting the vote: it will need to wait a few months. > Based on previous feedback, I thought the new version was materially different. But if a handful of voters from the last round respond and say that these changes won't change their vote, then sure, there's no point in rerunning this. Thanks, Adam
