From: Dmitry Y. <di...@ma...> - 2001-04-23 23:14:00
|
Hi Claudio and others, > While I agree that such changes should be discussed and BLR > additions cannot > be committed without discussion, I should remember people in > this list that > this offering was made a lot of time ago. Alexander worked on > the IB (not > FB) code but his patch can be applied to both engines. Since > nobody felt > interested, I "resurrected" the posting, Frank replied and > Dmitry Yemanov > contacted Alexander. So, to made my position clearer: Exactly. And, as far as I remember, a couple of developers have asked us to keep working on it and include it in the next version of Firebird. > - I do agree that such changes cannot be committed without extensive > discussion but if there's no comments when an offering is > done, the few > people that get interest in it try to do the best. Agree. > - I won't expect year 3000 for INPR/BORL to sync with all our changes. > [snipped] I won't either. > - I consider this function an "escape" function, I mean, one > that should be > used sparsely and only when there's no static solution. If > the applications > developer want to use it for anything... well, there's all > kind of people in > the world, we know that some are brilliant because others are > silly, by > comparison <g>. The facility to interpret some commands has > been raised in > IB-Priorities, IB-Architect and other lists several times. Although we've never had this ability in Interbase/Firebird, I can easily imagine situations where this command will be very helpful. Maybe anyone who used DBMS_SQL package in Oracle could agree. > - Alexander wrote that he didn't know how to avoid recursion > problems, if I > remember. Perhaps someone knowledgeable (disregarding the agreement or > disagreement) could help to make this feature smoother. He has solved them. Now the depth of recursion is limited, so every transaction can execute up to 50 nested dynamic SQL commands. It looks to be enough. > To put in short: keep changes safe and controlled, but do not stop the > enhancements. > :-) Good summary :-) Thanks for your comments. This explanation would seem satisfactory. Cheers, Dmitry |