Subject | Re: SQL Update where params |
---|---|
Author | Marco Menardi |
Post date | 2004-07-02T17:14:42Z |
I'm really interested in this answer.
As you can get from the first Helen answer, seemd that there was some
sort of "magic" for preventing locks in FB, but I never succeded in
making it work (the same that happens to you: B overwrites A changes).
So I had to set PessimisticLocking to true and relay upon the "dummy
update" that IBO performs, with all the problems from triggers fired
in the database (expecially in the "After Edit").
The TClientDataset way of updating using all the previous values
fields for the where clause seemd to me a much better solution, or at
least, one that could be widely used in all the situations a conflict
is unlikely but it's important not to overwrite other's work.
So I do hope this time I will get the definitive answer ;)
BTW, Helen, that was one of the topic I asked you for a specific Tech
Sheet some days ago... what about use this thread to write it? Who
better than you? ;))
regards
Marco Menardi
As you can get from the first Helen answer, seemd that there was some
sort of "magic" for preventing locks in FB, but I never succeded in
making it work (the same that happens to you: B overwrites A changes).
So I had to set PessimisticLocking to true and relay upon the "dummy
update" that IBO performs, with all the problems from triggers fired
in the database (expecially in the "After Edit").
The TClientDataset way of updating using all the previous values
fields for the where clause seemd to me a much better solution, or at
least, one that could be widely used in all the situations a conflict
is unlikely but it's important not to overwrite other's work.
So I do hope this time I will get the definitive answer ;)
BTW, Helen, that was one of the topic I asked you for a specific Tech
Sheet some days ago... what about use this thread to write it? Who
better than you? ;))
regards
Marco Menardi