Re: Revert function

From: Andrus Adamchik (andru..bjectstyle.org)
Date: Thu Dec 05 2002 - 17:08:35 EST

  • Next message: Craig Miskell: "Re: Revert function"

    Craig Miskell writes:

    > Well, in the face of deafening silence :-),

    Isn't it what is called "lazy consensus" ? :-)

    > I have made it so that when
    > a toMany relationship is modified on a Committed CayenneDataObject, the
    > persistence state of the object is changed to Modified. Standard code
    > that already executes for every updated object checks and finds that no
    > other attributes have changed from the snapshot, and thus avoids an
    > unnecessary update.

    Like you may have guessed, the reason that object state on to-many
    modification wasn't switched to "modified" was that nobody cared. Now that
    we do, I guess it is OK... Don't see any problems with it (given that you
    estimate that performance impact is small).

    > If I don't hear any screams of protest in the next day or so, I'll
    > commit it. :-)
    >
    > In the principles of XP-style babysteps, I'll leave it at that for now
    > (the other suggestions from Holger look interesting, but aren't required
    > for this bit of functionality).
    > And in the principles of Opensource "scratch the itch", I'll leave the
    > others suggestions for more interested parties to pursue as they see
    > fit. :-)

    So in case I missed something from the previous emails, your version of
    "revert" will completely revert all the changes, handling all possible
    relationship and attribute changes? I guess this is a complete feature by
    itself. The rest of "undo manager" things maybe treated as separate
    features.

    Andrus



    This archive was generated by hypermail 2.0.0 : Thu Dec 05 2002 - 17:08:39 EST