db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Fernanda Pizzorno <Fernanda.Pizzo...@Sun.COM>
Subject Re: Subject: Detectability of updates in DRDA
Date Tue, 07 Mar 2006 08:49:15 GMT
Bryan Pendleton wrote:

> Fernanda Pizzorno wrote:
>
>> What we are looking for is a mechanism that we can use to implement the
>> ResultSet#rowUpdated() method.
>
>
> I think I understand better now. Let me try stating it in my own words,
> and you can tell me if I am on the right track or not.
>
> - We want to implement ResultSet.rowUpdated() in the Derby Net Client
> - On the server side, we have a data structure which holds a bunch of
>   auxiliary information for each scrollable updatable cursor. In 
> particular,
>   this data structure has a boolean field for each row which records
>   whether or not the row was updated.
> - So once we get to the server side, we have the information that we need
>   to be able to answer the question.
> - That means that now all we have to do is to figure out what message the
>   client should send the server when the client calls rowUpdated(), and
>   what message the server should return.
> - Lastly, you have been looking at the "Update Hole" mechanism in DRDA,
>   because it seems to specify a way to convey extra information about
>   a row in the result set.
> - However, the precise specification for the "Update Hole" field in DRDA
>   has to do with the scenario in which the update to a row causes it
>   to *move* within the result set, so that it's no longer at the same
>   *position* in the result set, which is a different semantic than what
>   is needed for the rowUpdated() method.
> - But it seems like the mechanism is roughly similar to what is needed,
>   so you're proposing to extend it slightly to have an alternate sort
>   of "Update Hole" which has the semantic: "row is still in the same
>   position in the result set, but it has been updated in ways which did
>   not change its position".
> - Then, when the client received that special variant Update Hole, it
>   would return "true" for the rowUpdated() method.
>
> Did I understand correctly?
>
> thanks,
>
> bryan
>
>
Yes, that's what I meant.

Fernanda

Mime
View raw message