> > I'd like to question the need to support these features in
> newtorque for
> the
> > following reasons:
> >
> > - They are supported differently in different DBs
>
> We could probably find some middle ground here. DBs I worked
> with (not a
> huge collection though :-) all supports the ANSI standard, so
> we might do
> this as a start. If your DB does not support ON *** or does
> it in a strange
> way you can just remove them from your newtorque schema.
>
> > - I personally do not see much use in them (at least in newtorque)
>
> I usually prefer to have at least ON DELETEs in my schema.
> It makes for
> less code and better db integrity.
>
> > - Supporting them only adds unneccessary work.
>
> Only if you don't use it :-)
I just sent it to Leon personally and I'll cut'n'paste:
I do not know very many databases but of those I know only Oracle supports
CASCADE for example (do you know others?), does MySQL?, etc...
I think we do not need to support it HERE, I believe in most cases it would
be preferrable to perform appropriate action (i.e. cascade delete)
programmatically in java. It might even Make sense to add methods such as
doDeleteCascade to Peers so that it does the cascade delete and doDelete
will raise an exception if you are trying to delete a row with child
records. Alternatively we can add a parameter to doDelete and doUpdate to
tell it whether it should be cascaded or not. I think this way you'll have
much better and granular control over this (dangerous) behaviour. Makes
sense?
> > So I suggest that we drop support for them in newtorque. What your
> opinions
> > here?
>
> I'd like to see them stay and get developed a bit further,
> but if I'm the
> only one, I'll go with the rest of the group and remove them...
What do other people say? BTW I would apreciate info on support of these
features in other DBs.
> > Would it make sense to do some sort of review of what
> features we want and
> > do not want in newtorque? Any ideas on this?
>
> Just some ideas I can think of for now. This is more to get the ball
> rolling than a set of specs...
>
> Main features for newtorque:
> * Generation of sql schemas for multiple DB's
> * Generation of Peer/MapBuilder/DO classes
> * [Far Future] Generation of basic screens/actions
> * [Far Far Future] Visual Interface
* Dynamic generation/OM class compilation/SQL execution
* {?]per-table XML file breakdown
>
> SubFeatures:
> * A rich set of data types (String, Int, Double/Money,
> Date/Time, Blob,
> etc?)
> * Foreign Key Support [single/multiple rows]
> * Primary Key Support [single id row for ID Broker]
I would say:
* Primary/Unique Key Support
* Unique ID support
> * [Maybe] Generation of 1-m stuff in Peer classes
* Readonly Table/Column support
What else?
Fedor.
------------------------------------------------------------
To subscribe: [EMAIL PROTECTED]
To unsubscribe: [EMAIL PROTECTED]
Search: <http://www.mail-archive.com/turbine%40list.working-dogs.com/>
Problems?: [EMAIL PROTECTED]
