On Thu, 2004-10-28 at 15:16, Lars Heuer wrote:
> Hi Lars,
>
> > | Do you think, that we can develop a generic interface and the tolog
> > | specific helper object only supports a subset of the interface?
>
> > In theory we could. I'm not sure that would be as nice to use as a
> > less generic interface would, but we might be able to get round that.
>
> Maybe we could design the tolog helper object with generality in mind
> (i.e. not calling the package org.tmapi.tolog, but org.tmapi.query
> etc.).
> If we have more knowledge about the TMQL specification we should be
> able to extend the query package / helper object.
> Would this be okay for everybody?
>
That is what I had thought too - if we make a first version of the
interfaces based on what we know about tolog and what we expect TMQL to
be like, we can then make a second version later if we need to make
modifications to support TMQL when it is cooked.
> [...]
> > | release, because I belive we'll have to change the TMAPI interfaces
> > | for transaction support.
>
> > I'm not sure we do. I think only the outermost classes need to change,
> > probably, such as TopicMapSystem.
>
> Yep, that's what I meant. I don't think, we've to alter every
> interface, but some. I think TopicMapSystem and TopicMap are
> candidates. But my sentence wasn't exact.. . ;)
>
Do we really need to change those interfaces, we have the HelperObject
mechanism for these sorts of extensions. Why can't we just define a
"TransactionManager" helper object ?
Cheers,
Kal
--
Kal Ahmed <kal-5x+ggncFKvmB+jHODAdFcQ@xxxxxxxxxxxxxxxx>
techquila
-------------------------------------------------------
This Newsletter Sponsored by: Macrovision
For reliable Linux application installations, use the industry's leading
setup authoring tool, InstallShield X. Learn more and evaluate
today. http://clk.atdmt.com/MSI/go/ins0030000001msi/direct/01/
|
