Login | Register
My pages Projects Community openCollabNet

Discussions > dev > direct instantiation of TableManager subclasses

Project highlights: Architectural Overview

joist
Discussion topic

Hide all messages in topic

All messages in topic

Re: [joist-dev] direct instantiation of TableManager subclasses

Author dlr
Full name Daniel Rall
Date 2000-07-26 16:44:36 PDT
Message "Daniel L. Rall" wrote:
>
> Does it make sense to instantiate a TableManager subclass directly, or
> should I use the DataManager layer, even though it will be slower?
> Though I *can* access the ctor of the sourceXchange RfpManager, it seems
> to me that the *right way* to access it is through the DataManager.
> However, this is the only TableManager that I will be instantiating, so
> I am not sure that the overhead is justifiable...thoughts?

Okay, I was wrong about being able to instantiate a RfpManager without a
DataManager reference to pass to RfpManager's ctor. This wasn't
explicitly noted anywhere, but it now is in TableManager via a private
argless ctor. TableManager needs that reference to DataManager to be
able to call dataMgr.registerToCl​ose(DataSet) without throwing a NPE.
Issues noted, and hopefully fixed! :)
--

Daniel Rall <dlr at finemaltcoding dot com>

direct instantiation of TableManager subclasses

Author dlr
Full name Daniel Rall
Date 2000-07-26 16:00:16 PDT
Message Does it make sense to instantiate a TableManager subclass directly, or
should I use the DataManager layer, even though it will be slower?
Though I *can* access the ctor of the sourceXchange RfpManager, it seems
to me that the *right way* to access it is through the DataManager.
However, this is the only TableManager that I will be instantiating, so
I am not sure that the overhead is justifiable...thoughts?

--

Daniel Rall <dlr at finemaltcoding dot com>
Messages per page: