M
Martin Platt
Hi,
I've been working on designing new architecture to be
used for future development using C#.
We're going quite well with the architecture, having
originally decided not to use the dataset, datatable,
dataadapter type implementation. Instead we are using
our own implementation using business entities,
controlling processes, etc., so we have definite
business, data and presentation layers. We're happy with
this, but the further on we go, the more out
implementation seems like an alternative version of the
dataset. What I'd like to know is how does the dataset
compare to other implementations for large distributed
systems, as well as small desktop ones?
Part of the architecture would involve having the
business entities (the classes containing the data)
having their own internal representation of state
(unchanged, insert, update, delete) Any thoughts on how
this might be accomplished? I had a thought that be
could call a set method from a base class that would
compare, and set the values in they're different, and
also manage the state information. However, that becomes
a problem when trying to pass a property as a ref value.
Any ideas on how else this might be accomplished, or
should we just get the users of the subsystems to set the
mode manually (not my preference)
It would be great with the above to be able to override
how the set part of a property works, so we can always
make the property perform a check and keep internal state
up to date, but there seems no way of doing this that I
can find.
There must be a simple way of doing this - I can't be the
first person to have this problem, surely!
Any thoughts, suggestions, etc. would be gladly accepted.
Cheers,
Martin.
I've been working on designing new architecture to be
used for future development using C#.
We're going quite well with the architecture, having
originally decided not to use the dataset, datatable,
dataadapter type implementation. Instead we are using
our own implementation using business entities,
controlling processes, etc., so we have definite
business, data and presentation layers. We're happy with
this, but the further on we go, the more out
implementation seems like an alternative version of the
dataset. What I'd like to know is how does the dataset
compare to other implementations for large distributed
systems, as well as small desktop ones?
Part of the architecture would involve having the
business entities (the classes containing the data)
having their own internal representation of state
(unchanged, insert, update, delete) Any thoughts on how
this might be accomplished? I had a thought that be
could call a set method from a base class that would
compare, and set the values in they're different, and
also manage the state information. However, that becomes
a problem when trying to pass a property as a ref value.
Any ideas on how else this might be accomplished, or
should we just get the users of the subsystems to set the
mode manually (not my preference)
It would be great with the above to be able to override
how the set part of a property works, so we can always
make the property perform a check and keep internal state
up to date, but there seems no way of doing this that I
can find.
There must be a simple way of doing this - I can't be the
first person to have this problem, surely!
Any thoughts, suggestions, etc. would be gladly accepted.
Cheers,
Martin.