OOP Industry Standard

[dRAMmer]

Is there such a thing as OOP Standard? This is the way how we implement OOP concepts.

- Our classes are stateless
- When fetching data, methods returns "recordset" (ie. dataset, datatable)
- When updating data, methods accepts parameters of what-supposed-to-be properties of the class. To make it neat we pass a structure instead of a number of paramter. The structure's elements represent the supposed-to-be properties of the class.

I have read all the debunking articles and praise article about OOP (I prefer the latter). I guess we have implemented OOP fair enough. Anyone who can shed light on my inquiry is much appreciated!

[Dave Sell]

Stateless Classes? The very nature of a Class is an intelligent data structure which implies a state-ful Object. Care to extrapolate?

[dRAMmer]

Hi!

There are many articles that can be a basis for a debate about a class being stateless. But nevertheless you would agree with me that if you use a class in MTS then you can't have it as having state.

Now, .Net offers remoting, how should classes be defined for that then? I guess they should be stateless too (or not)?

[Dave Sell]

I think a Class that is stateless actually becomes a service. As far as .NET remoting, it is a .NET replacement for DCOM. I don't use COM+, but I do use DCOM, and my Classes are not all State-less.

Although I do have a DCOM Singleton which is stateless but only since it is more of a message-router than a state-ful Object.

So I guess you will have some state-less Objects in .NET remoting, but this attribute is not decided by choice of architecture, but rather by necessity of functionality.

P.S. As .NET remoting supplants DCOM, so Indigo supplants .NET remoting, according to Microsoft itself. Seems like this Web-Services bandwagon has everyone abandoning their loyalties.

[techgnome]

I think a Class that is stateless actually becomes a service. As far as .NET remoting, it is a .NET replacement for DCOM. I don't use COM+, but I do use DCOM, and my Classes are not all State-less.

Although I do have a DCOM Singleton which is stateless but only since it is more of a message-router than a state-ful Object.

So I guess you will have some state-less Objects in .NET remoting, but this attribute is not decided by choice of architecture, but rather by necessity of functionality.

P.S. As .NET remoting supplants DCOM, so Indigo supplants .NET remoting, according to Microsoft itself. Seems like this Web-Services bandwagon has everyone abandoning their loyalties.

Gads, I hope not.... we've got parts of our system where remoting will make sense.... but not as a WebService.... I think WS has gotten out of control and we'll see a swing back in the next couple of years. And eventualy it will be replaced by something else.

-tg

[Dave Sell]

Sorry I can't dig up the keynote, but as I recall it was centered around WCF (pronounced In-di-go lol).

In it, the M$ dood said that WCF is here to replace this-that-and the other, and one of them was definately .NET Remoting.

I shuddered to think someone would suggest replacing a technology like .NET Remoting (Remote instantiation of state-ful Objects) with a technology like Web-Services (Remote method calling of state-less Objects). (Can that even work??? I think not).

I know - the whole WS hype started to get out-of-hand about a year or 2 ago. It's mostly-pointy-haired marketing goobers who are responsible for this, similar to what happenned at Y2K. Both are real and valid mole-hills that have gotten swept up into irrationally-porportioned mountains.

[dRAMmer]

Thanks for the input! I'll leave this thread open for at least a week and wait for more opinions.