Still COM+ in .NET for multi-thiered programming?
-
Hi, There is a lot of talk these days about the .NET vs J2EE thing. I have taken a look on both sides and I was wondering if the COM+ architecture is still a fundamental thing in .NET for distributed computing? (Yes I know, .NET eventually emerged out of COM+2.0 or NGWS or whatever)... But for distributed computing, have things changed from COM+ ? Is there such a thing as J2EE that offers server-side containers being statefull (<-> COM+, as far as I remember?), how slick and integrated is everything? They must be a lot more elegant right now, can someone please give me a pointer to information in msdn about that? thanx for reading, cursing and flaming ;-)
-
Hi, There is a lot of talk these days about the .NET vs J2EE thing. I have taken a look on both sides and I was wondering if the COM+ architecture is still a fundamental thing in .NET for distributed computing? (Yes I know, .NET eventually emerged out of COM+2.0 or NGWS or whatever)... But for distributed computing, have things changed from COM+ ? Is there such a thing as J2EE that offers server-side containers being statefull (<-> COM+, as far as I remember?), how slick and integrated is everything? They must be a lot more elegant right now, can someone please give me a pointer to information in msdn about that? thanx for reading, cursing and flaming ;-)
Com+ is fundamental in distributed computing, whether you use .NET or not, the services provided by Com+ are the fundamental thing (like distributed transactions, object pooling, the easy thread development architecture, etc..), which were not supplanted by any new technology (yet). In .NET you can write Serviced Components (this are the Com+ components that we write today in Visual C++, Visual Basic,..), these will run inside Com+ and can take advantage of the services provided by it. Right now there isn't support for statefull components, in the sense of J2EE, but microsoft it's working on a new technology that permits something like this, I can't remember the name (something like "Object Space" or so). The idea of statefull componentes is just the opposite to the spirit of Com+, where one of the rules is "do not maintain state" (see Transactional COM+: Building Scalable Applications by Tim Ewald http://www.amazon.com/exec/obidos/ASIN/0201615940/qid%3D1013822174/ref%3Dsr_11_0_1/103-4000867-4379823) Hope this clarify a bit. (and sorry about my english :) Andres Manggini. Buenos Aires - Argentina.