Okay, last night I was roaming around Microcenter in thier notoriously huge computer book section (Easily the size of a mini B&N bookseller), I came across a book called "Win32 Systems ... " I forget. It wasn't typical win32 system programming. It was intensely geared towards networking. Topics like WNet, Winsock, RPC, TAPI, creating a BBS and much much more. I took particular notice of the RPC section, which was notoriously similar to DCOM with it's end result. Got me researching. DCOM was created as a COM solution to RPC (RPC is supposedly difficult -- thos in the book, it looked deceptively simple). My problem, as an assembly programmer (and C), is if I don't want to use DCOM, how can I use RPC? There appears to be a complete lack of examples. Anyone know of a place where I can learn more than rudimentary knowledge of the subject? I don't know if it's new and improved with modern COM+ or whatever... winsock, but for my version management software to write for the IDE project and possibly a remote capability, specially for the debugger, this may or may not be what I'm looking for... Thanks, _Shawn
Gee, I hate to be the teller of bad news, so I'll break it to you gently. DCOM is built on LIGHTWEIGHT RPC (Remote Process Communications). Basically, they are the same thing, just it can be lightweight because it starts by assuming only certain data types and such will be passed, so it can forget about all the other things being passed. The orgional COM specs were built with different process/same processor in mind, this was easily extended to different process/different processor by DCOM Mind you, I've yet to go out of process (or out of thread even) with my COM stuff, thus avoiding the whole RPC issue. But from what I have read, if you stick to the basic automation types (which are plentyful), you can just use the Universal Marshaller provided by COM. ----------------------------------- "Oh my god! Space Aliens! Donít eat me, I have a wife and kids! Eat them!"
Actually, I release DCOM was build on top of RPC... in such a way that really, DCOM is little different than COM, except is't out of processor boundries... However, to use RPC, I need not DCOM. Maybe I should just get the book, but it didn't focus on any one topic in any depth. It had an example of a calculater whose calculations resided in a process on a nother machine over some network. In less than 15 lines of code, he had a calculater that did its stuff over a network, and the calculater logic even so, was much less than 15 lines of code as well. I was just wondering if someone knew of a resource which explained some of it, or had RPC samples. I really don't want to resort to the COMplexities of COM if I don't have to... Thanks, _Shawn
I hope i'm not completely out of line, but if I remember well, Rpc are just a way to communicate from port to port using the sockets. The real way to communicate is by the sockets and rpc are some standard to use. Dcom must be a standard over the rpc standard, yet they all uses the sockets. Anyway, with visual studio, there's a help file named rpc.chm that shows a lot about rpc and how to program it in c++. I've seen a couple of samples, but I can't find them right now. If you can't find the files, I've put them at http://pages.infinit.net/souimet/asm/rpc.chm and http://pages.infinit.net/souimet/asm/rpc.chi (if it's required) This message was edited by Silas, on 3/23/2001 12:08:10 PM