Talk:IUCV

Jeff,

About the 'client/server predates...' assertion. The fact - IIRC - is that client/server interaction in distributed systems actually *predate* IUCV ! If you look at TCP (which is a form of connection based transfer between units) - it predates IUCV by about 10 years !

Ok.. The fact is, I was trying to keep this article as 'factual' as possible, not trying to outline the pros and cons or anything as such. I believe discussion about the merits of IUCV, or its disadvantages compared to other communication forms - namely its horrendous path length when executed on 43xx class machines, probably do not belong in the article...

But then again ;) I appreciate the input !

Ivan Scott Warren (talk) 01:42, 13 June 2009 (UTC)[reply]

Hi Ivan, Nice to communicate with you on this site as well! :-) I just wanted to emphasize the historical context of IUCV, which really did popularize at an early point a client/server style of programming between (virtual) machines. For what it's worth, IUCV came in with VM/SP Release 1, which was in 1982. TCP/IP as a networking protocol was still in its infancy, and wide-spread use of TCP/IP protocols to create applications was still well off into the future. And, if we really want to push the argument, there's VMCF, which was in VM since VM/370 Release 3 PLC 8. And that's in 1976! I don't claim there was nothing else out there of similar flavor (I didn't survey all platforms), but remember that even the PC didn't come out till 1980 or so.

What's really poignant is that I remember those releases when they were new, and I personally built applications using those protocols then! Time flies, doesn't it? Jsavit (talk) 01:52, 13 June 2009 (UTC)[reply]