On Tue, Jun 29, 2004 at 06:58:35PM -0400, Hendrik Boom wrote: > I have it on moderately good (but anonymous) authority that a few > years ago it was deliberate Microsoft policy for their C++ implementations > to be *in*compatible with others. I don't know whether that policy is > still in force now.
I worked there from June 2000, the annoucement of "NGWS" (which became .NET), to Jan 2003. I saw several product cycles, how it went from perfectly plausible explanations of how interop would finally be "normal" (as in Java web services from IBM consumable from Visual Studio and vice versa), to the usual "yeah, if you're a nutjob, you can make them interop with a lot of effort and in a limited way" -- and I never saw (except for two occasions someone deliberately say "we want this not to interop by design" (One was a manager two levels up and one level sideways from me; and the other was Bill Gates -- both saying "we want people to only think of our XML processor when they think of XML"). What ends up happening is just 1,000,000 neutral things, a bunch of sideways hoohah, and it just sort of ends up that way. If there is a deliberate attempt to sway the process it was at a level I never saw. To be fair none of the Java web services interact well mix-matching Servers and Clients either. I have this on authority from Britt McAlister, a super java expert who came over from @Home, who did some interop work between Tomcat, .NET, BEA, and IBM. All of the three Java-ish things are relatively-mutually-incompatible with other Javas as they are with .NET. I did come across the phrase "trench warfare" used to describe IE 3.0. One or two people sheepishly acknowledge malevolent intent. But a lot more of it is --- something else, not directly describable as "evil intent," but just "stuff that happens". Kind of like how politicians always end up the same no matter how they start out. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
