Another example: IE has a nasty habit of trying to be "too helpful" when interpreting URL's. Never mind that the data format is explicitly specified in the HTTP headers... if the URL ends in something which "looks like" a filetype extension (i.e. GIF or PDF), IE will ignore your explicit headers and substitute its own "better judgement". This is in _absolute_ violation of the RFC's and W3C specs, which clearly indicate how such data should be interpreted. However, it's perfectly in keeping with Microsoft's overall "We know better" attitude.
Meanwhile, in every case that I can remember, Netscape has been compliant. Every single time we ran into a problem, we could benchmark against Netscape and watch it work perfectly. Our code causes us to dig mighty deep into technical details, and yet I've never found a deviation between the RFC's, W3C specs, and Netscape. Doesn't mean they don't exist, but I've never personally seen one.
And No, I'm not a "Microsoft basher". I use their products to develop code and run equipment, and have done so for many years. They can and do produce some fine software. But that doesn't mean I don't get frustrated when they thumb their nose at industry standards. There's no excuse for doing that.