I agree with the robustness principle....
....And what I say, is that if a software fail due to bad input, and the programmer forgot to implement error checking, is still the fault of whoever that provide the bad data, not the fault of the programmer or the software in question.
These two statements contradict each other. The purpose of the robustness principle, in terms of things that work on the internet, is that you pretend the network is a goddamn warzone. That means the blame lies on the receiving app that crashes, and it's programmer - because they should have expected users to do naughty, impolite things.
About web development, it is in fact easy to do web pages that works in most browsers, including IE.
Validate your pages against HTML 3.2, no CSS and no JS, and your pages will work wonderfully in most browsers including old versions of IE like IE 4.0.
And there it is, full-retard. Everyone should just go back to lynx and Mosaic because one douchebag, one of all seven people on the planet who claims to be a programmer yet is still retarded enough to use MSIE, doesn't like it when his browser crashes on some mal-formed markup.
The browser CRASH when visiting bitcoin wiki thats a unacceptable fault of the wiki! The wiki are sending something that causes the browser to CRASH, be it something in SSL handshake, be it something in HTML code, be it whatever. But please fix it!
For sites that care about their audience, it would be prudent for the webmaster to sort out what the issue is. That doesn't mean it's the webmaster's fault, make no mistake about it the blame lies entirely with your browser, but to maximize your audience you have to decide how much time is worth it supporting older versions of software. Bitcoin's target audience is, for the most part, IT security-conscious people, who normally wouldn't be caught dead using a version of MSIE that crashes on malformed input.