There's a lot of anti-WYSIWYG going on here. Let's not forget that the problem is not with the WYSIWYG idea itself but with the failure of web development tools to deliver it.
WYSIWYG is perfectly viable (and essential) in application GUI development - nobody would expect to code up the layout of a Swing or .Net form.
On the internet: for historical reasons we are saddled with 'markup language', of one kind or another, which does not sit easily with WYSIWYG. If I am composing a user experience I really do not want to worry about document formatting and structure (HTML), I want to just put images and controls where I need them to be. Today this is not straightforward.
It turns out that *ML (e.g. XML) is more useful for information exchange (i.e. systems integration) than it is for making nice GUIs. Conversely, *ML of any kind is not best suited for making functionally rich applications.
Markup language, by its nature deals with document structure first and foremost. Internet applications inherited *ML as the default GUI but that does not mean that it is suitable for today's internet applications.
For this reason we have seen the development and proliferation of rich technologies like ActiveX, Applet and Flash. IMO, *ML is an internet legacy that will eventually go away or be fenced back into its comfort zone.
The alternative is that someone develops designer interfaces that are WYSIWYG and are easy to use, genuinely hiding the underlying *ML. Personally I think that *ML will eventually die as a GUI media although its offspring (e.g. XML) will (rightly) live on as integration protocols.
WYSIWYG is perfectly viable (and essential) in application GUI development - nobody would expect to code up the layout of a Swing or .Net form.
On the internet: for historical reasons we are saddled with 'markup language', of one kind or another, which does not sit easily with WYSIWYG. If I am composing a user experience I really do not want to worry about document formatting and structure (HTML), I want to just put images and controls where I need them to be. Today this is not straightforward.
It turns out that *ML (e.g. XML) is more useful for information exchange (i.e. systems integration) than it is for making nice GUIs. Conversely, *ML of any kind is not best suited for making functionally rich applications.
Markup language, by its nature deals with document structure first and foremost. Internet applications inherited *ML as the default GUI but that does not mean that it is suitable for today's internet applications.
For this reason we have seen the development and proliferation of rich technologies like ActiveX, Applet and Flash. IMO, *ML is an internet legacy that will eventually go away or be fenced back into its comfort zone.
The alternative is that someone develops designer interfaces that are WYSIWYG and are easy to use, genuinely hiding the underlying *ML. Personally I think that *ML will eventually die as a GUI media although its offspring (e.g. XML) will (rightly) live on as integration protocols.