>> Sure, it makes it a little easier for the dev, but in the end, you are just learning 5 times the amount of Opa when you could learn each thing.

I see it the other way round. When you are dealing with all the languages and formats we have now there is a huge amount of wasted duplication. Look at how you concatenate strings in Javascript versus PHP. Or calculate a random number. Or iterate an array. obj->method() or obj.method() or obj->val or obj.val ? How about encode a non-ascii character in a URI, or in XML, or in HTML? Which characters are reserved in HTML or XML or JSON or a URI or SQL or a regular expression? How do you walk a DOM in Javascript versus PHP (versus Ruby/JSP/NET/etc.) ? Frankly, I have better things to do with my limited time on this planet that memorize all these little conventions over and over again slightly differently each time.

If I can learn and use Opa and not have to meddle with the others, then it's a win. But if I have to use Opa to perform string concatenation to construct javascript or SQL. then it's a loss. I can live with having to concatenate HTML and CSS (but I'd rather not).

Wt is cool, but compare the listings for the Wt chat application versus that of Opa. How many orders of magnitude are there between the lengths of the two listings? 1? 2? The Opa listing would fit on a t-shirt. So I think that is at least a partial answer to your question of "why a new programming language required". Personally, I also find that most new language announcements are unnecessary. There are already many great languages available, could not a tool or library accomplish the same thing? However, in this particular space I'm less opposed to the introduction of a new language. The whole web app thing is already a rotten stinking soup of languages, protocols and formats (SQL, PHP/JSP/NET/Ruby/Python, HTML, CSS, XML, Javascript etc etc etc) and I am very sympathetic to any one-language-to-rule-them-all solution. If I can worry about one instead of legion I'll be better off. However, though I have not yet used Opa or Wt yet, looking through their example code listings I see HTML strings being constructed in the Opa listing, and Javascript code being concatenated in Wt. And for both CSS files seem to also be BYO. That isn't necessarily unreasonable, but you can see the slope we are on. So now I'm using Opa/Wt to write HTML (or Javascript) and I have to provide the CSS myself. Am I really ahead of the curve, or wearing mittens? How is this different than (gulp) PHP? What about the database? Does my Opa/Wt code need to construct SQL query strings? If I, or my designer, wants to add some new jQuery plugin that will be announced tomorrow, how is that approached? I've been fairly impressed with Opa from what I've read on its site. I'm hoping to get some time to actually play with it and maybe get to the bottom of some of these questions.

I think that dreamweaver will essentially become a photoshop add-on. This way, very smart graphic designers will make a beautiful graphic, click on the "Dream-Weave-it" button, and presto: A complete web page with rollovers from layers.

That plug-in already exists:
http://www.medialab.com/sitegrinder/