I thought the same thing at first but those hover over deals are not meant to be links. If you read the subtext they say things like "download the source from news" etc. So it is meant to be a one page sort of affair. I dig simplicity! download and hack!
Their front page reads like a pompous ad. Why can't people simply describe their product and let it speak for itself if it's so great? (or maybe it's not)
Here's a better version, IMO:
"The Kanamit Web Framework provides a templating engine and a persistence mechanism that runs on [list of supported relational databases].
This web framework leverages the power of Lisp well. We have developed a library that comprises many real-world usage scenarios. It's portable across lisp implementations. We got it running on several flavors of Lisp on Linux, FreeBSD, Solaris, Mac OS X and we are currently porting it to our faithful Symbolics 3600. We even got it running inside Emacs 22.
You should get a decent development environment. Emacs 22, Slime and GNU-Lisp are what we use, but if you really want to use Eclipse (some people compare it to Emacs) it should work.
You can also order our book here. It took a while to write, but we wanted it ready for the public launch of our project."
Not only that... The other co-prankster (the one with the press release) had already done their own co-prank (a book on the framework that went on sale the same day)... It wouldn't be nice from our part to leave him alone, so we had to play along.
Well... You know that all the major hard to do components of our little imaginary framework already exist: there already is a Lisp-based web server that could handle HTTP requests (we assume it would stand behind a cache such as Varnish) and there must be at least a good few RDBMS persistence mechanisms for Lisp by now. As for the template engine, I would tend to go for something like ZPT (Zope Page Templates) as mixing code in the presentation layer does not hit me as particularly fancy. AFAIK, there is no implementation in Lisp, so, it would have to be done. For the rest, the glue connecting the HTTP to the persisted objects, it would be rather easy to do patterning after Rails or Django.
So, I think if we decided to really do it, even without full dedication, we could have it ready for release in about a year.
