Someone (iirc Linus) mentioned in an interview that software is not if you build it they will come. If you care deeply about a project you will need to do what your consumers/users request otherwise your project won't go far (e.g. Andrew S. Tanenbaum thinks that he was unlucky with his timing, etc. and that minix is a much better design than the linux kernel, but who is currently using minix?). I think we all know who the leader and manager of the Linux kernel is. On the other hand, much/most of the current linux kernel does exist because of a do mentality rather than people asking Linus for added features.
My primary point is we cannot expect developers to flock to pyjamas just because it exists, there does need to a sufficient ecosystem that contains much of what the average developer looks for. However, I don't see script kiddies or any another similar stereotype thinking pyjamas (or GWT) is the next big thing given the amount of understanding required and the differences between old school pages with mostly static html and a rich internet application. As an example (in my opinion), one only needs to look at the confusion that arises with "pyjs doesn't work with framework X due to HTML error Y". If the error is with respect to jsonrpc, it is clear to me that the necessary handling must be implemented in the server code. However, what seems to be the case from anecdotal evidence from google searches is "pyjamas is broken because it won't do cross origin jsonrpc". My 2 cents

