New Flapjax release coming; plans to drop compiler

[Shriram Krishnamurthi]

In early July, we intend to release a new version of the Flapjax
library. This release will be backward-compatible with the current
release. We will fix bugs, improve performance, and provide better
documentation.

The principal change is that we wish to no longer support the Flapjax
compiler. More on that below.

As we discussed recently, we've been using Flapjax extensively:

http://groups.google.com/group/flapjax/browse_thread/thread/f935ab046daa349a

These projects use Flapjax as a library, eschewing the compiler.
While the compiler affords some syntactic convenience, we find that we
rarely use it. For instance, Continue 2.0 has 200 calls to lift in
3,000 lines of code. Moreover, feedback we've gotten from other users
indicates that they, too, prefer to use Flapjax as a library, because
it provides greater flexibility and better integration with other
libraries and tools for JavaScript.

If anyone has significant objections to our dropping the compiler,
please do let us know! (If there are none, we can put out the new
release sooner.)

The Flapjax Team

[Jaime Vargas]

Not significant objections, but I prefer the flapjax language syntax
and its succinctness.

[Shriram Krishnamurthi]

On Tue, Jun 10, 2008 at 12:22 AM, Jaime Vargas <j...@mac.com> wrote:
>
> Not significant objections, but I prefer the flapjax language syntax
> and its succinctness.

Breaks my heart, too. But the insularity of the current compiler
isn't a good thing. Ultimately we have to break down and do some
serious research to figure out how to smooth this integration. But a
research-less, naive compiler is not, I think, a good thing in the
balance.

Shriram