Re: WSAPI replacement draft

[Matthew Frazier <leafstormrush@...>]

On 01/23/2011 07:55 PM, Chris Babcock wrote:
> Warning: Long, hypothetical, web app oriented and not at all specific
> to Lua. (I'm fine with follow up going off-list, too.)
> [hypotheticalness omitted for brevity]

Okay, to be honest I'm not sure if I entirely get this, but it seems 
like what you're saying is that LASI could be a way for Web applications 
and servers to dynamically reconfigure themselves to support different 
modes of operation (i.e. coroutines vs. OS threads). The goal of this 
would be to allow a single server to run multiple applications, and tune 
itself to each of them.

This is one of the coolest ideas I have ever heard (although a bit 
impenetrable at first). However, it's kinda overreaching LASI's goals, 
which (for now) are just to be a well-specified and easy-to-use server 
interface. (That, and supporting both OS threads, forked processes, and 
coroutines in the same server is likely to be somewhat problematic.)

Though it reminds me a bit of the "Setup Environment" section of the 
"Python Web API 1.0" draft I have linked from the spec. The premise is 
that the callable request handler is obtained from *another* callable 
(since the draft is for Python, classes are likely to be used) that is 
passed a "setup environment" containing all of that information about 
the server ahead of time, so it could prepare appropriately before the 
first request even hits. That might be worth looking into.

> Chris
> --
> 51st century guy
--
Thanks, Matthew Frazier