This is a placeholder task to track IO::AIO patch for OSX.

http://search.cpan.org/~kevinj/AnyEvent-SCGI-1.0/
http://search.cpan.org/~orlandov/Catalyst-Engine-SCGI-0.02/
http://www.mems-exchange.org/software/scgi/

http://gist.github.com/192083

to deal with X-Forwarded-For

This is just because we don't support HTTP/1.1 in standalone. -0 option (force HTTP/1.0) would work.

http://robhudson.github.com/django-debug-toolbar/
http://www.brynary.com/2009/4/22/rack-bug-debugging-toolbar-in-four-minutes

This would be a middleware to install debug toolbar panels and provide a basic fallback so app frameworks can record stuff, as well as a set of plugins to override common module methods like Memcached or DBI.

response_cb doesn't implement poll_cb, and actually don't know how... it causes a deadlock easily

also to give an option of passing $self in the callback, maybe with an option or a different name like inline_object_with_self

We probably shouldn't use Path::Class here...

Tatsumaki did Transfer-Encoding: identity but it caused issues with Apache proxy, so i removed it, but this will then cause problems with Firefox though:

The problem was, if we omit Transfer-Encoding here, middleware::chunked chunks the response regardless of the response is HTTP/1.1 or 1.0, and Tatsumaki::Server and Plack::Server::AnyEvent don't do HTTP/1.1 response yet, so Firefox ignores the Transfer-Encoding: chunked and fails to parse the pushed stream.

Eh.

I guess the right thing is to let our AnyEvent do HTTP/1.1 as well as middleware::chunked not chunk the response, but kill keep-alive connection after the response.

deflate response doesn't get decoded properly by LWP clients and Firefox.

include fastcgi_params should fill most of fastcgi params but apparently PATH_INFO is not passed. Should be documented in FCGI.pm

.

if (__FILE__ eq $0) {
   exec 'plackup', '-a', $0, @ARGV;
}

would be a common pattern in .psgi script to be runnable it self. Make it a module?

When you stash wide character (utf-8 strings) in a response body the Standalone server gets killed. It's an user error but at least the server shouldn't die.

Integrating Plack::Test callback with LWP::UserAgent should be an interesting hack.

request_send needs a matching spec which requires host and port which Plack::Test mangles. I guess the client needs a second parameter to get those parameters for that.