[MTOS-dev] New Dispatcher On the Way

[Reed A. Cartwright]

Byrne Reese wrote:
> If creating a new call back will facilitate the integration of this code -
> then let's start in one of our next dev cycles to integrate that callback
> for you. Then we can do some additional testing and optimization.
> 
> Thoughts?
> 
> Can you submit a patch for the new callback in the core?

I'm a bit swamped with real life right now, and probably don't have the 
time soon to figure out how to add the right call backs.  (Unlike many 
of y'all I'm only a hobbiest when it comes to MT hacking/development.  I 
get paid to be an academic.)  I've never used callbacks.  However, I can 
explain what I think needs to be done, in case someone with more time 
can patch the system.  I guess I can propose something if SA doesn't 
mind if I submit an untested patch.

1. Establish pre_accept and post_finish callbacks.  Pre_accept happens 
before a connection is established and post_finish after it terminates.

2. Update the standard bootstraper to call those routines where 
appropriate.  Pre_accept should be called before 'new cgi' or 'new 
cgi::fast'.  Post_finish should be called after the connection is 
closed.  On FCGI this can be done with a call to the Finish command. 
Forking can be used for CGI support.  (I can then modify MT-Dispatch to 
be consistent.)

3. While your at it, modularize MT::Bootstrap based on what I use in 
MT-Dispatch, therefore I don't have to copy code anymore, I can just 
call MT::Bootstrap functions.  I can probably quickly propose such 
changes, and if someone gives me the 411 for creating a new callback 
types, I can add those as at the same time.

-- 
*********************************************************
Reed A. Cartwright, PhD     http://scit.us/
Postdoctoral Researcher     http://www.dererumnatura.us/
Department of Genetics      http://www.pandasthumb.org/

Bioinformatics Research Center
North Carolina State University
Campus Box 7566
Raleigh, NC 27695-7566

Cuiusvis hominis est errare, nullius nisi
insipientis in errore perserverare. --Cicero