The basic principal is that your program never stops running, but instead blocks on a loop, and continues it's progress whenever a new request is needed.
This is ment to aid in initializing generally unchanging things only once for several request, like database connections...
What you need to make sure is that no information in the initialization part of the program is oriented towards any client in particular, and that everything is kept enclosed and secured - objects destroyed and cleaned up, and so forth.. Generally the loop scope should take care of this...
Update: I forgot to mention that CGI::Fast inherits from CGI, which is more thoroughly documented. Their behavior is the same in terms of data processing, so the migration from CGI::Simple to CGI should be very similar if not identical to the migration to CGI::Fast...
You may also be interested in mod_perl.
-nuffin
zz zZ Z Z #!perl | [reply] |
By the way, what is the difference between mod_perl and fastCGI ? I mean I know how mod_perl works, but how does FastCGI? What are the differences ?
| [reply] |
| [reply] |