Peter Fraser wrote: > The web designer had web pages that he was trying to convert from Apache to > nginx. > Those pages were calling Perl programs from nginx using slowcgi. > > I was the one that was configuring nginx. > It would have made my life easier if a couple of points were added to the > documentation.
Documentation for what? :) > > 1) that running "slowcgi -d" give extensive tracing information of all > requests. > (but remember to /etc/rc.d/slowcgi stop" first) > 2) that all programs and their libraries have to be copied into /var/www man slowcgi already documents both of these. the consequences of chroot aren't stated explicitly, but the implications of "chroot(8) to /var/www" are "standard unix lore". > 3) #! Lines are examined in text files and that information is used to find > the programs, > But no search rules are applied so you have to type > #/usr/bin/perl > Not > #!perl This too has nothing specifically to do with slowcgi, that's "how unix works". It is, fwiw, documented in the execve man page, but there are dozens of programs on OpenBSD that can exec other programs or scripts. There's no way we're going to explain the rules in every one of them. If you have specific changes to request, we can have a look. The documentation can be terse at times, and sometimes you can only see what you're looking for after you already know it, but sometimes it's also just a matter of reading closely.
