Very Computer

I created a perl script, but no matter what I do, I still get the error from
the http server:
        access to /usr/local/www/cgi-bin/read-data.pl failed
        for dark_star.pbgi.com, reason: malformed header from script

In the process of debugging I eliminated all lines of the script so that the
ENTIRE code looks like:
        #!/usr/bin/perl
        exit;

Yes.. I have tried it with
        print "Content-type: text/html\n";

Still the same error.  What's up with that? The script has 100766 permissions
and the uid has read/execute permissions... this is identical to other
working scripts in the cgi-bin directory.

What is the black magic needed?

(perl 5, FreeBSD 2.01, Apache)

Peace,
Blake

Quote:> Yes.. I have tried it with
>    print "Content-type: text/html\n";
> What is the black magic needed?

  *silence* Well, *ok* :) - try:

  print "Content-type: text/html\n\n" ;
                                  ^^

  and see if that might not just help a wee bit.

--

<                                                                           >
< " Once the rockets are up, who cares where they come down?                >
<  That's not my department," says Wernher von Braun. " - Lehrer            >

instead of:

        print "Content-type: text/html\n";

HTTP requires a blank line between the end of your headers and the actual
document.  Therefore a single \n does not signal the end of your headers.
Two \n's will.

--
Richard Trott

I wonder if anyone else does this in Unix -- I have a number of the CGI
environmental variables set in my shell environment in my login file
(since I use tcsh that's actually .cshrc).  When I want to debug scripts
I just grab the query_string from the form submission, write it to a
file and get the content-length for that query_string.  Then I set the
CONTENT_LENGTH variable in my shell and I can start running the script
at the command line, piping in the query_string from the file I saved.

Does anyone else do this too?  I find it incredibly helpful when I want
to step through or run it interactively.

- kov

--

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

: I wonder if anyone else does this in Unix -- I have a number of the CGI
: environmental variables set in my shell environment in my login file
: (since I use tcsh that's actually .cshrc).  When I want to debug scripts
: I just grab the query_string from the form submission, write it to a
: file and get the content-length for that query_string.  Then I set the
: CONTENT_LENGTH variable in my shell and I can start running the script
: at the command line, piping in the query_string from the file I saved.
:
: Does anyone else do this too?  I find it incredibly helpful when I want
: to step through or run it interactively.

Hello Kevin,

 yeah,I do that too.

Another hint for you is to assume (using 'su') the userid that the
webserver runs as (the spawned child process.. the userid set in your
httpd.conf file!).

This will help you find such things as file access permission problems as
well..

Steff