Re: CGI, semicolons, and so on...
Charles Henrich <henrich@crh.cl.msu.edu>
From: Charles Henrich <henrich@crh.cl.msu.edu>
Message-id: <9312292151.AA20618@crh.cl.msu.edu>
Subject: Re: CGI, semicolons, and so on...
To: www-talk@nxoc01.cern.ch
Date: Wed, 29 Dec 1993 16:51:34 -0500 (EST)
In-reply-to: <9312292024.AA00540@hopf.math.nwu.edu> from "John Franks" at Dec 29, 93 02:24:12 pm
X-Mailer: ELM [version 2.4 PL21]
Content-Type: text
Content-Length: 1231
> only as an existence proof that it is not too difficult to design a
> syntax whose server implementation will have several important
> properties:
I agree wholeheartedly, my original intent was to make the scheme cleaner and
allow for a more efficent implementation. the stat() call on local unix
systems is a waste of resources, and on a AFS system a whole lot of wasted
resources. (We run AFS all over here)
My scheme was simply to have the server ignore everything beyond the first
semicolon found in the URL. That information would then be passed (in any
number of ways) to the scripts that get called. This format works quite well
in satisfying the following parameters:
> 1) Name and path of the file/script
Initial part of the URL
> 2) string to be put in PATH_INFO environment variable (if any)
The text beyond the semicolon
> 3) Is this file/script to be executed or treated as text
The srm.conf file. Or heaven forbid the execute bit on the file, after all,
unix has been using the EXECUTE bit to mean that the file is EXECUTABLE (go
figure) for the longest of times..
-Crh
Charles Henrich Michigan State University henrich@crh.cl.msu.edu
http://rs560.msu.edu/~henrich/