The suggestions made by Charles have nothing to do with the particular
problem he's bringing up, what he's done is made it so that the run-time
config files can define scripts in their directories. I don't see how your
suggestion of placing an = in the extra path information helps this unless
you are requiring every script to have extra path information. What Charles
suggested is to have a semicolon before the extra path information, again,
this does not help the particular situation he added code for unless you
require extra path information to scripts!
What am I missing about your proposal or Charles's that will alleviate this?
* There is also no reason that
* directories containing scripts have to be listed in configuration
* files and processed on server start up.
As a server administrator I would say that I should have at least some
control over what people are running on my server...
* Or that scripts need to be
* distinguished from ordinary files by a naming convention which the
* server presumably decodes. Adding unnecessary complexity to the
* server is undesirable. You have now added more code to your server
* and you still don't have the functionality (much less the simplicity)
* that a very minor change in the protocol would give.
Which change is this? I must not understand your proposal. I will add that
the Alias mechanism is already there, and that allowing ScriptAlias from
config files is a realatively minor change.
* I have to plead guilty to starting this discussion at much too late a
* date. The suggestions should have been made several months ago when
* the CGI standard was still in flux. One possibility is that this
* issue could be addressed in CGI/1.1. I would point out that nothing
* suggested requires any change in browsers and changes to scripts and
* servers should be very minimal. I think that relatively few scripts
* currently use PATH_INFO.
Well, for every tech support question I still get about the NCSA script
interface, I will testify that even the most minor change, especially the
incompatible ones, is potentially a very major change.
* I would be interested in hearing from server writers, like Rob McCool, Tony
* Sanders and the CERN server author. Also the views of script writers
* would be valuable. Are there examples of widely used scripts that would
imagemap would require additional changes.