This document is part of The Web at Nexor.

ArchiePlex Info

Introduction

ArchiePlex is a suite of gateways allowing World Wide Web users to lookup files available by Anonymous FTP. It was originally developed for Plexus, but is now CGI compliant, and can be used by other WWW servers.

ArchiePlex returns the references it found as full hypertext links. This allows you to retrieve files, or browse FTP directories directly. ArchiePlex also gives you special features to do specific kinds of search (substring, exact, case insensitive substring), and to have them sorted by hostname, or by date (latest files first).

To see ArchiePlex being used around the world, see the List of Hypertext Archie Gateways

If you experience problems, check if they are caused by recent changes.

If you're familiar with ArchiePlex, and consider running it at your site, check out the technical details.

The three ArchiePlex flavours

ArchiePlex has three different front-ends to the Archie search:

ArchiePlexSimple

This is a limited but simple interface. The user only sees a searcheable document. The server administrator determines the type of search executed, and the archie server to be used. This is especially useful for sites running a archie server themselves.

ArchiePlex

This is a more verbose interface, giving you full flexibity to determine the kind of searches executed, the archie server to be used, and the format the result is to be displayd in. The coices are made by clicking on links.

ArchiePlexForm

This interface uses the HTML+ fill-out forms. It offers the same felxibility as ArchiePlex, but you make your choices about the type of search and the archie server to be used by choosing from menus and clicking on buttons. This is the most sophisticated and efficient interface to ArchiePlex.

ArchiePlex Output

All three interfaces produce the same output: A hypertext document with active links to the hosts, files, and directories that archie has found. By selecting a file you immediately retrieve the file, and by selecting a directory you will be able to browse through the directory.

ArchiePlex Background

This all started out as a shell script, called by inetd, to be able to integrate archie into the WWW.

However, Bourne shell is not the best of programming environments, so I rewrote it in C, called it Warchie, and added the special searches and the manual. Today, Warchie is still in use by the Communications Research Department at the University of Nottingham.

When I switched from the NCSA HTTPD server to Plexus it occurred to me I could integrate Warchie into Plexus, so I didn't have to run it from inetd, and have integrated logging etc. So I rewrote it in Perl, and sent it of to Tony Sanders for inclusion into Plexus. He modifed it to fit into Plexus better and be fully HTTP 1.0 compatible. Bill Fenner has fixed some bugs, added some error checking, and made the code more secure.

Then when Mosaic started supporting forms I changed ArchiePlex to use forms too. Around that time I noticed that Rutgers University had taken the ArchiePlex code, and produced a simplyfied version for their server. I then rewrote these gateways to all use the same code.

Recently Guy Brooker (guy@jw.estec.esa.nl) wrote an AWK version of ArchiePlexForm, that uses the CGI standard, and can therefore be used by NCSA's HTTPD. It used Mosaic's internal images, and added two more configuration options, so I decided ArchiePlexForm should have these too :-).

Lee McLoughlin (L.McLoughlin@doc.ic.ac.uk) contributed code to sort the hostnames into a preferred order, and have special mappings for FTP sites that are reachable by HTTP too.

Oscar Nierstrasz wrote some small wrappers around ArchiePlex which made it run under CGI server. This urged me to change the base distribution and make ArchiePlex fully CGI compatible so no wrappers are required.