The URL Access Counter - A Bit of Web History

Variants of this counter program, the first to appear on select sites of the early World Wide Web, led to the development and widespread dissemination of the popular and much abused "hit-counters". The original was coded in C by Frans van Hoesel (hoesel@chem.rug.nl) of the Rijksuniversiteit Groningen for his WWW Expo project. Unlike its many descendents, this program painstakingly constructed a simple image, digit by digit, to display access counts. Since images, unlike text strings within an HTML document, are routinely cached within the user's browser, the Hoesel counter typically is insensitive to document reloads within a sesson and is fairly accurate in determining the number of primary accesses to a particular page. An instance of this counter, previously run remotely at MSU, provided access counts for the Expo homepage later hosted by sunsite.unc.edu at UNC, Chapel Hill. Visit Expo to see its visitor count and here for background info on Frans' Expo project. Go here for a description of the original program and a pointer to the source.

A port of the C version to PERL was accomplished by Dan Rich of Silicon Graphics (drich@corp.sig.com).

Erik Nygren (nygren@mit.edu) of MIT and NASA Ames Intelligent Mechanisms Group then modified the SGI PERL script slightly to execute from the host's cgi-bin directory. This is the execution mode used here.

Local modifications include a new locking mechanism to prevent accidental reset when the referenced page is being heavily accessed. For details on the ITC implementation, contact Joe Burch (jbb@virginia.edu), ITC Unix Systems, at The University of Virginia.