NCSA httpd

At the time, CERN httpd was available, but users had to compile it themselves; with NCSA httpd it was just a matter of downloading the binary and the configuration file, editing a few lines as indicated by the comments, and starting it - your own webserver within five minutes.

Also, development for CERN httpd had paused, and there was a great need for new features. The CERN httpd code base, a big library backported from Objective C, was powerful but not exactly inviting, which more or less excuses Rob for starting from scratch.

NCSA httpd rapidly became the most popular webserver and the focal point of innovations such as CGI, SSI, password- and IP-based access control. Due to its rapid development it also introduced many small design flaws and some bugs.

NCSA continued development after Rob McCool left for Netscape, and introduced some more innovations, but development only really gained its momentum with the independent code fork known as Apache.

NCSA httpd may have been critical in spreading the use of the Web: it made it very easy for prospective users to try it out on their own systems. But it was never a well-designed piece of software. If the Apache team had had the guts and foresight to base their work on the technically superior CERN httpd, they would have saved themselves and us a lot of trouble in the long run.