KnotPlot looks like somebody gave topology a graphics workstation and then refused to stop experimenting for thirty years. It is part mathematical reference, part software project, part image gallery, and part personal laboratory.
Robert Scharein built KnotPlot out of work connected to his PhD at the University of British Columbia’s Imager Computer Graphics Laboratory. The software visualizes and manipulates mathematical knots and links in three and four dimensions, and the site says almost all of its images were generated with KnotPlot itself.
That immediately makes the website more interesting than a normal gallery. The pictures are output from the subject being documented. Decorative knots, Brunnian links, torus knots, knot carpets, stereoscopic images, cubic-lattice knots, hyperbolic knots, Celtic knots, Moebius-strip knots, and stranger constructions sit alongside thesis material and technical references.
The software side is just as substantial. KnotPlot can load a database of more than 3,000 knots and links, let the user sketch knots in 3D, construct them with Conway notation, use a tangle calculator, generate special knot families, and transform existing knots. Current builds still exist for Windows, macOS, and Linux, which is remarkable enough for a program whose copyright line reaches back to 1992.
The site also retains unmistakable old-Web texture. Some galleries are polished and useful. Others are openly filed under a section titled “Pages with little or no interesting content,” because apparently deletion is for cowards. There are PostScript examples, POV-Ray output, thesis figures, stereoscopic experiments, and pages whose main justification seems to be that Scharein once wanted to see what a particular knot would look like.
That accumulation is precisely why the site matters. Modern software documentation tends to explain features. KnotPlot preserves years of intellectual wandering around the features: experiments, pictures, weird side paths, references, obsolete rendering techniques, and the research lineage behind the program.
For somebody studying mathematical visualization, early Web graphics, scientific personal sites, OpenGL software, or just the visual culture around topology, it is an unusually dense primary source.
It also demonstrates one of the Ancient Web project’s recurring lessons: a specialist website becomes valuable by refusing to flatten itself. It does not need to turn every experiment into a clean product page. The mess contains history.
More sites like this turn up in CacheRat’s 1,967 Ancient Web Domains research list, but KnotPlot stands out because the underlying software is still alive while the website continues to carry decades of accumulated research around it.