The strength of vector graphics is their ability to scale to any size, small or large, without affecting the quality of the image. This is because the image is based on object definitions, such as draw a line from point A to point B, and the only difference in sizing is the length of the drawn line. There is no “low, medium, high” quality to vector graphics: they’re always the highest quality.
Another advantage to vector graphics from a web perspective is that we can create relatively complex, and large, graphics, and the bandwidth requirements can be a whole lot less than the requirements for a comparable raster image. Additionally, vector graphics lend themselves to manipulation, both at the server and in the page once it is loaded. We can dynamically generate a raster image, such as a JPEG, based on changing data, but the process isn’t as simple as doing the same with a vector format, such as SVG.
A paper presented at XML Europe 2001 by Chris Lilley and Dieter Weidenbrück titled “WebCGM and SVG: A Comparison” lists the following requirements for a web-based vector graphics system:
Integratable with other web content
What worked for 2001 still works today. Though the paper addressed two specific formats, WebCGM and Scalar Vector Graphics (SVG), the concept can also include Vector Markup Language (VML), in addition to the three-dimensional formats such as Virtual Reality Modeling Language (VRML) ...