T. Berners-Lee and others (of The Internet Society) write of URIs: "A Uniform Resource Identifier (URI) is a compact string of characters for identifying an abstract or physical resource." The URL is considered a "subset" of the URI, by the way. The syntax for relative URIs have escaped my understanding for a large time.
What I see now centers upon understanding what the period ('.') and the double period ('..') means to Microsoft- or certain Unix-based command line syntax. In such syntax, the single period means "the current directory" and the double period means "the directory immediately above this directory." When I apply this syntax to the relative URI, they begin make sense to me.
In the relative URI, the forward slash '/' is used as a delimeter, marking the number of directories above the current directory. The forward slash is the obvious choice because of its use in Unix shells.
Now with the single period, double periods, and the forward slash, I have a "relatively" complete understanding of URIs. The following are examples of "Normal" URIs:
SRC="./bitmaps/foo.gif" ---the source is the GIF image "foo.gif" in the "bitmaps" directory. This directory is immediately below the current directory.
SRC="../bitmaps/foo.gif" ---the source is the GIF image "foo.gif" in the "bitmaps" directory. This directory is immediately above the current directory.
SRC="../../bitmaps/foo.gif" ---the source is the GIF image "foo.gif" in the "bitmaps" directory. This directory is two directories above the current directory.
"Abnormal" URIs:
The double period(s) and its delimiter(s) should always lead the URI. This means that SRC="bitmaps/../../foo.gif" would be what the Internet Society calls "abnormal."
The single period should only be used once so that "././foo.gif" would also be very abnormal. There can only be one current directory!
For a formal study, see "Request for Comments: 2396" (rfc2396) by the Internet Society. As of this writing, this document is located at http://info.internet.isi.edu/in-notes/rfc/files/rfc2396.txt.