One starting point might be to define the perceptual space the knot will be defined within.
All my knots are created within the few cubic feet of space in front of my body and as such are referenced to my 'self' in terms of spatial location.
This space has right and left sides. It has up i.e. in the direction of my head and down, in the direction of my feet.
Near relates to the space between the knot and my torso, while away relates to the other side of the knot away from the 'near' side.
Over relates to passing over the near side or up or top side of the knot while under refers to passing the back or away side or the under or the down sides of the knot.
As I am right handed, for me the perceived origin of a knot is the cord held in my right hand and I would instinctively describe the path of a cord through a knot starting from this origin point.
Using this perceptual space in conjunction with the Crossing Point notation, the description for the simple Overhand Knot might look like this:-
[language] = KHTML
[knot name] = "Overhand Knot"
[WKI] = "
http://igkt.pbwiki.com/Overhand%20Knot"
[function] = "Stopper knot"
[status] = "whole knot"
[overs index] = {3:6},C1,C2,C3
[origin] = "Right hand"
[space] = "Body normal"
[path]
[cord end] = "End#1"
[cord name] = "Cord1"
[direction] = l(eft), h(orizontal)
Over C1
[direction] = b(ack), u(p), l(eft)
Under C2
[direction] = l, f(orward), u
[direction] = l, f(orward), d(own)
Over C3
[direction] = r(ight), b, h
[direction] = u
Under C1
[direction] u, l, f
[direction] d, l
Over C2
[direction] l, b, d
Under C3
[direction] l [length] = "SP"
[cord end] = "SP#1"
[/path]
Were you able to follow the trace of the cord using the directions of l(eft) r(ight), b(ack) f(orward), u(p) d(own) and h(orizontal) and the compound directions such as "l, f, d" which is a three dimensional move left, forward and down simultaneously?
Would the instructions have been clearer if they had distance values included, for example, [direction] = r, b, h [length] = 4 and possibly curvature indicators indicating the tightness of the curve (perhaps in cord diameters)?
I realise this will immediately fail the Dave Root test of "the man in the street", but then nobody would expect to be able to read HTML without first having learnt the syntax of that language, despite this, HTML drives the internet today. Likewise, KHTML would have to be learnt in order to read and interpret it whilst delivering the potential of giving us a basic language with which to unambiguously describe a knot structure.
Doubtless we will need a few more whistles and bells to give us a full working KHTML vocabulary capable of describing the majority of conventional (i.e. functional) knots. Such a language should be able to be utilised to describe the complex (i.e. decorative) knots, but I will leave that gauntlet to others who might have a need to do such a thing.
Longer term, we could consider adding components to describe the structure of a dressed knot and components to describe methods of tying, but for the time being perhaps we could 'cut our teeth' on a language to simply describe the structure of a knot.
I have put a description of this outline definition of KHTML syntax onto the Wiki so that members can develop it further, see:-
http://igkt.pbwiki.com/KHTML