Most people, knot tyers included, believe that a knot is a means to an end rather than an end in itself. They believe that knots are tools which serve purposes over and above the knots themselves - for example, to bind two or more objects together, to suspend a heavy object in mid air, etc. In this context, if knots "work", if they "do the job", they are OK : the ends justify the means.
This is not a "wrong" attitude, it is just a restricted, narrow and naive attitude, which limits the scope of the knot tying endeavour.
We have attempted to discuss this issue in the Forum in the past, but it seems that when a view happens to be defended only by the "minority", the "majority" is suddenly turned into something of a
mob - perhaps because established religious fundamentalists can not tolerate heretics !
A couple of recent comments on what I dare to think on this, offer me the opportunity to try to explain my view yet another time.
The purpose of a knot is to be knotted, and to remain so. The purpose of a knot is to entangle one or more pair of rope ends, and to keep them entangled, and securely "locked" inside the knot s nub, even if/when it happens that some of those ends are tensioned. In other words, the purpose of a knot is to immobilize two or more tensioned ends of the entangled rope(s), so they do not slip out of the knot, even if/when they are tensioned.
The knot is a rope-made mechanism ( even the simplest knots incorporate many rope-made
simple machines ; levers, pulleys, wedges ), which can, and should, be studied as such. If possible, it should be analysed in simpler "elements", and the reasons behind its structural and functional properties should be investigated. ( Why it jams ? Is there some feature common in all knots that do jam ? Is there a way to design knots that do not jam ? )
People built the steam engine for a purpose, but the study of its properties, as a mechanism and a heat exchanger, contributed to the fields of physics called mechanics and thermodynamics - which allowed us to built better steam engines, among other things !
I could also express the same view otherwise :
"What is a knot ?"
"I do not know - please, tell me one kind of knot."
"A bend."
"OK, then. A bend is a tangle of two ropes which joins the one end of the one with the other end of the other."
Now, I believe that the unbiased reader got the idea - and I will proceed with a question which asks for a knot for such a particular "purpose"

:
A midline-to-midline TIB-to-TIB bend, able to withstand loading from any pair of its four ends.
From time to time, I try my hand in such a knot, but I have never came across something that caught my eye - so I now think that there may not be such a knot, because there can not be : the "purpose" itself seems to be too complex, and it leads to too complex knots (*), which can not be considered as practical.
(*) For example, they should incorporate one or more ugly unloaded bights stemming out of the nub and gawking around - the ones I had called "
dL s bights" :
http://igkt.net/sm/index.php?topic=5269.msg34591#msg34591 http://igkt.net/sm/index.php?topic=4425.msg28055#msg28055 The simplest example of a "
dL s bight" - which, just because it is very simple, it is not sooo ugly...- is the bight of ABoK#1492, shown in the attached picture.