Well, that's one way to see if anyone is paying attention -- put in
an error that should call for a report! Good show, Derek, yes.
Now, I suggest that post's photos remain as now revised,
showing both incorrect & correct: (1) Derek's report won't
make sense in the absence of the image & (2) there's merit
in showing the simple change, to guard against other's doing
it by mistake on their own.
Also, please note that the mistaken image however presents
the knot symmetrically (and which that particular knot lends
itself to) -- the two Overhand components have the same shapes.
This is a good thing for seeing several alterations, as follows.
The center of this structure can be seen as interlocked bights
and that form taken in silhouette leads to four possible interlockings:
with crossings Over/Under one way there are two directions the bights
can be seen to flow, SPart-to-end; and then the crossings are reversed
and again there are two directions to take -- 2x2 =4.
The mistaken knot is in fact Ashley's #1031/1048-like, near what
Asher named "Shakehands", as Asher loads the opposite ends from
what Ashley shows (& implies qua bend) in 1031.
The correct image reverses direction to this over/under structure
and gives the Versa-Vice loading.
Reversing the Over/Under state for the mistaken image gives the
Ring Bend, "strong form" (my term for loading exterior parts).
Reversing direction of this reversed O/U results in a knot more
like Versa-Vice but hard to set symmetrically (as is e.g. Ashley's #1453).
To the four variations just outlined, there comes the switching
of which end of the Overhand component is loaded, which
yields four more; and since the components are asymmetric
then one can toss in asymmetric loading to double the 8 =>16.
Whew, knot invention can be exhausting (and this sort of rigor
takes some of the fun out of it, for me).
--dl*
====