big mistake was to always start with a Tumble Hitch.
I finally reread the title of the post and realized
that I needed to begin with a Draw Hitch.
I can now even clearly see the "eye glasses" in the third diagram below.
. . .
Sorry I am so slow to understand, but it has been fun!
I like the knot, now that I understand it.
But you STILL don't --albeit you're getting closer.
2 B quick :: you need to load what you show as Tail,
and in your Step 3 be taking the longer end (what
you mistakenly load) out-away-right_of shorter end
(to be loaded) and make the slip-tuck from behind,
coming forwards (so, opposite to your arrow).
In presentation, I lay the to-be-loaded S.Part U-fold
on the FRONT of the object, and ... .
But given your Step 1, in Step 2 show the brought
around front U-fold put around the 1st (nothing more),
and give the black arrow straight up for its path
in next move,
which will cast the "cat eyes" INTO THE S.Part's U-fold.
(looking at the cat-eyes, one can see that with the
correct twist to them one can put the 2-B-loaded
strand such that it turns around the other strand
and of course also the frame).
Bit by bit,
but do note that I continually said to begin as
for trad. knot and not in the revised orientation
of "tumble".
NB :: the knot of your latest venture does NOT
look so good --
robust vs. pull-through capsizing!
If one goes cross-eyed in those cat eyes --i.e.,
twist them (crossing this U-fold's legs thus) before
they surround the other U-fold,
then there is some cordage to block pull-through;
but I'm not sure that it's fully adequate, and folding
the slip-tuck even if that doesn't capsize out of being
knotted will mean inhibiting release (so, too, can the
trad. Highwayman's H. fail :: holding, but no longer
releasable).
One must also carefully consider how this knot
is to be used --materials and circumstance ::
i.p., might the release require pulling on both
ends --the Tail (of course) to release, but then
the S.Part to completely unknot things?
Might the knotted structure have to be pulled
AROUND the tied-to object? (which might not
be so easily done).
And all this from some 50' or so distance,
rope running over surfaces and with its elasticity
and variable friction vs. surface not able to give
steady good force into the knot (which worked
well in the arm chair in the Ivory Tower!) ??!
And so far below, one is wondering what has
happened up there, not knowing if perhaps
a state has been obtained for which now one
should pull the S.Part to finish release, else
pulling Tail will try to take the tangle back
around the object.
--dl*
====