I disagree and stated why, for this application.
But I will further revise my recommendation based on seeing what
material --natural fiber (jute)-- he is working with. And in this, I
can see a Xarax “additional … re-tucking” working well for some
alternative but like knot to the Dbl.Harness --viz., the S88 bend.
I think that the basis of the S88 --viz. Ashley’s #1406 “Whatnot-1”–
is perhaps best, given the jute; Xarax’s S88 extends this base with a tuck
of the tails. The S88 seems to be easier to set tight with pull on both
SParts & tails (iteratively) than the Dbl.Harness, and might be surer
to jam!? The latter perhaps requires more flow of tension in the
knot and with a frictive and not-so-strong natural-fiber material it’s
better to go with an even simpler structure (which in turn would work
less well/surely in slipperier material). The S88’s tails are perpendicular
to the axis of tension, so are pretty easily cut off at the not body; and
they leave behind a workable structure of the Whatnot/Grass Bend
even should the final tuck be lost.
But this is an awkward knot to tie, and is not done so all that
material-efficiently. Go with the [i]Whatnot-1 /i!
Cf.
Hmmmm, further fiddling with these options in some sort of small,
natural-fiber cord (about 3mm diameter, 6-strand laid), points strongly
to the Whatknot-1. It’s quickly tied and can be set manually to a nice
snug, jammed state in this frictive stuff; I should think that it will do
the job well in that jute. Take care in setting the knot to balance each
side’s pull & nip on the other. This knot has a very small profile, with
the tails trimmed off!