Bends

[zoranz]

Hi Dan & Luca!

I tried #1425 (for the first time) using Luca's (first) diagram, and to my surprise got a nice facility, very strong and reliable bend, the only thing I do not like that it is very difficult untied (prone to jamming?). I'll put this band in my library; but will tie it exclusively looking at the diagram (never under table or in the dark ). In that sense i want to tell that diagram is very clear! (Personally i don't like additional lines made by hand/pencil...)

In meantime I found Xarax's pic which resembles somehow the original Ashley sketch; and original from Ashley, too.



And question for Dan: why do you think that butterfly has not place among best bends?

Regards, ZZ

[Luca]

Hi zoranz,thank you
 
I concur,#1425 can be very difficult to untie:maybe you can try leaving the portions of rope adjacent to the tails just a bit loose during the tying;if  after the load it still should be difficulties,during the untying you can try pulling on the tails in order to loosen the "closure of the shell" (using the methods for #1425 described in my previous posts, the knot in fact, when it is tight, it "closes" as a sort of bivalve;if Dan Lehman in his previous post referred to this particular movement when he mentioned the capsizing, then this also happens using the second method I explained, contrary to what I wrote in my last post;the method proposed by Ashley in ABoK in fact implies that the knot already has his form before it is tight,without this described movement ),but if the knot is already seriously jammed, I think this attempt is irrelevant(I tried a end loop version of this bend, it can jam in an even worse way!).
I also agree that the additional lines in those diagrams are very ugly!

                                                                                                                 Bye!

[Dan_Lehman]

I tried #1425 (for the first time) using Luca's (first) diagram, and to my surprise got a nice facility,
very strong and reliable bend, the only thing I do not like that it is very difficult untied (prone to jamming?).

You should not believe it to be "very difficult
to untie": it is in fact fairly non-jamming, and
easy enough to untie --where "enough" means
that, although it takes some working, the working
is available, in many circumstances.
The method : push the side-by-side *binding*
wraps of the tails away from each other,
and then try pushing them back together WITH
the two SParts; bit by bit, at first, you should be
able to return material from the SParts into the
knot and the tightness is gone.
Oh, you can also try pulling the tails apart from
each other, roughly perpendicular to the axis
of tension, which opens the knot.  (And after
such pulling apart, one might try pushing the
bights that turn around these tails back over
them (i.e., towards their ends), returning
material from the tails into the knot --to the
same "binding" as above.

Which binding is the nice thing, keeping the knot
secure when slack/untensioned.  The knot can be
tied with slack in this binding, so that untying is
all the easier (but giving up the slack security
--something that one might not need).

And question for Dan: why do you think that butterfly has not place among best [end-2-end knots]?

Well, it is an asymmetric knot that allows of various
dressings, and it seems to garner praise for the silliest
of reasons --it's known, and there is rumor of great
characteristics (but unproven).  So, there is more slop
in tying this knot than there will be in symmetric ones,
IMO.  (Although on this forum we had a long debate
about whether #1452 (nb : '52', not '25'!), aka
Ashley's Bend, jammed; that resolved itself when
finally the jamming tyer & I realized he'd tied it in
a way I didn't think would naturally arise.  (That it
can be dressed to jam, I find a *feature*,
not a problem; but one needs to know ... & choose!)   

In fact, I have to acknowledge that (a)symmetry is
no guarantee of strength characteristics; indeed,
there might be versions of the butterfly knot that
prove nicely strong (there have been some good
results, but we so seldom can tell how anything
was actually dressed & set that ... who knows!?).
There is a form in which one side forms what can
be regarded as a minimal timber hitch shape
and the other a "pretzel", and in this case it seems
good if the former nips its tail directly --then, both
overhand components look to be in good shapes,
even though different; maybe both are strong!

But the question returned is how one can come
to such apparent adulation of that popular knot
and not have equal or higher regard for Ashley's
#1408, 1425, & 1452?  And why!


--dl*
====

[Dan_Lehman]

Hi Dan,

A quick method for tying  #1425 :
this method in some way is a bit similar to the method shown in ABOK
for #1425a/Hunter's bend ...

You really can't believe this, can you?
This is anything but "quick"!

I can :quick,fast and rapid;and I would add also easy ...

I remain convinced otherwise!   

In what you present, a tyer
(1) must work both ends simultaneously,
(2) maintain a curved crossing --i.e., left then back right--
of the two SParts,
(3) wrap the two tails around just so
with regard to (2),
(4) and tuck them out in a particular way,
(5) and then get it all to capsize & dress nicely!
.:.  I call bulllloney!     ( --or "magician"!)

... damn,I thought I made ​​a small gift for you!

Oh, then I am thankful (that it was not a BIG one)! 

... by Ashley for this very good,but "not pratical", "decorative" bend
...at this point I should ask you if you have a better method!

Ashley's assessment is no better than Harry Asher's
of the version of SmitHunter's bend in which the
tails cross --a simple change, there, that both improves
strength (my weak surmise) and ability to be loosened
(my surer observation).  IMO, there are reasons to use
this knot; and it leads to a quickly tied-in-the-bight
eyeknot.  His notion of number of crossings is bunk
--who cares, really.  It ISN'T easy to remember, for
a casual user; but for someone who might find it
good for tasks, the learning isn't so bad, et cetera;
not to gain the slack security aspect.
(And it should have led him to variations!)

As for tying, I admit to having a mental block on this,
recently --had a rough idea of how it went, but ... .
But, I'd simply (practise, practise...) do as I do for
all such interlocked-overhands knots (usually;
some eyeknots are potentially "TIB") : form one half's
overhand component and then reeve the other
end through it according to which knot is desired.
(And here I'll note that Ashley is ambiguous on the
dressing(s) available for #1452, unlike his noting
issues with #1408.

I'm glad I was however in some way a source of inspiration!
But I must confess that I did not understand what you mean by
 "#1425 as a bight hitch ".

Firstly, "bight hitch" is an idea of mine regarding
those end-2-end knots in which one end forms
a bight (e.g., the venerable sheet bend --which,
as best I can tell, was originally a hitch to a sail's
"clew" (roundish reinforced hole), and then adapted
to join lines).  Especially where one rope is larger
than the other, it really seems that one is tying
the thinner to the larger, more than some equal
joining, hence the invocation of "hitch".

You're diagrams showed bights at one point, and
I just went along with that thinking of trying to
work out something quick in tying to a bight;
I could tie to it with an overhand pretty quickly,
but not so quickly then turn the bight into the
matching component!
(And this is an example in general of a problem
I face when going with good intentions to convert
a tied-in-rope record of a *new knot* into ink-on-paper
record : I see some "what if ... " possibility, and instead
of freeing the first rope, I end up now with TWO ropes
literally "tied up" and unable to be used, until ...
that inked record is made.  .:.  most frustrating!    )


--dl*
====

[xarax]

...it leads to a quickly tied-in-the-bight eyeknot.

...bulllloney!     ( --or "magician"!)

[Dan_Lehman]

...it leads to a quickly tied-in-the-bight eyeknot.

...bulllloney!     ( --or "magician"!)

Yes, okay, I owe you.
But go back to that "symmetric fig.9" end-2-ender
(by which I mean using Ashley's #525 stopper form)
which I equated to THIS #1425 interlocking of two
overhand (~= 9-1  ) and you advised me not to try
to win others to this notion of equivalence !
The eyeknot makes half-way redress of my casual
*equivalence* in that the SPart's through path to eye
leg IS a symmetric fig.9, leaving only the tail
to have a convenient abbreviation to overhand.

In the attached photo of the (snugly set) stopper form
of #525, consider the left end as the S.Part and
segment-1, flowing anti-clockwise into (next visibly
delineated) segment-2a..-2b ("b" end of segment-2
diving back through a loop), and down to the bottom
segment-3, which penetrates the loop formed by 1-2
to continue as segment-4 and exit into right-side eye.

For the "TIB" eyeknot(s), form 1-2 and then
make a bight of the working end and lay the tail leg
alongside the span from s-2 to 2-3 (so, vertical, given
photo's orientation).  NB : 3 options occur here :
the tail leg can be to the left, or to the right & in
front of 2b-2a, or ... behind; in all cases, the tail leg
is adjacent the S.Part part 2b-3, vertically oriented.

Now, simply complete the tying of the shown single-strand
stopper but using this just-formed bight.  The tail-side of
the eye flows into an overhand and the S.Part's full
passage into the eye forms the fig.9 ; I regard #1425
as capturing the *spirit* of this general knotted geometry,
of the loaded part passing though those "binding wraps"
and up around parts, and not pulling directly against each
other.  (One could of course tie #525 with a bight and so
have a fully "sym.fig.9" eye knot; but I see benefit to
keeping the tail side from making the full journey.  It
can even be that the tail's path is further shortened,
to a mere turn, but ... .   .:.  horses for courses.

--dl*
====

[Dan_Lehman]

Why did I post my image?
The orange & white #1425 by Xarax serves better!
Again, simply see, e.g., the right side as where the
eye will be, and then have the orange tail trace
farther along the white strand out into this space
(fusing itself to the white to close the eye beyond
the photo's reach).  And the 3 versions can be
seen as keeping that white tail left of the orange
or moving it on either side of the orange strand
to position it just to the right.

The goal of this eyeknot is to be easily tied (esp.
easy when TIB) and secure when slack AND easily
untied.  I found that when loaded to break force
in HMPE cord that it was no longer easy to untie
(I have not tried, as I want to preserve the surviving
knot --opposite one in test specimen broke--, but I
can see that it will be --at best-- (very) hard to loosen;
in HMPE, force flowed extensively along the SPart
and drew that binding wrap extremely tight!)


--dl*
===

[xarax]

   If THIS is how ABoK#525 leads(sic) to something, anything can lead to anything else ! 
   I do not see any significant saving, in relation to the amount of the material required/consumed, in anything "abbreviated"/ "simpler" ( but, in fact, conceptually, much more complex ! ) than the full retraced ABoK#525. However, this double-line ABoK#525 nub can be dressed in many ways ( as all those knots where the lines of each link follow adjacent and parallel paths, so that they can be twisted = rotated around each other ), and its bulk is 100% offset, regarding the loading axis the eyeknot. I do not like this unbalanced - to say the least - form ( a fact mentioned by Ashley, too : "the stem is a bit off center").
   A "similar", but more axially balanced eyeknot, generated by retracing a symmetric stopper ( ABoK#582 )( "similar", in the sense that we retuck the bight through the nub of the basic overhand knot stopper twice, in order to make the knot less prone to jamming - and, possibly, as a by-product=bonus, increase the diameter of the Standing part s first curve, too ) is shown at (1).
   
  ( I repost the picture of ABoK#525, and of the symmetric stopper ( ABoK#582) , the retracing of which generates the "Full" loop shown at (1) - and a detail of a picture of transformations of the fig.9 knot, shown at (2), to help - so to speak...- the reading=deciphering of the previous posts... . See also (3)(4). For further inquiries, visit (5))

1. http://igkt.net/sm/index.php?topic=4095.msg24597#msg24597
2. http://igkt.net/sm/index.php?topic=3838.msg22777#msg22777
3. http://igkt.net/sm/index.php?topic=3838.msg25931#msg25931
4. http://igkt.net/sm/index.php?topic=3838.msg25938#msg25938
5. http://www.holidayinn.com/hotels/gb/en/san-diego/sanmm/hoteldetail?qAdlt=1&qChld=0&qRms=1&qIta=99617383&qPSt=0&qSmP=3&qWch=0&qSHp=1&qBrs=6c.hi.ex.rs.ic.cp.in.sb.cw.cv&qSrt=BRAND_SORT&qRpp=25&qRRSrt=rt&qFRA=1&srb_u=1&icdv=99617383&sicreative=21152247964&sicontent=0&siclientid=2038&sitrackingid=428967873&
     

[Luca]

Hi Dan,

I remain convinced otherwise!   
In what you present, a tyer
(1) must work both ends simultaneously,
(2) maintain a curved crossing --i.e., left then back right--
of the two SParts,
(3) wrap the two tails around just so
with regard to (2),
(4) and tuck them out in a particular way,
(5) and then get it all to capsize & dress nicely!

1) I'm beginning to suspect that i'm using a wrong method  when I run my slipped Reef knots to tie my shoes ...
2) In order of the successful realization of the bend,is not important to keep this crossing between the standing parts that is created when the bights are placed one above the other;at the end of the tying, the standing parts can safely exchange the right / left directions(with respect of as shown in my diagrams) in which to be pulled in order to properly tighten the knot.
3) See 2)..
4) It does not seem very different from how it is done using the "b&q" method for the Zeppelin or the Hunter's.
5) I guess would depend on the material and the greater or lesser rigidity and/or the diameter of the rope used: in my experience all capsizes and dresses nicely and the knot assumes a compact form (almost without any slack ..) in a spontaneous way,only by pulling on the tails.
                                                                                                     

 
I call bulllloney!   

Uh..thanks!   NOW I know that I'm really a Sr. Member!

(or "magician"!)

There are methods to tie knots that are more respectful of the real structure of the final product, and that better help to understand this structure,and also help to create a sistematic basis from which to realize all the possible variations of a given model (OK , I explained badly: let's say your method to achieve the four variations of the DL locked Bowline and the three bights method by xarax are examples of what I'm trying to explain).
These are the methods that xarax, following what he writes here ( http://igkt.net/sm/index.php?topic=4698.msg30357#msg30357 ), perhaps would define as "natural".The method proposed by me for # 1425 is perhaps not very educational/explanatory for the knot tyer in this sense, and therefore can be considered as "magic";in the sense of: "Et voila!", and the audience is surprised without understanding what has actually happened..
This is surely a limit of this method (but a certain systematicity is still applicable here as well,;after all, the method is a variation of the "magic" method for #1425a described in the ABoK (which I find to be susceptible to errors(that can lead to different knots/variations)to a greater extent with respect of the method in question..)

As for tying...I'd simply (practise, practise...) do as I do for
all such interlocked-overhands knots... : form one half's
overhand component and then reeve the other
end through it according to which knot is desired.

These things relate to the private sphere of the knot tyer .. but confidentially I tell  you  that,with regard to these interlinked Overhands-based bends(when I'm alone,and I have not seen by anyone( )),personally I act as you describe only if I am forced (mostly in the case of the realization of end- loop versions of these bends(which, this is true, implies at least two methods for each bend to learn ..)).

You're diagrams showed bights at one point, and
I just went along with that thinking of trying to
work out something quick in tying to a bight;
I could tie to it with an overhand pretty quickly,
but not so quickly then turn the bight into the
matching component!

Thank you, now I've got an idea ..or maybe not?(In his drawings Ashley perhaps has not  proposed something like that?)

And this is an example in general of a problem
I face when going with good intentions to convert
a tied-in-rope record of a *new knot* into ink-on-paper
record : I see some "what if ... " possibility, and instead
of freeing the first rope, I end up now with TWO ropes
literally "tied up" and unable to be used, until ...
that inked record is made.  .:.  most frustrating!    )

I thought that only I had this problem! (Given the scarcity of ropes that I own!)

                                                                                                                                Bye!

[xarax]

There are methods to tie knots that... help to create a systematic basis from which to realize all the possible variations of a given model

   Unfortunately, knot tyers ( well, most of them...) are not interested in anything "systematic"... I have tried to show how, starting from a reef or a thief knot or a Carrick mat "base", one can tie many interlocked-overhand-knot bends, some of them being unknown, but in vein...(1)(2)(3)

1. (retucking the Reef knot )                 http://igkt.net/sm/index.php?topic=3204.msg19380#msg19380
                                                          http://igkt.net/sm/index.php?topic=2826.msg19395#msg19395
2. ( retucking a particular Carrick mat ) http://igkt.net/sm/index.php?topic=3086.msg18601#msg18601
3. ( #8-shaped links )                          http://igkt.net/sm/index.php?topic=3148.0
4. ( retucking the Thief knot )               http://igkt.net/sm/index.php?topic=3611.0

   Now, here is an interesting challenge for the young and/or senior member : With the one rope, form the overhand knot shape, in all its four variations ( this shape has 3 crossings, so, starting from the first one and keeping it unchanged, one can tie 2 x 2 different Pretzel-shaped "bases" - one of them is the overhand knot itself, and the other three are Pretzel-shaped bases topologically equivalent to the unknot ). Then, with the end of the other rope, trace all the different paths through the three openings of this shape : enter into any of them from any of the two sides, exit from the other side and then enter into another, and so on.  WHICH of those paths generate symmetric bends ? How many are they ?
  Of course, this is half the required job... , because the overhand knot itself, and the overhand knot shape, in general, can have the other, nice, symmetric #8 - shaped form ( shown in the attached picture ). The procedure described previously should be applied to this shape, too. It has 4 crossings, so the possible variations of the first "base" are 2 x 2 x 2 = 8. The end of the other rope has now four openings, so the total number of all possible paths is bigger.
  THAT is a systematic basis to realize all the possible variations of the "overhand knot shape" model...

[Dan_Lehman]

   If THIS is how ABoK#525 leads(sic) to something, anything can lead to anything else ! 

One might posit that as a theorem (if A-thing then A-else),
but, really, this particular "symmetric fig.9" form --and in
an older thread on this site, I give the equivalence between
#521, 525, and another-- surely presents a 2/3 of itself
the common overhand form.  And it does so just as
does #1425; the continuation that one would give to this
end-2-end knot to completely realize the #525 components
is simple, by a tracing.  (Now, Ashley's image is less than
compelling; and #525 is loaded asymmetrically, qua stopper.)

I do not see any significant saving, in relation to the amount of the material required/consumed, in anything "abbreviated"/ "simpler" ( but, in fact, conceptually, much more complex ! ) than the full retraced ABoK#525.

Then it must be by choice (that you don't see)!
The abbreviation in the eyeknot --which has but
*half* of the full #525-- enables the trio of positions
of the tail's part, and in all cases omits need to figure
how to trace farther along --which additional extent
would contribute little to the knot (but would present
the tail for simple binding (tape, hog ring, seizing)
to the S.Part).

And you do note the possibilities:

However, this double-line ABoK#525 nub can be dressed in many ways

but go wrong re "offset" --that "binding wrap" part
surrounds the axis of tension.

and its bulk is 100% offset, regarding the loading axis the eyeknot.

Further, you seem to be confusing stopper & eyeknot:

I do not like this unbalanced - to say the least - form
( a fact mentioned by Ashley, too : "the stem is a bit off center").

The stopper, btw, made in e.g. a hand-over-hand climbing
rope (exercise, otherwise), or as a placement marker in
a dockline (I've seen such nub uses), has the advantage
of being non-jamming, and possibly (?) stronger than
the common overhand.


--dl*
====

[xarax]

 

...this particular "symmetric fig.9" form...

  You had presented more than one "symmetric" form of the fig.9 "stopper" ( see the attached picture, for the most interesting of them ), but, personally, I do not find any of them very stable ( i.e, very suitable), as a "basis" of an eyeknot - in comparison to the topologically more complex, that is true, but geometrically more symmetric ( so, in a sense, more "simple" ) ABoK#582 "stopper". Ashley missed the monumental opportunity to utilize this "naturally" symmetric "stopper" in the case of a bend ( the Zeppelin bend, of course - perhaps the biggest mistake in his knotting life, IMHO ), we should not retrace his omission in the case of an eyeknot - although the derived bulky/portland knot (1) can not be compared to the Zeppelin beauty.

  The abbreviation in the eyeknot --which has but *half* of the full #525-- enables the trio of positions of the tail's part, and in all cases omits need to figure how to trace farther along --which additional extent would contribute little to the knot ...

  I do not say that it there is no abbreviation, I say that this abbreviation is : a, NOT 50% ( so, not *half*, with any number of asterisks...), and, b : it is not worth the conceptual complexity the asymmetry on top of an already naturally-asymmetric stopper adds, compared the conceptually simpler fully retraced ABoK#525. On the one hand you do subtract some parts of the knot that are not very important to its working as an eyeknot, indeed, but on the other hand you add complexity, in the form of the attention during tying and dressing, compared to the simple / dumb task of just tying a full double line = TIB ABoK#525 eyeknot in no time. ( Because I suppose we are interested in tying this eyeknot as TIB, most of the time, so we do not need to "figure how to trace futher along".) Moreover, I claim that the retraced ABoK#582, although topologically more complex, is nevertheless geometrically more symmetric ( so, more "simple") and thus more suitable to serve as an eyeknot .

...you seem to be confusing stopper & eyeknot:

I do not like this unbalanced - to say the least - form
( a fact mentioned by Ashley, too : "the stem is a bit off center").

  I used the word "stopper" to denote the middle line knot which is loaded from both ends / sides that has this topology, not the end-of-line form which is loaded from the one end /side. How else can one name such a knot ? I was just talking about the TIB eyeknot made by a double = two line "stopper", or by a retraced "stopper", in that sense. Whatever happens to this midline "stopper", happens to the eyeknot derived from this "stopper" - if the bulk of the material of the former is offset, regarding the axis of loading, the later is offset, too. In other words, the more symmetric, carefully dressed forms you had presented of the ABoK#525 are not very stable, and the less symmetric, "natural", self-dressed forms are not symmetric enough to serve as a 'basis" of an eyeknot. Tie the retraced ABoK#582, without even thinking of similar "abbreviations" ! ( which are tempting there, too...) .

  { I have a folder, in my computer, labelled "Beautiful knots". There are some very simple and slim knots there ( the Tumbling Thief, the Zeppelin and the Double Harness bends, the Double Locked and the Single Locked Cow hitches, the Gleipnir binder, the simple and the double simple-hitches-a-la-Gleipnir, the TackleClamp hitch, the 8-8 loop, the Helical Constrictor loop ), but there are also some more complex and not-so-slim ones ( The Oyster, the Strangle and the Illusion bends, the Lee s locked, the Alpineer s TIB, the Lee s C, the Tweedledee and the double=two collar Water bowlines, the pseudo-Zeppelin loop .) I had recently added the retraced ABoK#582 eyeknot, as shown in (1) - although it will never loose the weight it should, I am afraid... }

1. http://igkt.net/sm/index.php?topic=4715.0

[xarax]

  I think that we should had expected the instability of this particular symmetric form of the Fig.9 "stopper", given that the corresponding ( I claim I "see" a genuine correspondence here...) symmetric form of the Fig.8 "stopper' ( the Pretzel-like one, shown at the attached picture, generated by a symmetric cut of the closed 4.1 knot ) is also not very stable. After all, as I had noticed at (1), there is a unexplained pairing between the Fig.8 and the Fig.9., then a gap, which should not be there !  ( and so it should had been expected that it was not expected...) and then a pairing between the Fig.10 and the Fig 11.
  A first explanation for this instability was attempted at (2) - where one can also see a picture of the retraced Fig.9 knot. 

...this knot is not very stable ( so it seems not very suitable as a basis of an end-to-end or an eye-knot ), because the oblique / diagonal element that connects the two linked bights / the two sides of the knot... does not stem out from / is not attached at the centre of the nub, so it can slide on its surface, and be transported towards the one or the other side.

   When this oblique / diagonal element passes straight through, and in between, the two interlinked overhand-knot-shaped parts of the knot, as in ABoK#582, and not around them, on the surface of the nub, as in ABoK#525, this problem disappears.   
 
1. http://igkt.net/sm/index.php?topic=3838.msg25911#msg25911
2. http://igkt.net/sm/index.php?topic=3236.msg30470#msg30470

[NotSure]

   I suspect that the so-called "Well Pipe hitch" ( a grandiose name for a mediocre knot ) is the less clever hitch there can be in this Universe : just many wraps, and two half hitches at the end : an ingenious knotting "solution", indeed !    However, I am not sure about this. What I am sure about, is that with 12 wraps and ANY means of connecting the Tail end to the Standing end, there can be NO hitch that will fail to do what the "Well Pipe hitch" - or, for that matter, ANY other known hitch - can do... How one can be "a king", when the job he does could have been done by anybody in his kingdom ? On the other hand, that explains why most kings, once they grip the sceptre, they succeed to not let it slip through their hands - their job is as easy, and perhaps easier, than of any other subject of their kingdom ! In other words, with 12 wraps, there is no hitch that will let the Pipe fall into the Well ...
   Most knotting problems either they have many absolutely correct solutions, or they have none - and bends, i.e., end-to-end knots, are no exception to this rule. Now, I believe the following is a legitimate question to ask : If, in some cataclysm, all of knotting knowledge were to be destroyed, and only one bend passed on to the next generation of creatures, which bend would it be ? I believe it should be the Zeppelin bend.

You don't need 12 wraps for the Well-pipe Hitch, 2 or 3 will suffice, LOL. Obviously though, the more turns you use the greater the friction gripping power will be.

I like the Well-Pipe Hitch specifically because it is entirely composed of simple wraps. This maximizes the surface area for friction contact as opposed to a cross-lashing or "cross-gartering" scenario (which raises a significant portion of the rope away from the surface to be gripped). See this thread on the KC Hitch here for an extreme example on the cross-lashing effect:

http://igkt.net/sm/index.php?topic=542.0

And knot4u's successful rebuttal to it here (that is, if you were to test each hitch with the identical length of rope used to form them - meaning more friction turns and surface contact will be available for a non-cross-gartering style of hitch):

http://igkt.net/sm/index.php?topic=542.msg17072#msg17072
AND http://igkt.net/sm/index.php?topic=1889.msg13675#msg13675

And yes, xarax, you could indeed use any knot to finish it off. In fact, most of the time I'm inclined to use my favorite finisher, an inverted Sailor's Hitch tied to it's own standing part (instead of the 2 half hitches/clove hitch to finish it). Which brings me to my next thought:

While a hitch would suffice to join 2 greatly dissimilar sized lines together, it is also mostly a uni-directional affair and not terribly secure when loaded in different directions or when loading the tails. For a very secure connection, 2 loop knots interconnected will do, OR 2 anchor hitches.

Check it out! When 2 Sailors Hitches (inverted to Clifford Ashley's portrayal of #1231) are linked together, it makes a very secure, 4-way loadable, and super, super easy to untie connection regardless of any extreme loading applied and/or wet/frozen rope/fingers.

On a side note, I must say that I try very hard to avoid tying anything of a "Zeppelin/Rosendahl" or even "figure 8" type of knot whenever possible. I find those knots are incredibly hard to untie when sufficiently loaded. There are almost always better options for the task at hand. The ONLY exception to this preference of mine would be for when I have trust issues and want to inspect (from 50 feet away ) a knot that somebody else tied for me....

[roo]

On a side note, I must say that I try very hard to avoid tying anything of a "Zeppelin/Rosendahl" or even "figure 8" type of knot whenever possible. I find those knots are incredibly hard to untie when sufficiently loaded.

The Zeppelin Bend shouldn't be difficult to untie after a hard load.  If you need a bend that can be untied while the line is still under heavy tension, you might try a Trigger Bend.