per jarnos:
Ashley presents #1074 as a double loop knot, and you can use that as such, if you don't have to tie the both loops around something.
Ashley does illustrate #1074 as double "eye" Bowline - with the 'eye' and the 'bight' set to equal lengths (that is, there is the actual 'eye' and then there is the 'bight'...the bight is drawn through the
nipping loop until it matches the length of the eye). In a sense, the 'bight' is acting a composite 'tail'.
Couple of very important points that need to be made clear:
All written and posted in good faith...
1. Ashley published his book in 1944. Its now 2021 (happy new year!).
So that's almost 80 years ago.
At the time Ashley published his book ('ABoK') he did not conceptualise that #1074 could be employed in mountaineering/climbing applications. He only appeared to conceptualise sailing/yachting applications. We can see this with other knots/hitches presented in his book. For example, #206 Crossing hitch is very useful for belaying a climber - and indeed, its use is mandatory teaching in many climbing training organisations. And #1763 (Prusik hitch) was not conceptualised as being useful to climbers, rescue technicians or tree surgeons (despite Kark Prusik having already published his 'Prusik hitch' several years before Ashley).
Ashley appears to have understood that #1074 is not 100% stable and secure - but, when arranged in a 'load balancing' configuration (with load shared between the 'eye' and the 'bight' via an attachment) - it is secure and stable.
What does this all mean?
It means that Ashley is not the definitive (or final arbiter) for knots and knot applications. Merely because Ashley has published some commentary does not by itself limit or constrain other possibilities.
Ashley presents a knot with same name, #1882, as a single loop knot, uses other loop of the knot than I in the photo, and dresses the knot differently. It is not a PET knot.
2). In #1882, it is the same knot - just applied in a different way. The knot is shown 'upside down' relative to illustration #1074. And all load is now attached to the bight, instead of the eye.
I am unclear why you think #1882 (ie #1074) is not 'PET' (Post Eye Tiable?
By definition, it is still a 'PET' knot.
The depicted application shown at illustration #1882 does not alter/disturb its status as a 'PET' knot.
I would also point out that in the term 'PET', the letter 'E' refers to an
eye (not a loop).
Think of an eye bolt - its not called a loop bolt (its an eye bolt).
Same goes for eye splice - its not referred to as a loop splice,
We would need to alter the term 'PET' to be 'PLT' (ie Post Loop Tiable).
...
With regard to Alan Lee's original post - I very much like Lee's link Bowline.
Although it can be 'fiddly' when employed as a tie-in knot for rock climbing, like all knots, once mastered, it presents no issues to those who are prepared to invest a little time mastering the techniques..and its TIB which is nice. It is also interesting from the standpoint that all rope paths turn around 2 rope diameters within the knot core (ie no sharp bends).
I am warming to the 'Triple bowline' - as presented in Robert Chisnall's variation - and have been undertaking many field trials. In my view, it is superior to #1080 Bowline on a bight (as a tie-in knot for rock climbing).
It has 4 rope diameters inside the nipping loop and is 'TIB' (however, it is not bi-axially loadable as per #1074).