I learned this knot as the GRB - "Guide's Rappel Brake" and also as an autoblock. There are several entries each year in "Accidents in North American Mountaineering" that describe failure to properly use a mechanical belay device. I recall no accidents as the result of a faulty autoblock knot. This knot is aligned with the KISS principle from software engineering. It is remarkably simple and effective.

Interesting to read this story of its origin.

Interesting that you mention "Accidents in North American Mountaineering". This is the climbing equivalent of the Startup Post-Mortems story that was posted the other day. They publish them every year and they're a great read. Like startup failures, they tend to cluster around a few common mistakes--typically failure to anticipate weather conditions and overconfidence in one's abilities. Definitely worth checking out if you enjoy climbing and mountaineering.

The KISS principal is used outside of software engineering, I believe it was popularized by the US Navy.

It's called an "autoblock" by climbers today in the US: https://en.wikipedia.org/wiki/Autoblock

That appears to be a slightly different knot. The final stage where the bottom loop goes through the top isn't done for an autoblock.

> The final stage where the bottom loop goes through the top isn't done for an autoblock.

That'd be the french knot which only blocks in one direction. The Machard knot which blocks in both directions does not loop through itself, the karabiner goes through both loops to close the knot.

Sailors have known and used "jamming" hitches like this for centuries. It's the basic principle behind a sheep shank for example. Marseilles is of course a port city, so I wonder if he found some inspiration along the waterfront.

This was not the first jamming hitch used in climbing--that would be the Prusik, which is still used today.

The Prusik is trickier and slower to tie than the autoblock, and tends to stay tight once it is tensioned. The autoblock runs looser and is easier to loosen up once it has been tensioned, which is why it's generally preferred over the Prusik as a rappel backup.

true, but there are also very nice climbing locations in Calanques just off the Marseille city (among other). one of most beautiful places to do climbing i've seen so far (climbing white cliffs that don't heat so much with sun above/nearby dark turquoise waters beneath).

An alternative to this is the mechanical ascender / 'jammer'. I've used the Petzl version when caving: http://www.petzl.com/en/Sport/Ascenders

They're practically bullet proof, although they can take a little longer to bite when coated in mud :)

As a third hand for your rappel, a toothed ascender seems like a bad idea. Petzl makes another device called the Shunt which is more suitable to third hand use. Although in the case of icy ropes, maybe teeth are a hard requirement.

The Shunt is great, but there are some potentially dangerous gotchas in its operation. Never use it with ropes of different diameter. As a rappel backup, the shunt can fail if you place it above your belay device. The risk is that a panicked operator will hold the device open. In the correct mode of operation, you place the shunt below your belay device. If you're disabled, the shunt will lock off. If you panic and hold it open, the force you apply to the device will arrest your rappel.

Yes, it's applications are very specialized, this is true.

I've only used ascenders when climbing up free hanging rope, and they work very well for that purpose (I've done several hundred meters of vertical ascent in a single trip using this technique).

Oh, I remember using the Braid variant even it's a little more difficult to make with one hand.

Cool. Never used it before. A prusik seems easier, especially since I learned the one-handed technique. I wonder what the trade offs are of each.

I think autoblocks are significantly easier, especially since you never need to unclip the autoblock from your harness, limiting the chances of you dropping it. The hold is also much looser than a prussic, which is an advantage when used as a rap backup and probably a disadvantage when used for ascending.

I've found it much more friendly as a third hand. I haven't done extensive testing, but I have found it feeds more freely and is much easier to release or mind, while still being happy to bite when you want it to.

A prusik is still my go-to for most other friction knot applications.

One difference between the two it that the Prussik is much easier to release after it's been jammed.

I'm curious as to how this (which I know as a "French Prussik") could possibly be jammed. IME, and that of the guides who recommended it to me, this is far less likely to cause jamming problems than a traditional single or double prussik.

For ascending, I believe their recommendation was to use a traditional single prussik for the harness, and this for the leg loop. This is certainly easier to move up an untensioned rope than a traditional prussik, but may also slip slightly more readily, I think.

Personally, I've used it mainly as a backup when abseiling; I can't understand why it's not universally used there (attached to leg-loop and below the abseil device) - you just keep your hand on it and it will allow descent, let go or allow it to move up and you are stopped.

If a knot is the only thing between you and a very long fall, "easier to release" might not be a feature.

Not true. It is called the autoblock for a reason. You tie it by wrapping it around your rope. To descend, you simply apply a slight amount of downward pressure to the top of the knot. This is the easy to release part. If you remove that pressure, by just moving your hand off the knot, it instantly grabs and stops all downward movement. A Prussik will retain most or all of it's tension, making it difficult to use in ascending or descending. I only use a Prussik in a situation where I want no movement at all for the duration, such as a backup to an anchor.

This knot seems to work by friction. Is there any knot with the same application that works without friction?

All knots work by friction. Try to imagine any knot working with friction-free rope (which doesn't exist).

> Try to imagine any knot working with friction-free rope

Yes that was my question :)

Ok, well in that case the answer is 'no'.

You mean: "I can't imagine it". That is something else :)

I can imagine lots of stuff that doesn't exist (interstellar spaceships with earth as their origin, pink elephants wearing tutus, time machines and so on). But when I limit my imagination to those things that can be constructed using the laws of physics as we currently know them knots without friction drop right off the menu along with the items above.

So if your question was serious then I consider it answered, if you want to seriously know whether something that can't exist can be imagined then the answer is yes, you can imagine it (and so can I), but I can't imagine making one because that would violate the laws of physics somewhere along the line.

You could tie a knot in frictionless rope easily enough, but it would come apart without any work at all and it would not hold anything and you wouldn't be able to grip the rope on either end to tie the knot in the first place. Friction and knots go together unless you're living in a mathematical ideal world, which we don't.

You didn't ask if it's possible to have a knot perform the same tasks (break an uncontrolled descent) without friction, you asked if there were any examples of such a knot.

Is it possible? I can't rule it out.

Are there existing knots that do so? None that are commonly known, no.

Can it be that I've seen fishermen use the same to tie a hook to a line?

If they're tying it to the end of the line, they're probably using the improved clinch knot, which has visually similar wrapping: http://www.netknots.com/fishing_knots/improved-clinch-knot/