International Guild of Knot Tyers Forum
General => Practical Knots => Topic started by: xarax on September 01, 2010, 04:01:48 PM
-
http://sites.google.com/site/puripant/research/knot-design
-
Okay I like this. I'm someone who mentally visualizes geometric shapes and I'm pretty damn good at it. Knots, however, sometimes can drive me crazy when I try to visualize them. So something like this I like very much. When is this available for the public?
-
There are only two or three knot design programs like this that I am aware of, - programs that can achieve a 3D knot visualization AND some degree of self-dressing and self-tightening of knots as well - and none of them is commercially, or even academically, available at the moment, as far as I know. We have to persuade the authors that there is some willing audience or market niche waiting for their products, and it seems we have not been very successful in this yet... :) I think that people in knot forums are a little more conservative in their attitude towards practical knot designing than they should be, and they continue to draw and examine knots more or less the way Ashley did, decades before the advent of the modern computer tools...
Why would I settle for a computer program that makes unrealistic reproductions of knot structure and tightening when I can just tie the knot in real rope? The real thing is cheaper, quicker, clearer, and there's no interface to learn.
-
This software seems to be useful. I've discovered that everybody learns differently. For example, I found out on this this site that at least a few people gave up on the Jug Sling because they think it's too complex. I think it's just a learning issue for how their particular brain operates. Currently, to learn a knot, someone must be able to understand a pic, a video or a written description (yuk!). If you can't learn a knot by any of those methods, then you won't be learning the knot.
This software provides yet another method for learning. If you don't need another method for learning, then that's good for you. Meanwhile, it should be easy for you to understand that others may need another method. The trick is convincing people that this software is useful.
(On the other hand, it seems like programming a relatively complex knot into this software would be a beast. :-\)
-
Has Puripant Ruchikachorn made the software available? He mentioned in his thesis it was a Java programmed application. Does he have a website with this available for use? I can see this as a way to build tutorial sketches for those of us who are challenged to draw circles even with a tin cup.
Matt
-
Agreed! I'd love to get ahold of this software. would make learning things, and planning things much easier.
-
Has Puripant Ruchikachorn made the software available? He mentioned in his thesis it was a Java programmed application. Does he have a website with this available for use? I can see this as a way to build tutorial sketches for those of us who are challenged to draw circles even with a tin cup.
Matt
It seems to be a recurring theme in these forums to look for an awkward, homemade software solution when a hardware solution is called for. Even if you can't draw a circle in MS Paint with the Shift button pushed, you can probably work a scanner or camera just fine.
The only software requirement might be for emphasizing or tracing rope edges and removing extraneous texture to turn an unclear knot image into a clear knot diagram.
All of this is currently available.
-
I wonder what roo would have said if he was one of the first to witness the discovery of the car..."
You clearly misunderstand my position on this. I view the software in question as a marked step backwards in terms of accuracy, realism, ease of use, and presentation.
-
I view this thread misplaced from Knot Theory & Computing !
:(
-
You clearly misunderstand my position on this. I view cars, in relation to horse carriages, as a marked step backwards in terms of ease of use, clear air, noisy cities, parking space, Gulf oil spill, Iraq war, greenhouse effect, climate change, not to mention rising insurance premiums... Smiley
As expected, this is going nowhere. Let me know when you find software to automatically dress up a knot photo or scan to automatically trace over rope edges and remove texture.
I already have the ideal knot modeler in actual rope, it's the image export functions that could stand some automation.
P.S. I agree with Dan on topic placement. I assume that he's already alerted a moderator.
-
I'll take it from the derailment that the software isn't available. I'm jealous of those of you that can sketch things out .. I can use a scanner, but to capture a knot in process with a flatbed scanner would be interesting without the knot settling to the glass. Also, I like the idea of rotating the image in space like they did in the demo.
I appreciate the help if anyone knows if this is available, and I'll track the thread in the theory forum if it moves.
-
Here's where this program excels over pics of real rope: The program seems to allow easy editing of knot diagrams.
I've tried to teach a knot by taking a sequence of pics of real rope. Somebody inevitably responds and needs, for example, a pic between step 2 and step 3. Well, at that point, I have to go back into my garage, get some rope, pull out my camera and reconstruct the whole arduous process. Meanwhile, I have a feeling that most people don't even bother trying to figure out my pics. By the way, I'm no good at hand drawing knots on paper.
In contrast, this software allows me to be on a train, or wherever, and teach a knot with animation and without having access to a rope or a camera. That's just one example of this software's usefulness. Generally, I don't like stifling innovation.
Many of the best inventions did not have readily identifiable value at their origin. For example, Ronald Wayne relinquished his 10% share of Apple for $800 in 1976. :-[ I'm not saying this software is like Apple. I'm saying innovations don't have the benefit of hindsight.
-
Here's where this program excels over pics of real rope: The program seems to allow easy editing of knot diagrams.
I've tried to teach a knot by taking a sequence of pics of real rope. Somebody inevitably responds and needs, for example, a pic between step 2 and step 3. Well, at that point, I have to go back into my garage, get some rope, pull out my camera and reconstruct the whole arduous process. Meanwhile, I have a feeling that most people don't even bother trying to figure out my pics. By the way, I'm no good at hand drawing knots on paper.
In contrast, this software allows me to be on a train, or wherever, and teach a knot with animation and without having access to a rope or a camera. That's just one example of this software's usefulness. Generally, I'll never be one to stifle any kind of innovation.
Pictures have the same problem that this program does: The edges of the ropes are just weak and soft transitions from one texture to another. There needs to be black lines to make the rope stand out, and the texture needs to be deleted as it just hinders the eye.
To me, it seems much easier and faster to reconstruct a missing step with rope, rather than opening up a program that creates unnatural-looking geometry with an interface to learn that may or may not even have a selection of objects to supplement the diagram in the case of a hitch, for example. After that, images from either source would still need to have texture stripped, and edges enhanced, so there's no advantage either way there.
If edge enhancement & detexturization is accomplished, it could be applied to rope photos and scans as well.
-
What I wish, and the above mentioned article promises to deliver, is a more intelligent 3-D software program that "understands" the physical dimensions of the rope and tightens the knot...
Rope "understands" this. If you pretend that there's a microchip in your rope, would you use it?
-
I agree with roo... This is fundamentally, a solution looking for a problem.
The human hand, coupled with our advanced cerebral cortex, allows for the
manipulation of a line of cord, the recognition of esthetic and practical patterns,
and the creation of new knots and ties.
And... Our (human) machine works just fine, thank you.
JD ~ Two Cents
-
Perhaps this software could lead to the development of accurately calculating stress points in a knot. That would be really cool. Currently, we experiment, which is the ultimate answer of course. However, it would be nice to have a full understanding of the "why" when a knot breaks. More knowledge about stress within fishing knots interests me. By the way, xarax, that's a good point about the difference between raster pics and vector pics.
I'm disappointed that this software is met with such dedicated resistance on the International Guild of Knot Tyers Forum. I would prefer to think people here would want to encourage innovation in this ancient discipline of knots. It's depressing to be around people who stubbornly think "nope, we're all set, nothing new to learn".
-
I consider knotting an act of artistic expression.
I'm not sure why we'd want to turn such a thing
over to a computer program.
Then again, art appreciation is on the downturn
(nationwide), so I suppose handing over the
genesis of art to computers is the next step.
JD ~ ((sigh))
-
I don't understand how it's possible to read the posts in the "Practical Knots" forum and come to the conclusion that knot tying is purely art. I see knot tying as being about half art and about half science. There is no turning over anything over to a computer if you don't want. If you want to reject the new thinking that a computer program can facilitate, then you are free to continue to do so.
-
I consider knotting an act of artistic expression.
I'm not sure why we'd want to turn such a thing over to a computer program.
Because you're not turning it over to a computer program. The computer program is a tool in the hands of a person. Sometimes that person would be an experienced knot tyer, sometimes a beginner.
Consider the possibility that you could configure a knot with a computer program, tell it the dimensions of your cord and have it tell you how long your cord needs to be.
Consider the ability to pre-design compound knots and be able to experiment with the joins to be sure that to achieve your design that the cord is continuous and doesn't form unexpected closed cycles.
Not that I expect this program does these things, I just want one that does. 8)
Further don't underestimate the power of abstraction and manipulation. I used to tie my instructions with rainbow cord but decided that edged cord made for a clearer picture. Now to be consistent I need to retie my earlier efforts. If I had had a computer program that would render my instructions it would be a simple matter (if supported by the software) to re-render the images with a different cord type (maybe even rainbow cord with contrasting edging! 8) I have seen instructions where the illustrator uses a cord segment of a different colour for the current step. You can tell when it's a physical splice and not photo-manipulation and the labour involved in those efforts, well... not anything I'm going to try anytime soon. But to add that into a computer program? Much simpler. And if you decide/discover that some particular style or method is better that would affect your entire body of work (NTSC vs PAL vs HD vs 1080p vs 3D or whatever the next new thing is) and you could just push a button to re-render your work in the new format...