Registered: Nov 2004
Posts: 15

Hi.

I am not a gizmo tech, so I don't know if this is an interesting discovery or not, but I thought I would share it here and see if anyone else finds it amusing.

I have a logitech optical mouse for my laptop. Just now I was working on a paper and had the NKS text on my desk. I picked up the mouse and was using the book cover for a mouse pad, when a malfunction occured. I discovered that the optical sensor in the mouse DOES NOT WORK when the mouse is over the CA printout part of the dustjacket. It acts just as if it were not in contact with a surface at all.

The black surface of the dustjacket works just fine. The yellow and orange triangle portion does not work at all. I tried the mouse on a yellow legal pad, thinking it had a problem with the color, but it works on the legal pad. Not on the printout of the CA!

I also tried it on CA printouts in black and white on the inner pages, and it does work on those. What is it about that ray of yellow and orange triangles that confuses my poor mouse?

Regards,

Richard

__________________
Richard T. Harbaugh,
Nightcleaner

Report this post to a moderator | IP: Logged

Registered: Mar 2004
Posts: 132

My god, you're right, it doesn't work at all.

I did some analysis on pages 214 and 215 and found nothing like what happens on the cover of NKS. I am using a logitech single laser mouse, and it's true, there is absolutely no pointer motion as soon as the mouse crosses over from absolutely any pattern (that I've tried) into that reigon of the cover.

I love accidents.

Report this post to a moderator | IP: Logged

Registered: Nov 2004
Posts: 15

Yes, I tried the patterns in black and white and got no effect also. Just now I tried the inside cover endpapers, which has the same color as the dust jacket and a similar altho smaller pattern...no effect.

But there is an effect on the hard cover under the dust jacket, again only where the yellow/orange ray is printed.

Also I notice that the pointer on the screen sometimes jumps wildly to the edges of the screen when on the yellow pattern, an all or nothing effect.

Cryptographers may be interested.

Richard

__________________
Richard T. Harbaugh,
Nightcleaner

Report this post to a moderator | IP: Logged

Registered: Mar 2004
Posts: 132

Interested in what applications to cryptography there could be.

The only pattern that consistently tricks the mouse besides that cover is contrasting stripes. The pointer jumps as to follow the direction of the lines. This behavior varies with speed and always responds. The pattern on the cover seems a completely diffirent case, where there is no response despite variations in speed.

Report this post to a moderator | IP: Logged

Registered: Nov 2004
Posts: 15

well I'm not sure, as I don't know what causes the effect, but if there are conditions which are hidden from laser scans......

interesting about the stripes. I don't seem to have any striped surfaces at hand. What colors, widths, velocities?

There is a small jumpy reaction of the pointer to movement on the ray. It seems almost random, I cant seem to get it to repeat. It works best with small repetitious motions of about 1cm, and the curser moves slightly but not necessarily in the right direction, then jumps all at once to the edge of the screen, rather chaotic I might say. Larger motions do not seem to have any effect.


Also the cursor seems to work normally when the sensor is near the top of the ray, and there is a rather sharp cutoff point below which it does not work. There is no similar edge effect near the bottom of the book. Maybe it has to do with the width of the ray. Seems that it takes effect when the sensor opening is about the width of the ray, so maybe it is not part of the effect, only has to do with the sensor picking up the edges of the ray.

Richard

__________________
Richard T. Harbaugh,
Nightcleaner

Report this post to a moderator | IP: Logged

Registered: Mar 2004
Posts: 132

This is quite amusing.

Stripes can be found in many varieties in NKS itself, from 1d CA's to the simplist non periodic constraint on a black and white array. My guess is that when Logitech's mice run over a contrasting striped pattern, the internal algorithims that determine distance traveled since the last sample have trouble with an accurate value for distance traveled along the lines – because any speed would look more or less the same.

Because of this, the only diffirences it ends picking up in that direction only just correspond to the random details of the surface our mouse is gliding over. It's kind of like taking a snapshot of fresh snowfall on a sunny day, and then walking some distance and takiing another snapshot at the same angle. For the most part it is impossible to tell from the sparkling surface of the snow just how far one has walked before taking the second shot.

But what makes our mice plow through all patterns execpt our friend rule 110's repetitive background on the cover of the book? Is there something else going on?

Report this post to a moderator | IP: Logged

Registered: Nov 2004
Posts: 15

yes that is the question, but of course it may be some blind alley. We don't know what algorythms logitech mice use. I thought about the problem in a similar way, but I'm not sure the analysis of how the mouse works is correct....I mean it may be, but I just don't have any information.

Why does the mouse work on blank surfaces? For example it works just fine on white or black paper with no visible pattern at all. I have discovered that the mouse will not work on the recorded surface of a CD, although it works on the other side. It does not work very well on my yellow and orange Mathematics of Classical and Quantum Physics book, (Byron and Fuller, Dover Publications 1970).

I would suspect it is the color, not the pattern, which would makes it less interesting from an NKS point of view, except for the fact that the mouse does work on the endpapers of NKS, and they are the same color as the ray, to my eyes.

Picking up surfaces at hand, I find a papberback novel by LaVyrle Spencer (the cover assures me it is a New York Times bestseller, but I have not read it) where there is a picture of the author on the back cover wearing a yellow sweater. The mouse will not work on the sweater, but it does work on LaVyrle's face. There is no pattern in the sweater.

Something about glossy yellow surfaces? And diffraction grateings?

The mouse works on a flat mirror, does not work on a convex magnifying mirror.

I guess we should tabulate this data.

code:

surface     color     texture       pattern        mouse works
mirror      silver   flat          reflective      yes     (makeup mirror)
mirror      silver   magnifies    reflective       no       (makeup mirror)
paper       yellow   glossy       sweater         no       (LaVyrle)
paper       yellow   grainy         NKS              yes     (NKS endpaper)
paper       yellow   rough         NKS              no       (NKS hardcover)
paper       black     rough         NKS              yes     (NKS hardcover}
paper       yellow    glossy        NKS              no      (NKS dustjacket)
CD            silver     smooth      diffraction      no      
CD            silver      smooth     no diffraction yes

__________________
Richard T. Harbaugh,
Nightcleaner

Last edited by Richard T. Harbaugh on 01-28-2005 at 08:13 AM

Report this post to a moderator | IP: Logged

Registered: Jan 2005
Posts: 17

Jesse has likely identified the primary effect - a frequency interlacing, evidencing some resonance between the distance across the pattern, and the speed of 'flicker' in the duty-cycle of the mouse. Akin to timing a machine gun to pass bullets through a propeller, instead of into it, or to slip photons through glass, instead of being absorbed by glass.

I'll point out in passing that randomness has little to do with that effect. Quite the opposite, in fact; it requires regularity.

The CD makes sense if the diffraction-grating effect is consistent. It could also cause the same interlacing - In principle, you could figure out Logitech's duty cycle by calculating the whole number ratios of the stripe-sizes that make it fail.

The mirror is the one that I don't understand - wouldn't it sense just itself, necessarilly not moving?

Silver and aluminum (the two popular mirror materials) are both more reflective in IR than visible light, so if wavelength matters, my intution is that those mice must use something other than IR.

-Tom

__________________
-Tom McWilliams

Report this post to a moderator | IP: Logged