The Call of the Open Sidewalk

This was in response to someone on Reddit who suggested that software patents on video compression methods were somehow more reasonable than other software patents because it was hard to come up with such methods. It seemed blogworthy.

Digging a trench is really hard too. How long does the trench have to be before I can get a patent and prevent anyone else from digging a trench that long?

Video compression patents are an excellent example of why software patents don't really make sense. If I somehow decode some compressed video the result is a bunch of coloured dots on a screen representing a moving image. The output is pure unmitigated data.

Now the people that believe in software patents will tell me that I can't use my computer to interpret the H.264 bit stream in the H.264 way without the permission of the 1600 or so patent holders. I can interpret it in an infinite number of other ways. That is OK. Just not the one way. That is even if I have read the standard and do the programming myself. There is just something magic about the one interpretation.

I happen to believe that the invention of the general purpose computer is on a similar level to that of the invention of fire. If I believe in software patents I also have to believe that I can only use this incredibly powerful invention in very circumscribed ways. Imagine the reaction if the people that used to spend time tenderizing meat with rocks tried to get a royalty from the people using fire to do the same thing. The fire users might patiently explain that the world had changed in a very fundamental way. They might instead suggest that the people trying to make their wonderful new thing as useless as the old thing just go fuck themselves. I will leave it to you to decide which group I fall into...

At any rate, suggesting that software patents are real in some way is deeply offensive. You should expect to get called on it...

posted at: 17:22 | path: /politics | permanent link to this entry | Comments (0)

I sold my car the other day. I wasn't really using it. This has forced me to deal with some day to day issues in a different way. Every month or so I would take the car to a large supermarket and load it full of all the heavy stuff I did not feel like carrying home from a more local store. This is no longer possible.

They make something for this application. A classified ad web site and $5 got me this:

It worked. It lacked both the capacity and the load carrying capacity of my car trunk. The solid wheels meant that I had to pay attention to avoid having the cart hit discontinuities in the pavement. Having to pay attention to the surroundings seriously detracts from a walk. It also had an associated old lady vibe. I am an old man. That is entirely different.

Here is my current grocery hauler after a triumphant trip out and back:

It is rated at 360 KG (800 lb). The pneumatic tires simply glide over discontinuities in the pavement. Once you get it up to speed you can hardly tell it is there. This is how a guy gets the groceries home from the store.

The wagon shown was originally sold as something to help with the incomprehensible things people do to interfere with the plant life in their yards. As a result it would be quite reasonable for people to assume that my super hot and high maintenance trophy wife had simply taken the BMW and left me with the shopping.

On the way home from my first shopping trip with the "garden cart" I had a realization that made me feel a little less superior to conventional shopping cart users. A light, two wheeled cart was simply essential to grandma's mission. The environment had changed in a significant way and I had not noticed. Grandma had to deal with curbs.

The local disabled community fought long and hard in the battle against curbs. Today there is not a single significant discontinuity from my stairs to the aisles of a supermarket 1.4 km away. That is kind of remarkable and represents a triumph. I now feel a bit bad about sitting out the conflict.

The morals to this story:

Sometimes different communities have more in common than they realize.

Some unintended consequences are good.

The pneumatic tire is a really clever and significant invention.

posted at: 12:58 | path: /politics | permanent link to this entry | Comments (0)

I recently had to design a reflector for a light fixture. After I was done I ended up with two dimensions that depended on one another. After some thought I had a true mathematical insight. The solution to my problem was the same as the solution to the intersection of a line and a cone. After futzing with the rotations (nothing was at right angles) I would of ended up with two equations in two unknowns. Using the more or less mechanical process of algebra I could of solved those equations.

My abstraction led to algebra. That is because I spent a lot of time as an undergraduate in the faculty of engineering. As a result the fact that I have access to a computer did not really help that much.

So what would I of had to know to be comfortable in solving this problem by programming?

1. Addition and scaling.
2. Cartesian coordinates.
3. Distances in 3D Cartesian space.
4. Simple iterative optimization.

That is pretty much it. I didn't even have to know about real numbers if I knew anything about scaling integers. Note the absence of geometry or algebra.

My point is that if we want to teach people how to solve problems with computers we still have to teach them math. But it would be different math. That is more or less the dilemma we face.

Anyway, in the end I solved the problem with some stiff cardboard and scissors. I learned how to do this quickly and accurately from my father. I guess there is some sort of moral here too...

posted at: 08:15 | path: /politics | permanent link to this entry | Comments (1)

This was originally a Reddit post. As so often happens when I reread something I wrote it just seemed to get better. It would be a shame to starve this blog of such insightful commentary:

Up to four years ago the CRTC was more or less harassing the monopoly telcom companies and leaving the new ones alone. The theory was that with enough leveling the new ones would get significant market share, the CRTC could step back, and Canada would have a competitive telcom market.

It didn't really work for reasons that are not important anymore. When the government forced the CRTC to let up on the monopoly companies the obvious happened. The market went from sorta competitive to not at all competitive. The results of many years of effort were reversed in a few years.

It seems that the government was being willfully stupid in this. I guess we can only hope that they have some clever long term plan to let things get so bad that the political will will arise to actually fix things.

Simply allowing more foreign competition as the report suggests will only mean that the names of the companies will change. We don't need more competition. We need to make competition possible.

When the US government broke up AT&T back in the 80's they didn't just stop at that point. They worked to create a legal and technical environment that would allow competition. In particular they created standard interfaces on both sides of the monopoly company. On the phone side they created the phone jack. On the long distance side they enforced technical standards that allowed any company to provide long distance service. Competitive markets erupted and the cost of phone equipment and long distance fell dramatically.

Things are simpler these days. Everything is data. We need a standard interface for access to last mile infrastructure. The government should simply prohibit the connection of buildings to the curb unless the resulting infrastructure can be shared in a competitive way. We need a phone jack for the data age.

This should not only apply to fiber but to radio based connectivity. Spectrum is a limited resource. Simply giving new spectrum allocations to companies who might someday provide some sort of service is irresponsible and wasteful. Voice is just a application.

This was a response to a CBC article about a report. As so often happens with the CBC the article seemed to be all about how bad the Conservative government was. I think I have to give them this one. This was an instance of policy that made me wonder if the government was really run by space aliens bent on the destruction of the human race.

After having thought about this a bit the interface stuff seems easy. Ethernet jacks can transfer 10Gbps in an entirely standard way. Just encapsulate standard size Ethernet frames in jumbo frames with a service number. TV stuff does multicast (the local telephone company does this already for their TV service). No one cares anymore about separate phone service. There; the technical part is done. If someone like myself can come up with something that works off the top of their head then it is not a hard problem. We just need to do the political stuff. Perhaps the "We need a phone jack for the data age." statement can be used as a rallying cry.

Currently the CRTC just comes down from heaven from time to time and forces the people with stuff like fiber to the curb to allow their competitors to use the system. Without the technical standard this ends up being kind of counterproductive. It is always treated as a pollitically motivated surprise by the owners of such infrastructure. As in "surprise sex". A huge battle ensues.

Really, how hard can this be?

posted at: 23:20 | path: /politics | permanent link to this entry | Comments (0)

For most of my life Environment Canada reported wind chill using units of Watts per square metre. If it was 1500 W/m^2 it was uncomfortable. If it was 1800 W/m^2 you would want to think about covering up exposed skin. If was 2300 W/m^2 then things were sort of life threatening.

At some point it got cold and windy in Southern Ontario. Environment Canada thus reported wind chill values there for first time in a long time. The people of Southern Ontario responded to the use of the units of "Watt" and "metre" with fear and confusion. This generated the political will to come up a new way of expressing wind chill and is the reason this was posted to the political category. What happened next had nothing to do with science.

What was wanted was a way to express wind chill as a temperature. The political people asked the science people to do this. There was a problem. The problem was that expressing wind chill as a temperature makes no sense at all.

People have a body temperature around 37C. If the air temperature is 37C then it does not matter how fast the wind blows, there will be no wind chill. As the temperature decreases then the wind chill will increase (if there is any wind). If the wind then increases there will be more wind chill. This ignores various factors like evaporation but the point is that temperature and wind speed are the factors that cause wind chill. By pretending that wind chill is actually a temperature we are treating one of the things that cause wind chill as the wind chill effect itself. Getting cause and effect backwards is bad. Thinking that cause and effect are the same thing is a lot worse.

So here we have a case where some smart people were asked to do something senseless. In some cultures there would be a fight that might last for years. Canadian culture is famous for compromise. In the case of wind chill I feel we came up with a truly world class compromise.

I like to imagine that the discussion went like this:

INT. MEETING ROOM - DAY

POLITICAL PERSON (brightly) We need to come up with a way to express wind chill as a temperature.

SCIENCE PERSON blinks. S/he looks thoughtful for a moment.

SCIENCE PERSON (resignedly) That does not make sense even at the level of simple logic.

POLITICAL PERSON (brightly) Then how can we make it make sense?

SCIENCE PERSON (slight eye roll) It might be helpful if we knew exactly why do we need to do this?

POLITICAL PERSON (still brightly) We need a way to express wind chill in a way that is more intuitive.

SCIENCE PERSON (incredulous) Expressing it as power per area is too abstract?

POLITICAL PERSON (seriously with firm eye contact) Yes.

SCIENCE PERSON (gazes at a point on the ceiling) OK then. Why do we need to associate this with a temperature? Why not create some sort of unit-less index? For example we could make zero denote the case where there is no windchill and five denote the case where frostbite is possible. Then the index would almost always be a single digit from zero to ten.

POLITICAL PERSON
Good. Good. Just thinking out loud here. What would happen if we made this index negative. You know, it gets smaller as the windchill increases. We could, say, make the index for the frostbyte windchill the same as the temperature that a person would get frostbyte if there was no wind. We could do that for all the other wind chill conditions.

SCIENCE PERSON (irritated) Very funny. The problem with your scheme is that wind chill does not work that way. If there is no wind then there are another set of variables that determine if someone gets frostbite.

POLITICAL PERSON
Great point. Still thinking out loud here... What if the person is walking?

SCIENCE PERSON (confused) What?

POLITICAL PERSON (excited) Yes! Most people do not just stand around outside in the winter. If they are walking then there will always be some wind! So can we do this?

SCIENCE PERSON (peevish) In theory yes but it would still not be a temperature.

POLITICAL PERSON
I can work with that. We will just leave the "C" off the index we provide to the public.

SCIENCE PERSON (desperate) Making a windchill index that looks like a temperature is being deliberately misleading.

POLITICAL PERSON (smug) Perhaps, but that is a judgement call. Report on my desk by Monday. Have a good weekend everyone!

OK, that is almost for sure not how things transpired but the result is accurate. In Canada the number given for the wind chill represents the temperature that would produce the same cooling for a person walking at 4.8 km/h (3 miles per hour) in calm conditions. The people involved even put people on a treadmill in a cold room with a fan. By measuring skin temperatures they came up with an impressive looking formula that takes temperature and wind speed as parameters and produces a negative number that looks like a temperature. This ended up being sciency enough that the United States has now adopted this formula for their weather reports.

Interestingly enough the hack involving the walking speed is no longer mentioned on the explanation page for the wind chill value on the Environment Canada web site. They now claim that it represents the "feels as cold as" temperature for the zero wind condition. ... which of course makes no sense. In time everyone will forget the thinking behind the formula and it will become a legend.

An interesting thing here is that people in places that actually experienced a significant amount of wind chill (Canadian Prairies) generally preferred the straightforward W/m^2 way of expressing windchill. My theory is that people in those places have a better intuitive understanding of the difference between temperature and wind chill. I personally still prefer the wind chill expressed as W/m^2 even after 9 years of the new system.

posted at: 19:44 | path: /politics | permanent link to this entry | Comments (0)