Sunday, January 15, 2017

Metric system silliness.

I'm not sure even where to begin with the metric system. It was a great idea in principle, but doesn't actually work out in the real world. And the acrobatics that its advocates so through to keep it alive just make things worse, and are rather silly.

So the idea behind the metric system was to have one system that unified measures of distance, volume, mass, temperature, and more based on the natural constants of water. So that any unit can be derived from others based on water. And further to have ranges based on multiples of 10, with set prefixes so that addressing things at different scales need not see inflation of digits. This definition didn't last but you'll still see memes lauding the metric system's superiority because of these relations. 

The water based definitions for the metric system didn't last because they were too naive. The idea was that one or two measures plus water would let you derive any third measure, but it turns out that in the real world, it's not that simple and you really need pretty much all the other measures to derive a missing measure. Not to mention that water is seldom as pure as we'd like.  So the water definitions were dropped. From official adoption in 1795 this idea didn't make it 4 years before being replaced.

But before the metric system was even out of the gates it already started to be crippled. Maybe it's always struck you as odd that the default mass unit has the kilo prefix. It is odd. The original mass unit was the grav. Unfortunately grav sounds like graff, and graff was a title of minor nobility at the time, when nobility was not in favor. However the milligrav has already taken on a more colloquial name of gram. So they went with that instead, skewing the alignment of the base units and their prefixes. 

The system that replaced the water system was prototypes. That is to say they made specific objects and set them aside somewhere to be the definition. Of course they didn't make just one of these, they made many as identical as possible, though a specific one was named to be the actual factual definition. Obviously switching to this system removes the smug superiority from the metric system. It also means that the system is based on items that can get lost or destroyed, which isn't great. It also, as it turns out, was still too naive. These prototypes were replaced at least once, to improve the material science, but that's still too naive, because our technology just wasn't in place. As the duplicates of the prototype when returned and compared to the prototype, were no longer the same. Which really undercuts any faith in the system.

As a next step for the metric system they want to turn them back into natural constants. This has already happened for the meter. They have chosen to define in terms of the speed of light in a vacuum. The consequence of this that I think is the silliest, is that this means you can't measure the speed of light in a vacuum with the metric system. What ever result you get the answer is by definition, not by measurement. This seems like a silly thing to do with such an important constant. Additionally there are just the basic questions about this constant and our knowledge of it. Have we ever has access to a perfect vacuum? Have we had sufficiently fixed endpoints in any sort of vacuum to get a good measure of this? Don't we measure everything else with light, so how can we calculate margins of errors for calculating the speed of light, when those margins depend on the speed of light? And I'm not enough of a physics guy to get deeply into relativity, which fixes it, and makes it so that it the wavelength of light from moving things that change and not the speed, but also has time and space deform. Which when you go from pure mathematics of theoretical physics to the real world still has perhaps some degree of hand waving going on.

For the Kilogram, one idea at least, is to make a prototype in the shape of a sphere out of an atomic molecule, measure it's diameter, calculate the number of atoms in it, and then base the definition of the kilogram based on that. So here material purity rears it's head, and in this case not even ions are allowed. Next you have that they will calculate the number of atoms in a sphere. based on it's diameter, cause nothing says imprecision like including an irrational number. But then you have a similar problem to light, determining the mass of electrons, protons, and neutrons. The mass unit will be defined in terms of the combination of all three of them. So really you can only measure their proportionality. And once you have their proportionality, you can never measure the mass of anything again. All you can do is count it. The only free variables when considering the mass of something will be what it is and how much of it is there. If you have those answers then the mass is by definition, and the proportionality of those 3 particles. Even the mass of other subatomic particles might be in fixed if quarks have fixed masses. This seems monumentally foolish and roundabout to me. 

There is probably more silliness in the history of the metric system. However the question that's been avoided is why is the imperial (or any other) measurement system better? Simply stated because they don't try so hard to be perfect, and certainly no one one goes on about why the imperial system is better based on it's vast and constant technical superiority. The smug adherents to the imperial system normally put forth it's cultural superiority, which is an argument that defeats itself. 

I think really the thing that irritates many (including myself) is that it speaks to the hubris of people ready to declare themselves in nearly full understanding of the universe. Time and time again such people have been found to have been lacking. And even worse sometimes their well intentioned ideas survive and are used as ammunition against improved understanding, despite having been originally put forth to promote understanding. The math used by theorists is precise and perfect, the real world is not, and only fits approximately within precise models. As we learn those approximations improve. They may never be perfect, but that's good since it pushes us to look deeper. The ethos of the metric system is opposed to that, and I guess that it why it irritates me so much and find humor in the foibles of the supposedly perfect system. 

Sunday, January 1, 2017

Why Pluto is still a Planet.

In the early 2000's as we started discovering additional large objects in the outer solar system, astronomical displays started to clump these objects in with Pluto. The rate at which we were discovering them made it seem likely that there would be many. And who wants to memorize 30 different planets? Which if you're lazy and don't want to be embarrassed when little kids know all the planet names but you, who are the actual astronomy professional, don't, means that Pluto has to go. That's it. Don't be fooled everything else is just a smoke screen. Pluto had to go so that a bunch of people who were full of themselves wouldn't have to keep memorizing new lists of planets that might subtly change order. Think about how often these same people, who should know very well that Pluto has a trans neptunian orbit, refer to Pluto as being beyond Neptune.

There is a case for Pluto's demotion. We didn't know how big it was when we discovered it. We thought it was bigger. We didn't know it was in a belt of objects, some of which were big. After it was discovered Ceres was briefly considered a planet, until we realized that it was part of a belt. And while it remains the largest and the only one large enough to pull itself into a sphere, it remained demoted to the rank of asteroid, until this whole kerfuffle, so again there was a reason to be considering these things.

But at the same time as we were finding new things in our solar system we were finding "Planets" in other solar systems, and even looking for ones that weren't captured by a star, that are called rogue planets. This is the back drop of the state of the art science at the time. We were looking for bodies outside of our solar system that didn't orbit stars that we were calling rogue planets. Remember that cause I'm going to come back to it. Also it's very possible that in other start systems we will find bodies in all sorts of exciting and strange orbital configurations.

So with a reason to consider the designation of things in our solar system, and the desire to limit how often school children embarrass them, the IAU set out. The definition that they cam up with is "(1) A planet is a celestial body that (a) is in orbit around the Sun, (b) has sufficient mass for its self-gravity to overcome rigid body forces so that it assumes a hydrostatic equilibrium (nearly round) shape, and (c) has cleared the neighbourhood around its orbit".

Of the planet definition criteria b is a reasonable criteria. It is a clear dividing line that objects throughout the universe will be on one side of or another, and as such will be useful with all the extra-solar astronomy that is going on as well as with what's going on in the Solar system. This criteria is very broad though and in addition to what had been considered planets before it also captures stars and many (but not all) of the moons in the solar system. 

Criteria a defines planets as being only in orbit around the Sun, and as a consequence there are only planets in our solar system. No extra solar planets, no rogue planets. Planets are just something we have. This is rubbish of course. Further it is complicated by just what that means. In orbital dynamics there is something called the Hill sphere, which is used when looking at two bodies to see if objects nearer the smaller one are orbiting it or the larger one. Right now the moon is orbiting the Earth, and the Earth the Sun. As little as ~1.27 million years ago when the moon was closer, the earth orbited the moon not the sun. Without wanting to do the math I would guess that for Pluto and Charon it is the case that they orbit each other and it's only their system that is in orbit around the sun. So this is the sort of things we should expect to see out there, so if we apply this definition to other systems and other suns it still will eventually break. This criteria wasn't forward thinking towards improving a definition for science. It's stated really naively so you have to guess that they are talking about Hill sphere's for the criteria, otherwise everything in the system orbits the sun. Also this criteria is part of the dwarf planet definition, so if Pluto and Charon are in orbit around each other then they are not even Dwarf Planets, or the framers we being even more naive and less scientific than that. 

After what I did to criteria a, I'm sure that you are expecting that I think criteria c is a pure mess. It is. It's so bad and so naive that people keep coming out with "interpretations" of it, to fix the problems. So let's begin with the big one. Pluto and Neptune. Pluto crosses Neptune's orbit. How can you have cleared your orbit when there is something that big that crosses it? Then you have things like Earth and Cruithne, an asteroid in a 1:1 orbital resonance with the earth that crosses our orbit but isn't technically in orbit around us. And add on to that trojans. The original trojans were in Jupiter's orbit, but it's such a thing that we've generalized the term for all the planets, and they are objects in the same orbit, at the same orbital period, that are in the right spot to roughly never get cleared out of an orbit by a planet. That's right we know, and knew when the definition was made, that no planet could completely clear it's orbit. This definition is a lie. It's a lie that was put in place to kill Pluto and keep the rest of them out. And beyond the problems in our solar system there are lots of possible interesting orbital dynamics that we might find out there, that won't fly with this either. Sure a lot of them won't be super stable, but what's the threshold of stability? Nothing in our solar system is actually stable, so it can't be forever, and then after that you let in a lot of interesting things that we are going to find out there.

So the current definition is unscientific and garbage. As such I refuse to accept it. And revert to the previous state in which Pluto was a planet. If some one tries to argue for the demotion of Pluto, there is a simple question you can use to break them, "What is the difference between a rogue planet and a rogue dwarf planet?". You don't have to give them anything harder to think about than that question which the framers of the new definition, left unanswerable, which counter to the mandate for even considering the definition. In response to searching for and finding lots of new things they revised the definition to not be applicable or otherwise be nonsense to the new? No. Just No. If a future definition gets put forth that's good and demotes Pluto, I'll accept it. But I refuse to accept that trash just like I refuse to accept other garbage that wants to claim to be science.

What is required for a good definition? It has to consider all the types of things we're looking for, so it should consider objects in other sun systems as well as those in between them. It should be based on the intrinsic properties of a body not extrinsic ones. We're not going to have perfect knowledge of everything in a solar system when we first see a body, so it's designation shouldn't rely on other bodies that aren't clearly playing a major role in what's happening with that body. Also the definition should be very specific. It should start as a hard computable thing that is technically precise that can be summarized for humans, rather then something naively high level that might melt down when you actually get into the details. 

I don't know if a body will ever put out a good definition that takes off. But I have some predictions. I predict that we'll define some new term like planetoid to account for objects between solar systems. Of objects that pull them selves into spheres, I think the divisions are more like to keep Pluto with Earth, than Earth with Jupiter. I think that after we define terms like planetoid (potentially with sub groupings) we will then define planet and moon based on orbital relations so that the Earth and the Moon can be the same type of object but will also have terms to describe them in terms of a start system. (I think moon and satellite will be split in this situation). And I think that in systems of mutuality like Pluto and Charon, the two are likely to hold the same status. 

Wednesday, December 14, 2016

Leap seconds clever but stupid.

For a variety of reasons the Earth isn't always perfect about making it around it's access in the expected amount of time. The implemented answer to this is Leap seconds. Periodically we add a second to our time keeping systems. Since 1970 we've added 26, and we're slated to add the next at the end of the year. I think that this system is stupid for a variety of reasons. (also let's just state off the bat that all numbers in this post will be approximate)

The first thing is the question of why adjust. Why do we need astronomical noon to be perfect noon in part of London? Not all of London mind you, just part of it. At a latitude of 51 degrees 30 minutes, a second's worth of the earth's rotation is only .179 miles. London is over 600 square miles, which is equivalent to circle with diameter 27.6 miles. which at that latitude would take 154 seconds to rotate through. At the current rate of leap seconds it would take over 200 years for the perfect moment of noon to make it all the way from one side of London to another! And it's not even perfect noon since the sun is always slightly to the south. And with s system that takes so long to pay off London might be growing faster than the need for leap seconds to keep noon in London. Will it matter when we've got colonies all over the solar system where in the sky the sun is at noon in London?

Another big problem with this is the effort it takes to actually implement this for all our time keeping systems. Human's don't notice that big a change but computers sure do. So in all our computer systems we keep having to code in and fix weird edge cases like adding a second. Adding this leap second is hard and it's buggy. A stupid amount of human effort goes into pulling this trick off. And probably more effort goes into fixing the things that break when it fails. And this time Google's solution is real time to not implement the extra second but to slowly stretch the seconds on other side.  Hopefully nothing high precision will be using googles clocks that day. Plus there is the question about what they do going forward. Normally things keep track of all the time unit seconds since the beginning of 1970, and have a table to look up when there have been extra seconds, but if google computers don't ever live that extra second then how are they eventually going to take care of it. Are the all the computers just eventually going to register an error of a second and skip ahead? The effort that is involved just doesn't seem worth it.

This also has the gross effect of eliminating all other units of time. If the span between 11:59:30 and 12:00:30 is a minute, then how long is a minute? It depends on which one. If it's one of the ones that has had a second added then it's 61 seconds. And so on for larger units. This already impacted years with leap years, and moths weren't regular anyway, but now everything above a second is not just a calendar indicator instead of an actual unit. And because of weird implementations like google's, a second isn't even really that any more.

One potential solution for this is to do it in time zones. Right now to tell you your local time. your computer has to look up the time zone and modify the time based on that and then look up the leap second chart and modify the time based on that. Why modify the time twice? Why not just make leap seconds or any other minor adjustments for local pride part of the time zone? That way locale's that want noon to be special can finally have it that way, and then they can keep it that way for as long as they desire.

But what we're really going to eventually drive at is global / universal time. The inefficiency of having so much space/stuff dedicated to one person is eventually going to falter and we'll end of with round the clock people trading off space/stuff throughout the day. Not to mention globalization which increasingly creates communities in many time zones that all are on one schedule. All the effort that has gone into local time, and keeps going into maintaining local time is eventually going to be set aside. And then we'll look at the legacy of the leap second look up table and ditch it. Hopefully by then it won't be so late that making that change in and of itself won't be a major headache of an undertaking.

Tuesday, November 29, 2016

Flying Cars.

Back to the future day was over a year ago, so where are our flying cars?

One of the biggest things holding back flying cars is the human element. We're not that great. And air traffic control for humans would suck. Transition to a future without human drivers. Then that problem goes away and we can have flying cars. We're already doing extensive work in automated drone piloting. Is it really that hard to imagine flying cars as a next step?

Self-driving cars, have a lot of perks in terms of infrastructure. They reduce traffic, so road expansion to mitigate traffic is reduced. They can park on their own, so there doesn't have to be parking everywhere, since your car can drop you off then go find parking. Flying cars increase those perks. At the point where we have flying cars roads are only need for long haul transportation, and so we can spend a lot less money on them. And if you reasonably assume vtol, parking lots can be extra dense. 

And if we want to cast our gaze further forward. Once we have ubiquitous flying cars why would ground level even still be an important thing. we can all have aerial egress. no more going to a central shaft to go up and down. At that point cities might end up looking more and more like the worlds of the Jetsons, or the 5th element. 

Perhaps we'll even moving to ocean based habitats, either floating or in platforms high above the sea, leaving the land to be a nature preserve with only small groups of luddites left. But, of course this magical future has the pitfall that as technology becomes like magic, a crash in society takes us back to square one. In which case our ocean based super-society will be just a legend to the luddite remnants of civilization left on the mainland. But that particular scenario was deliberately picked to sound like the legend of Atlantis.

Getting back to reality: As flying cars kept failing to materialize we kept thinking of them as an increasingly distant prospect. But if fully self driving cars are mainstream by 2020, it's not improbable that we might have flying cars mainstream by 2030. We've gotten so used to predictions of a dull future that an exciting one might just sneak up on us, and that is pretty wonderful to think about.  

Tuesday, November 1, 2016

Apple's Laptop Event

I've seen a lot of circular and confused comments about Apple's laptop event last week, and I thought I would chime in with some of my own thoughts as well clarifications of some of the things I see confused.

To me the biggest question, is what about desktops. There were reports that the iMac has been delayed, but we've heard nothing about the Mac Pro or Mac mini. So we still need answers and it doesn't seem that likely that we'll be getting them this year. If the desktops were addressed, then with a straight face you could tell all the people claiming to be ultra-pro, but demanding in a laptop that they are being silly. With the MacBook Pros still not stepping up to fill all the desktop roles they really have a hole at the top of their computer lineup.

So from a line-up perspective what happened? Apple got rid of the 11-inch MacBook air, and the old thick 13-inch MacBook pro. They left the entry level 13-inch MacBook pro in place but limited it's customizability. Replaced mid-line 13-inch models with the 13-inch without the touch bar, and the high end 13-inch models with the 13-inch with touch bar. They left the entry 15-inch without discrete GPU in place, and replaced the versions with discrete GPUs with the new 15-inch with touch bar. And it looks like they left the 13-inch MacBook air and MacBook alone. If you weren't familiar with the specs of the old laptop lineup or that they left those old MacBook Pros in the line up. It can look like the prices shot up, instead off it being the case that they just didn't make new entry level machines.

As to whether or not they are falling behind Microsoft, with their announcements. I can imagine that for those that draw all the time the Surface Studio is great. And for a company that has been doing a full court press on the laptop front about touch it's great to see them stay true to that message and finally deliver it on the desktop. I hear that they drawing isn't as good as on an iPad, but they are sticking with their gimmick. I don't know that I've bought into the whole touch thing and I'm not sure everyone has, so I think it's ok that Microsoft is better at their gimmick then Apple. But for Microsoft they are a one platform company, where Apple is not so I think it makes sense for them to find a way to unify their platform when Apple is not.

I haven't played with one of the new macs yet, so I can't speak about the touch bar first hand. In the presentation it seemed way better than the pre-announcement rumors I had heard, which made it sound like it would just have buttons with changeable labels. I think that it's all going to depend on the apps, as to whether or not it takes off. I think it will be a bump in the road for the people that do still use keys in the top row with any frequency, which I do because of my correct editor choice of vi, but I hope it will just be an adjustment, and that what there is to be gained is more than the lose.

For the ports, I think that it's great that they went with Thunderbolt-3 / USB-C. I think that thunderbolt-3 offers more than just straight USB-C, and it's good that it's there, and the ports can all be used for whatever. And I think that the industry has spoken and USB-C and/or Thunderbolt-3 are the way of the future. It's annoying that it's everything all at once, and before we've even finished off the last of the magsafe-1 devices. But I don't think that these ports will become "mac" ports the way firewire or the first two thunderbolt versions were, which is great.

As for the memory limit of 16GB. I am constantly amazed at what can be done by tablets and phones with less than a quarter of that. And one of the reasons to use more memory is to make up for other bottle necks, which have all gotten faster. I do think that there is a place for a mac with more memory, but it probably shouldn't be a laptop anyway.

And doubling back to the complainers. I saw people both complain about the "price change", and that the high end wasn't high enough. Their high end prices didn't really change, so you if you're making both those complaints you probably weren't in the high end before. And if it really is a big limit on you, why are you trying to do all that much on a laptop anyway. It should probably be a minor annoyance unless you're trying to do way too much on a laptop.

I also think it's funny to see so many people talking about how Apple is alienating their core customers. They are the most profitable company ever and their market share in the industry had been growing until they got to the heart of this drought. I imagine that they know exactly who their core customers are, it's probably not who it was 15 years ago. There are also a lot of internal equations that could see someone leaving the mac platform, and I think that those people who have already made that decision will use a new release to re-affirm it to themselves, when nothing could have kept them on the platform anyway.

In general, I think that there is a lot of unnecessary doom and gloom (as long as they do something about the desktops). I think that the size and efficiency improvements will be a big plus to a lot of people. The new task bar provides an interesting opportunity. Finally have a secure enclave in the mac is interesting (and from the point of view of someone that supports them scary). It is disappointing that apple is moving to the model of having the old model as the low-end/entry model at the end of a drought.

Monday, October 24, 2016

Smart Watches

I've had an Apple watch for about a month now, so I thought I'd share my thoughts, on it in particular, and on smart watches in general.

Why did I buy the apple watch? I wasn't buying a new smartphone this year for a change, but saw from Apple's entry into the bluetooth headphone market, that the remote was a thing of the past, and thing to fill that gap would be a watch. So if the new ecosystem was going to be one with a watch, I wanted to buy it in this off year, to make next year cheaper, and to be a position where I could make a better judgment about the proposed ecosystem. I also kind of think that health trackers of some sort are the future, so might as well, jump in.

I probably don't use the apple watch typically. I'm not an alert fiend or a workout hound. I sleep with the watch on and have it on for most of the day. It frequently gets some time off (and charging) right after I get home, and again while I shower in the morning. And with this routine, and with rise to wake off, I don't have problems with the watches battery life.

I think where the watch really shines for me is as a digital watch. It's probably the best I've owned. It's alarms with it's taptic engine are nice and discrete. It's easy to have many of them and to set them up in same-complicated fashions. It's replaced my iPhone for most of my alarms. It's timers are also good. And it's generally more convenient to get to then my phone for the time. I don't think that any watch has lasted on my wrist so long.

It's a fitness tracker, but I don't watch that that much. I'm not trying to use it for that much. My curiosity about such things is fed by it. I think that it's a good supplement to my phone. And it meets my idle curiosity. But a large part of my interest in this function stems from the death of my mother. She died alone and her last hours are a mystery to us. We used some of the tracking in her phone to shape up that period. But a watch that's almost always on me will, I imagine do a better job.

I haven't gotten to use it as a remote yet. I can but it's somewhat silly since I'm not using bluetooth headphones. Even when I do, then model I think is most compelling has an on wire remote. So we'll see if this use pans out. maybe I'll be able to do enough or maybe it'll be slower for all the things the on cable remote can do, but too awkward to do the more complicated stuff.

While my phone is a device I spend a lot of time staring at I think the watch is a device for glancing at. I really like the intimate notifications of it tapping me. and using it for apple pay can go either way depending on the terminal location. For health monitoring and discreet notifications, and maybe as remote for things on less accessible objects. Overall I think that it's a plus. If you're not sure, I'd consider waiting until series 3, comes out and you can potentially get a used series 2, on the cheap.

I think there is a future for wearables. I imagine in the future, I'll wear headphones that work for all of my devices. have a watch for when I'm on the go and need to interact with something, and really only pull out my phone for entertainment/productivity, when I'm stopped somewhere for a while. I once thought that what we should have is a personal area network, that lives in a brick in our bag, and that can manage choosing between cell or wifi, and have lots of endpoints we can interact with. Now that's what we're getting but it's going to be our phone.

Thursday, August 29, 2013

Attack of the self driving cars.

I've seen multiple posts so far lamenting the dark side of self driving cars, the lost jobs. And this argument starts to spiral into if car automation actually gets that good, then a lot of the problems that have gotten in the way of a lot of other automation will be solved, what are we to do. And this raises the legitimate question of what happens to the masses when it's cheaper then a living wage to build and have a robot do everything.

Let's start in the short term. Most cars are not driven by professionals, this means that if we can come up with one to many jobs related to these new cars they may be absorbed by the dislocated work force. The first of these jobs is UAC pilots for when cars get into trouble. Secondly these cars will likely have a much lower tolerance of dysfunction then people have with their cars, so the need for mechanics will rise. And really where the magic is with that translation is that this change will be spread equally over the cars that are professionally driven and those that aren't so it won't need to be a 1-1 exchange for jobs.

Now to look at what happens as robots take over everything. 

First thing to do, expand. The needs of how many people or service people a mostly automated society can take care of are bounded by the geographical area that the people are in. So if we decrease density and expand we can grow human jobs that way. Of course this means expanding int new types of areas which will require new types of innovations so the innovation sector will expand.

The second thing to do is change the education system. As economic activities are increasingly reduced to ownership, innovation, and education, we need to move increasingly towards free educations and then towards paying people to become educated ( rewarding higher levels of pursued education with higher compensation). So that as industries die, those that can are motivated to move up the path of education. That way we will always move towards a maximized innovator class and move people to their optimal level of innovation so that people who can learn more don't take away opportunities from people that are maxed out.

The third thing, is to avoid the accumulation of ownership into the hands of individuals. One way to do this is to adjust the rules for inheritance, to favor the distribution of inherited wealth both to as many parties as possible and to those that are not already wealthy. And if anything is left to fall to the state, either hold it in a social trust, or distribute it as possible, depending on your politics.

The fourth thing you do is to manage the population in both directions. If there are too many people you increase the disincentives toward childbirth. If there are too few people, you make it a pro to have kids.

You don't do all of those things overnight so we need to start moving in those directions before we are too messed up. But maybe I'm completely wrong. Those are just my ideas.