Monday, April 30, 2012

What's not important.

Recently, a link circulated, regarding a legal decision. This was a decision from the 5th Circuit Court of Appeals, by a three-judge panel, and there was a dissenting opinion. In the dissenting opinion, the judge cited The Princess Bride.

Naturally, the internet only cares that one of their favourite movies is being mentioned by judges, and isn't this wonderful? However, it's definitely worth taking a look at the entirety of the decision. (PDF link) I'm not expecting you to read it, so I'll summarise the facts. Do bear in mind that I'm a layman, and I may misunderstand things.

Monday, April 23, 2012

Same-sex... divorce?

Yes, this is something we need to think about. The US has a patchwork of tolerance as things stand; some states allow marriage without regard to gender of the participants, while most don't. Some recognise out-of-state same-sex marriages, some don't. And then there's California, with its horrible mishmash of things that are still being decided in court.

Now, personally, I regard the various same-sex married couples I know as legitimately married. They've made a committment, just like I have, and that deserves respect. That they don't reap the tax advantages I do from the Federal authorities is a travesty. That said, marriages don't always last. What happens when a same-sex married couple decides they need a divorce?

If the state certifies such marriages, it shouldn't be a problem. But what if the state recognises such marriages, but doesn't perform them? Is it obliged to provide for divorce of marriages it doesn't perform?

And what if the state doesn't even recognise the marriage? I dread to think of the problems which could arise should the respondents accept an answer of "you were never really married in the first place", and then attempt to re-marry.

This is a problem that deserves much more attention than the "none" it's recieved so far.

Tuesday, April 17, 2012

What's going on

You may have noticed that more discursive posts are appearing. I'm trying to make that a thing.

If a post shows up at 1.30pm on a Monday, California time, then chances are it's actually been in the can for a few days. I'm trying to keep enough ahead that I have a buffer of posts ready to go out. Posts at other times will be the usual haphazard posting.

Monday, April 16, 2012

Lifetime, and what it means mechanically.

A large proportion of new cars on offer these days come with what are called "lifetime" fluid fills in the transmission. What this is held to mean is that you shouldn't need to change the fluid in the lifetime of the car, unless you've abused it. Leaving aside for a moment the issue that virtually everything we do to our cars is technically abuse, and should have us on the "severe duty" schedule for servicing, what exactly is the lifetime of the car?

Well, any mechanical system has a lifetime. It will, eventually, succumb to age, wear, corrosion, and so on. This is true whether one is talking about the hinge on a tin of Altoids or the engine of a Volvo, despite the expected lifetimes with proper care being vastly different (and that difference is currently undefined; we don't know the lifetime of a properly maintained Volvo engine, because nobody's managed to wear one out yet). The lifetime depends on the lives of the individual components, and their ease of replacement. A good example of this is a perennial debate among car enthusiasts: the question of timing belts versus timing chains.

Valve timing has been a live issue ever since the camshaft migrated from its former position in the block to the cylinder heads. When it was down in the block, you simply put in a gearchain with suitable ratios and you were golden; the cams pushed the pushrods, which rocked the rockers, and the valves did their job. Sadly, all that extra mass was inefficient, and limited the engine's speed. Nowadays, the standard is for the camshaft to live up above the valves, and push on them either directly or via rocker arms. The pushrods and their play have been eliminated. Unfortunately, this setup means you can't use a geartrain to drive the camshaft; it would be too heavy and too awkward. A new solution was needed.

The two most common options are a chain, not unlike a bicycle chain, or a toothed belt. Both have their advantages and disadvantages. The chain is more durable; a metal chain will last longer than a mostly-rubber belt. The chain is lubricated, which the belt can't be thanks to oil's deleritous effects on rubber. However, the rubber belt is quieter, and it's a lot easier to change. Both are perfectly capable of maintaining valve timing within acceptable precision right up until they fail.

And fail they will. It's a fact of life, a fundamental law of physics, that everything breaks. Within a mechanical context, you have enemies; motion is the enemy. Anything that moves will wear. The lighter it is, and the more it moves, the more it will wear. The more complex it is, the more it will wear. A timing chain is a hideously complex assembly of very small parts, moving at high speed with close tolerances. The rollers of the chain wear. The links can stretch or even break. The sprockets wear. And because the chain is so closely integrated into the functioning of the engine, it's a pain to replace. Frankly, it's highly unlikely to be economic to repair an engine in which the timing chain is worn to the point of problems. It's generally held that a timing chain, unlike a timing belt, will last the lifetime of the engine; but looked at this way, what that really means is that the lifetime of the timing chain determines the lifetime of the engine. Any wear is repairable, but there is always a decision to be made as to whether the expense of repair is excessive compared to the benefits arising from performing said repair.

In contrast to the timing chain, a timing belt has a definite intended lifespan and is then expected to be replaced. It's generally hung out on the front of the engine, which exposes it to the harsh conditions in the car's engine bay but which also makes access for replacement easy. There are more oil seals to maintain with a belt setup; and remember that oil will destroy the belt in short order; but the cost of the belt replacement is essentially trivial. In this design, we've accepted that failure will occur, and planned for it by making replacement easy, and setting a schedule of maintenance which will in almost all cases result in the weak part being replaced before it fails.

What we're seeing here are two different philosophies of design. The chain represents the effort to reduce the chance of failure, at the expense of making failure catastrophic should it happen. The belt represents acceptance of failure as inevitable, and planning to remain one step ahead of failure. Both of these viewpoints are valid; I can easily accept a person looking at the problem of valve timing and, by impeccable logic, reaching exactly the opposite conclusion from me, even though I also used logic. The difference lies in the relative importance of various factors to the person making the decision. My preference is for a timing belt, on the grounds that it can be repaired more easily; others prefer chains because they need repair less frequently. Both preferences are logically arrived at, and that's why this issue remains a perennial flame-war. Both sides are talking past each other, because what they consider their killer points are what the other side thinks of as unimportant.

Going back to the "lifetime" fluid fills in transmissions, it's now clear that what the manufacturer really means is that the fluid fill determines the lifetime of the transmission. Remember, it's mechanical; it will wear out, and eventually break. Whether you consider the expected service lifetime acceptable is your decision; but remember, next time the salesman touts the lifetime fluid fill, to ask him what the lifetime IS. You should have that information to base your decision on. Or, you can opt for a car that has scheduled fluid changes, and accept that you'll pay more in recurring fees for a transmission that has the potential to last longer.

It comes down to accepting a black box, or accepting a slightly more finicky device that you can tweak to your satisfaction. Some things work well as black boxes; an iPod is a masterful black box. Other things work better as openable boxes; I prefer to at least have the option to pick and choose the components in my desktop computer.

Wednesday, April 11, 2012

What I'm actually paying for.

We live, let's face it, in the age of the internet, when it is essentially trivial to get hold of any data. If I like a song that the radio's playing, I can, should I so desire, obtain a copy of it without needing to pay anything above and beyond the cost of keeping an internet connection active. I don't, though. Generally, I go to iTunes, and I pay money. Why is that? How is iTunes beating "free"?

The first part of the equation is ease of access. I already have iTunes; most of the time, it's running, since this machine is so over-endowed with processing power and memory as to not notice any program save some games and my web browser (thanks to the memory leak). So buying music from itunes has very low overhead, especially given that I don't have iTunes taking payment from my credit card; instead, I'm allowed a budget by means of gift cards. Everything happens automagically, and that's a big advantage. Compare the "free" method: search a stupendously large results set, some or all of which may be mislabelled. The legitimate method is giving me a massive advantage in convenience.

The second part is safety. I can reasonably assume that any music I should happen to download from itunes is merely the music, and not some horrendous foul beast that will render my computer a drooling, leprous zombie at the command of some nefarious thug-for-hire, and probably won't be the right music anyway. And if I should get infected by means of iTunes, I have redress. I have paid for a service, with reasonable expectations of what will and will not be delivered, and so I have a means of getting compensation should Bad Things occur. Again, the legitimate means has the advantage.

The third part I've touched on, but it bears repeating: there's a reasonable expectation that things I buy via iTunes will work in iTunes. Generally, they work in other programs as well, but that's not something I expect; it's merely a bonus. Anything acquired via the shady means has nothing to provide an assurance of functionality, and nobody to complain to should it turn out to not be functional. Amazingly, "paid-for" is coming out well ahead of "free" again.

What all this adds up to is simple: I'm not really paying for the music so much as I'm paying for convenience in acquiring it, expectation that it will do what it says, and expectation that it won't harm me. I'm paying because I see my time as better spent listening to the music I like than it would be spent tracking down a good enough copy in the seamier corners, and fixing any problems arising therefrom. The same reasoning applies to why commercial software I use is paid for; it's simply less hassle that way. While I COULD get it for free, there are potential pitfalls and inconveniences that I simply don't consider worth dealing with. If I feel the price of a given piece of software is too high *cough*Adobe*cough* I simply find a cheaper alternative (which can be free; the software I use for image editing, for vector drawing, and for 3d modelling and raytracing are all free, provide their source code for free, and require improvements to said source code to be freely available) or do without.

So while some would have it that only a mug pays for software, I see it differently. I see my time as too valuable to waste troubleshooting something that would probably just work straight out of the box if I laid out a little cash. This is before we go into the ethics of paying creators (an unambiguously good thing to do).