My main guess (I'm not an expert) is that it's to ensure that the double-window system is fully redundant. Without the hole, in the event of an outer-window failure the inner window would already have gone through a variable and probably large number of hours of stress from differing air pressure. With the hole, the middle window is (hopefully) effectively factory-fresh with zero service hours at the moment it takes over. That would be especially good if the failure of the outer window had anything to do with a quality defect in that window (or mishandling of it) instead of being completely adventitious: in that case, maybe it's not the only window pane in the aircraft with a dodgy service-life rating? If the outer windows have to be replaced every n flying hours, then not stressing the middle windows would also make sure that they don't have to be replaced as often. (This seems a bit clearer from reading the io9 version http://io9.com/why-is-there-a-hole-in-airplane-windows-17036... , though I don't think it's stated explicitly there either.)
The purpose of the smudge pane should be obvious.
There are two panes and a hole left to explain.
First for the panes:
The article says the two panes are for safety reasons and I buy that. Equally important in my opinion is another reason.
It's the same reason you probably have double panes in your home: thermal and noise insulation.
Air has terrible thermal and noise conduction. So you need two panes and air between them.
Why the hole:
Remove moisture, I buy that.
Allow pressure to equilibrate between the passenger cabin and the air gap between the panes.
I buy that too but the article does not really explain why this is necessary.
If the two panes would be sealed absolutely tightly the pressure between them would stay at normal level in any case.
As an aircraft climbs, the air pressure drops in both the cabin and the outside air — but it drops much more outside, as the aircraft’s pressurization system keeps the cabin pressure at a comfortable and safe level.
Without the hole and with perfect seal if the pressurization fails for some reason you'd have maximum pressure gradient on both panes. The redundancy of the second pane would be lost.
Just to put a number on it, the cabin pressure at high altitudes is given by Wikipedia as around 0.75 atm. (It varies by aircraft model.)
If the inner cavity between the two windows did not have a hole, the full sea level pressure would be pressing outward against the outer window. But with the hole, only about 75% of that pressure is present.
Additionally, if there were no hole, the inner cavity between the windows would be positively pressurized with respect to the cabin. This means the force on the inner window would be backwards from what it would have to bear if the outer window failed.
I think it's saying if the two panes were sealed without a hole at ambient pressure when the plane is manufactured then the outer pane at elevation would have ~1atm of pressure pushing out on it and the middle pane would have ~1atm pushing in on it. With the hole the outer pane only has cabin pressure pushing out on it, which is ~75% of 1atm.
If "what to make X" is your design choice, you would not want to have it always pegged at 1 atm, as it would be if it was/could be completely sealed at sea level. It's simplest to have it equalize with the cabin pressure.
When looking at designs of these types, I find it's instructive to think about what catastrophic event the designers are trying to mitigate. In this case, my guess is some sort of decompression where the pressure inside and outside of a window is rapidly equalized and the people onboard don't have time to react in an intelligent fashion.
I think the purpose of the hole is to ensure that stress does not build up on the redundant pane so that it is only used in the case where the main pane fails. When you build redundancy into a system, it's important to ensure that failure of both pieces doesn't happen at the same time. Either can fail on its own, but they can't both fail.
So you want the middle window to be as close to the ideal condition as possible when the outer window fails. By adding the hole, you accomplish this. Moreover, once the outer window fails, the hole ensures that the middle doesn't need to take the full pressure differential and allows the pressure to equalized in a somewhat controlled manner. Yes, the pressure in the cabin will drop, but there are oxygen masks that will drop if it gets too low and the pilots can always descend to a lower altitude. The people onboard will have the time necessary to take both necessary actions.
This explanation makes a lot of sense--I agree that not having the hole removed redundancy--but I'm not sure some of your details are exactly right.
> If the two panes would be sealed absolutely tightly the pressure between them would stay at normal level in any case.
I think higher pressure inside the plane would push out on the inner pane, pushing it closer to the outer window, so the pressure between would go up compared to the outside.
That is to say, cabin pressure would be highest, pressure between the panes would be in the middle, and pressure outside the plane the lowest. But ALL THREE pressures would be lower than air pressure on the ground.
> Without the hole and with perfect seal if the pressurization fails for some reason you'd have maximum pressure gradient on both panes. The redundancy of the second pane would be lost.
The air in between the panes would provide some "padding" so the outer pane would receive less pressure than the inner pane, but there's not much air between the panes so the padding, and difference in pressure, would likely be small, so yes, redundancy would be lost.
"That is to say, cabin pressure would be highest, pressure between the panes would be in the middle, and pressure outside the plane the lowest. But ALL THREE pressures would be lower than air pressure on the ground."
I think this is incorrect. Assuming the windows are sealed at ground level, the pressure between the panes would be slightly below ground level at altitude, due to flexion of both inner and outer panes; the pressure in the cabin is substantially lower than ground level; and outside pressure is the lowest of all. Thus the pressure between the two layers is the highest of the three compartments at altitude, subjecting both layers to considerable stresses with every pressurization/depressurization cycle. Now if you sealed a near-vacuum inbetween, that could help with thermal insulation and prevent condensation, but would cause more stress on the ground.
It seems likely to me that the sill is designed to resist pressure principally from inside the plane to outside the plane.
If sea level pressure was trapped between the two windows you'd have pressure in the opposite direction on the inside window whenever the cabin altitude increased, which would happen every flight and induce cyclic stress on the sill that it wasn't designed for.
> It's the same reason you probably have double panes in your home: thermal and noise insulation. Air has terrible thermal and noise conduction. So you need two panes and air between them.
I think it's the lack of air in double-paned windows that helps. Usually double-paned windows have a vacuum so that air molecules can't move from the outer pane to the inner pane. Surely air is a fine conductor of heat if we consider our weather system. And surely it's also a fine conductor of sound, if we consider our hearing system. In soundproofing applications, open air channels are the top priority, and the simplest way to dampen the sound is by adding mass. Sound doesn't travel at all through a perfect vacuum though (no sound in space), so typical double-paned glass should have an effect there.
> a separate but related function of the hole: to release moisture from the air gap and stop (most) fog or frost from forming on the window.
You could get around that by having the panes airtight and produced in a moisture free environment, but that would be more complicated. Think about failure rates in double paned windows which don't have to undergo the stresses on an aircraft. I'd assume even the outer panes aren't completely airtight to keep things simple:
Agree, the answer to the question was not articulated properly.
Right at the beginning they should have said that the hole is for "releasing the moisture from the air gap and stop (most) fog or frost from forming on the window."
The hole actually dilutes the safety wee bit: "In the extraordinarily unlikely event that the outer pane fails, the middle pane takes over. And yes, in that case, there would be a small leak of air through the breather hole—but nothing the aircraft’s pressurization system couldn’t easily cope with."
TL;DR: Aircraft window hole compromises safety to improve the view :-)
Um... that's not how I read it. It's about controlling the failure modes. The pane with the hole is the "backup"--without the hole, it would be stressed roughly the same as the outer pane during normal operation. This has a major consequence: when one breaks, the backup will probably be in the same sorry state as the primary before it broke, so it will break soon too! Not a very good backup!
Sure, because they are sharing the normal load, it is less likely to break in the first place. But, you'd have essentially doubled the strength, but removed the backup. That's not a good trade-off.
We have stored many terabytes of unaggregated transaction records in Vertica and analyzed large subsets fully ad-hoc, on-the-fly in less than 5s. We have also used BigQuery to directly power dashboards with few-second response times, also analyzing billions of records at a time.
On the other hand, we both build and use summary tables with Spark. (in a relational format to boot, and using Spark SQL).
I think you would benefit from re-evaluating the assumptions you made 4-5 years ago.
Most of the performance differences are in how these systems handle representation, I/O, and execution scheduling. In other words, implementation. The challenge in closing the gap is that the architectures across systems are quite different so you can't just add performance that is largely derived from fundamental architecture.
Columnar MPP needs to be carefully defined to answer the question. Some important data models that fit in these architectures operate on data types that are not meaningfully order-able at a mathematical level i.e. you can't sort them. A lot of columnar implementations, and virtually all in open source, assume sortability as a property of the represented data types.
tl;dr: Columnar MPP sometimes is not far outside what you can express with Parquet/Spark/MapReduce/etc, just much faster, but there are data models supported in some advanced Columnar MPP systems that are not usefully expressible with that stack. It depends on the platform and the use case.
Sorry I don't think the last paragraph you said is true. I dont see any special data types or models that can be modeled by MPP architecture but cannot be modeled in Spark. In short, I don't believe there is much difference at the physical execution level between MPP and Spark.
I also don't understand what you mean by orderable. Spark does not require records to be oderable. Maybe you can elaborate?
The problem isn't that "nobody" ever hears from the server again. The problem is that "not everybody" ever hears from the server again. If you aren't sure that the server received your write, but another system (that does have access to the server) depends on the result of that write.
The point is that temporary network partitions happen and you can't guarantee delivery without expensive coordination. Sure, exactly-once is possible in environments where network partitions don't occur. Unfortunately, we don't live in that world.
In Europe we could always buy mix and match SIM cards with mobiles since the early days.
Just how you get the mobiles is different.
In some countries you can choose between paying the full price or having a kind of virtual contract, which means you get it cheaper but it is locked to the operator for two years. After which you get the code to unlock the mobile.
It never occurred to me until now that you could use drones to commit petty theft. Not even of someone else's goods--stolen credit card + random residential address + drone waiting to pluck the package off the doorstep = profusion of "brand-new-in-box" laptops into the secondary market.
I suppose the challenge would be hiding the drone, especially before drones become a ubiquitous consumer-good... but I suppose you don't have to meet them at home; any flat rooftop is a valid place to drop a package to go collect, or a valid "overnight parking" destination.