Why The Moon?

[calbiterol]

That's what I think, even though I've been arguing about the small stuff here, too...

[Spyman]

However, Spyman, I disagree with you when you say the moon is a good step. Cut out the middleman. Go for Mars directly.

If we could make it to Mars directly that would be really nice.

 

But lacking a lot of knowledge about all the technical tasks with spacetravel and building extraterrestrial bases, I simply vote for the easiest problem, (to be on the safe side).

 

Learn to crawl before walk and walk before running and so forth...

[Ndi]

To be honest, my DVD-player crashed a few weeks ago, and it's not some piece of crap from Wal-Mart. Every component has the possibility of crashing - whether or not it is a software issue. Overheating, miscalculating, manufacturing defects.

 

OK, let's do a headcount. How many active components in your house? How many clocks? 10? TV, DVD, camera, phone, car? Do you drive a car? Know how many active processing units a car typically have? Let's estimate a total of 30 to be safe. A few weeks, 30 devices. I'll skip the fact that the DVD runs a decoder that can actually have a software issue that is not update-able from Huston. Few weeks (say, 3?) gives us a 21 day window, 30 devices, averaging MTBF(you should be familiar with that) at 630 days, roughly 15.000 hours.

 

This is for fire-and-forget-the-sucker-who-already-payed stuff that is actually OK to fail from time to time. Not carefully tested hardware billions ride on.

 

Do you have any idea what happens if your ECU (car computer) goes crazy on highway speeds? No, scratch that, I'm afraid you'll interpret that and tell me you drive for years. Let me remind you. Power steering is engine-powered. Power brakes are engine powered. Like many other systems in the car, such as stability, ABS and so on. Failure means that the car jolts violently and uncontrollably at 200 KPH, and you are out of brakes, helps and soft steering. If in a curve, the steering suddenly jolts as the wheels straighten.

 

Did that happen to you? Does that happen often? No. Why? Because people pay attention and test it first. MTBF on that is huge. Cars 20 years old, sporting 5 computers (real computers, with CPUs and flash storage, not clocks) never had a fail point.

 

Frankly, this kind of offended me. I work with computers and file storage every day

 

It is your right to feel offended, for whatever reason. The fact that you work with computers doesn't make you an expert. Don't try to pull this into the flame area, a) because it was a generic reply and b) because I really feel you are mistaking.

 

We'll elaborate on being right.

 

I also know that this is due to the fact that I have a number of different working copies spread across multiple media and computers, some of which go through a hefty everyday beating.

 

Then you also realize this has nothing to do whatsoever with a Mars mission, they don't use the same architecture.

 

I certainly know what I'm doing - I've been programming computers since I was 11, much less using them.

 

Again I say, I never said you never saw a computer, I said that these things work and they have incredible uptime. They run file systems over the Internet if needed be and there are storage solutions that are virtually indestructible.

 

Also, I know exactly how to force a file to corrupt itself.

 

This is relevant how? "Can be destroyed" and "will not self destruct for an estimated amount of time" are different concepts.

 

Multi-mirrored systems are much more difficult than non-mirrored, but certainly not impossible. Viruses take out systems - even multi-mirrored ones - on a regular basis.

 

You are mixing up software corruption and hardware failure. If you write gibberish, you read gibberish. Viruses are no concern on such a mission. Viruses don't run on the MarsRoverFirmWare.

 

As far as non-mirrored goes, I'm currently running Windows on an extremely unstable NTFS file system, and I'm waiting for it to catastrophically fail (again).

 

I'm sorry, I really try to stay flame free, but nobody will *ever* convince me you are an experienced professional when you claim that a self-healing, stable file system that runs servers with huge uptime and load is just a barrel of fun with you. Millions test that FS every day and it's been deemed stable. When a million people walk downstream and you are the only one going the other way you are either a visionary or you should take a look at your strategy again.

 

I have more experience with dead file systems than I would ever want to have. If there is one thing I've learned in my life, it is that EVERYTHING has the potential to fail.

 

Quit twisting my words. I never said a mirror is unkillable, I'm saying that given sufficient attention it will not fail on it's own to a point where data is unrecoverable KNOWING it's not a PC, it's not a commercial product and they don't use failing disk. I'm saying that a NASA-designed storage CAN make it safely and reliably to Mars and back. Not that it never made it before and much farther away on deep space crafts.

 

You somehow made it look like I'm accusing you of being incompetent because your NTFS gets mangled data when you work from different unreliable and/or virused computers.

 

The company where I work has huge databases, gigabytes in size, being accessed continuously for over 10 years now, with logs going back even more years. No catastrophic failure ever occurred, nor are any of those mangled. It CAN be done.

 

No virus, either. There are such things as routing, DMZ, firewalls and bridges.

 

--

 

RAID mirrors can be made quite inexpensive. Flash disks that are commercially available and you play football with have over 1.000.000 hours MTBF. One 1Gb costs around 12$, complete with shiny handle, nice connector and stuff. How much of a mirror can you buy with NASA's money and a discount?

 

Let's take a look at someone who's better than both of us, so we'll lose the competitive component of the argument. Microsoft states:

 

This expression can be formally derived (see for example [MGB74]). Its intuitive meaning is that the mean time to failure of the mirror is the mean time till the first failure MTBF/2 multiplied by the inverse of the probability of a second failure during the repair time, which is equal to MTTR / MTBF. With a 5 year MTBF and a 3 hour MTTR, the mean time between failures of a mirrored disk, MTBFmirror, will be more than 30,000 years!

 

Two disks, in mirror, with realistic 3-5 years MTBF, runs 30.000 years. And that's for a 25.000 hours MTBF (roughly 3 years continuous use). Use 1.000.000 hours MTBF flash and recompute. Then use 100 flash disks. What's the MTTDL now?

 

Let's also add that a good flash disk has a rated 1500 Gs Operating Shock. No vulnerability to magnetic field. No air cushion heads. No thermal recalibration. Virtually no cooling. Ni instant power consumption for seeks and spins.

 

So it's settled. Mirrored flash disks as well as processing units are reliable enough for the trip.

[D H]

Let me remind you. Power steering is engine-powered. Power brakes are engine powered. Like many other systems in the car, such as stability, ABS and so on. Failure means that the car jolts violently and uncontrollably at 200 KPH, and you are out of brakes, helps and soft steering. If in a curve, the steering suddenly jolts as the wheels straighten.

 

Well, it does happen. Computer-induced ABS failures have plagued Ford and General Motors. Software glitches have caused the Toyota Prius to stall. Lawyers see a pot of gold in this arena. There are now lawyers who specialize in problems related to automotive computers.

 

The simple fact of the matter is that avionics failures remain one of the leading causes of spacecraft failures. This is particular so with unmanned missions, which do not undergo the level of scrutiny to which human-rated spacecraft are subjected.

[calbiterol]

First of all, read DH's post. I won't touch on the automotive example because he did that already.

 

Second, no, I do not know what MTBF is. I'm not a Nasa employee. I don't have a degree in aerospace engineering (yet), or a pilot's license (yet), or am even in college. I'm a high school senior (grade 12, since "KPH" tends to mean "high school" is not used as a term). I was in a bad mood when I responded, and for that initial frustration, I apologize. I did not think you were flaming. I just despise being thought less of because I am (physically) younger than others - which is why I like online forums so much. Again, I may have been angry in the last post, but it was not totally because of you, and so I apologize. Now, back to the meat.

 

Then you also realize this has nothing to do whatsoever with a Mars mission, they don't use the same architecture... This is relevant how? "Can be destroyed" and "will not self destruct for an estimated amount of time" are different concepts... Multi-mirrored systems are much more difficult than non-mirrored, but certainly not impossible. Viruses take out systems - even multi-mirrored ones - on a regular basis.

All I was saying is that files can be corrupted fairly easily. You had previously said that files are very difficult to corrupt; I was refuting that specific argument. I was not applying this to the Mars mission concept, and I should have stated that at the time.

 

 

I'm sorry, I really try to stay flame free, but nobody will *ever* convince me you are an experienced professional when you claim that a self-healing, stable file system that runs servers with huge uptime and load is just a barrel of fun with you. Millions test that FS every day and it's been deemed stable. When a million people walk downstream and you are the only one going the other way you are either a visionary or you should take a look at your strategy again.

Actually, Microsoft will, if you want me to send you a link. I had a known software issue, for which Microsoft offered limited help. I did not say that NTFS itself was unstable, I said that my build of it was. This was certainly not (directly) because of Microsoft. In rare cases, a certain series of circumstances can lead to corruption of the file system. That is what happened. For 4 months, every tim I tried to boot Windows I got the blue screen of death. When I finally managed to force Windows to boot, the file system was still unstable (albeit useable), and my system would frequently crash and was permanently running at the half-processor-speed associated with power saving on laptops. I was not hit with a virus or other form of malware. I finally got around to reinstalling Windows and repartitioning my hard drive over the Christmas hols, and now my computer is again working fine.

 

 

Quit twisting my words. I never said a mirror is unkillable, I'm saying that given sufficient attention it will not fail on it's own to a point where data is unrecoverable KNOWING it's not a PC, it's not a commercial product and they don't use failing disk. I'm saying that a NASA-designed storage CAN make it safely and reliably to Mars and back. Not that it never made it before and much farther away on deep space crafts.[/qote]

If I twisted your words I apologize, but it was due to a misunderstanding, not willful intent.

I'm not saying that NASA-designed storage cannot make it to Mars, or the Kuiper Belt, or what have you. I'm saying that it has the potential to fail, and that Nasa-designed hardware, software, and/or firmware have all failed in the past. For examples: Apollo 13 was a hardware issue, as was the fire on the Gemini craft that killed all inside (while still on Earth, I might add); The Mars Polar Lander was deemed a software issue (lack of consistency with units). I will admit that I cannot think of a firmware failure off the top of my head; hence "and/or."

 

You somehow made it look like I'm accusing you of being incompetent because your NTFS gets mangled data when you work from different unreliable and/or virused computers.

Well, yes I did, because that's exactly how it felt to me. Whether it was from your specific choice of words, for a misconception or misunderstanding, or just by dumb luck, is irrelevant, because neither of our actions can be undone.

 

The company where I work has huge databases, gigabytes in size, being accessed continuously for over 10 years now, with logs going back even more years. No catastrophic failure ever occurred, nor are any of those mangled. It CAN be done... No virus, either. There are such things as routing, DMZ, firewalls and bridges.

Again, I am not arguing that it cannot be done. I am arguing that there are times when these things have failed.

There is, after all, a reasion that the people who designed and programmed these craft are "sitting on the edge of their seats" until the mission was more or less successful.

 

 

What's the MTTDL now?

To be totally honest, I understood very little between this and the last quote. I'm guessing that MTTDL means "minimum time to data loss," but the other acronyms are beyond me; therefore, so is the math, etc.

 

So it's settled. Mirrored flash disks as well as processing units are reliable enough for the trip.

I have no argument there; again, I haven't been arguing that it cannot be done. I am very much an advocate for extraterrestrial exploration and I find no reason to believe that a straight Mars shot would fail, given enough time and research for preparation. I think that a moon base would be a valuable asset in that preparation, as well as a more cost-effective manufacture and launching point than the surface of Earth.

 

_____

 

 

Though I have enjoyed our debate, Ndi, I must point out that we have both gone far beyond the scope of the original argument. The question was originally "Why the moon?", and we now sit arguing over the viability of hardware systems for extraterrestrial exploration. As hypocritical as this may sound, let's make an effort to bring this debate back to the topic.

 

Cheers,

Calbit

[swansont]

Second, no, I do not know what MTBF is.

 

Mean Time Between Failure

[Ndi]

Well, it does happen

 

Indeed. It is, however, rare enough, which was my intended point.

 

Second, no, I do not know what MTBF is

 

As swansont pointed out, it's Mean Time Between Failure and has nothing to do with NASA, it's how you measure hard drive reliability. 25.000 hours MTBF means that on average it goes south after roughly three years (3x365x24 hours). Another acronym used was MTTDL (indeed Mean Time To Data Loss). If you have a mirror, one disk fails so another is used. You don't lose data unless the second fails too, while the first is being rebuilt. This gives you a window of -say- 15 minutes for the secondary to fail. The window is 15 minutes every 3 years. Miss it and a new disk is in place. A third hard drive means that two HDDs must fail in the 15 minute window and so on. A threesome is -for all intents and purposes- invincible. (quit thinking dirty)

 

Apollo 13 was a hardware issue[...]

 

Decades have passed since, and we don't do things the same way. Components are *much* more reliable now; Just take the warning bulb versus million hours LEDs, in addition to storage reliability. Also, back then a surge was a disaster on its way, now everything is stabilized and protected.

 

I'm a lot more concerned on mechanical component failure - like a gasket.

 

 

beyond the scope of the original argument

 

Indeed.

[D H]

Here are a couple of articles/editorials that pertain to the question raised in the original post.

 

This one is by NASA Administrator Michael Griffin.

http://www.chron.com/disp/story.mpl/editorial/outlook/4485901.html

[ The Columbia ] report declared that if we are going to send humans into space, the goals ought to be worthy of the cost, the risk and the difficulty. A human spaceflight program with no plan to send people anywhere beyond the orbiting space station certainly did not meet that standard.

 

President Bush responded to the Columbia report. The administration looked at where we had been in space and concluded that we needed to do more, to go farther. The result was the Vision for Space Exploration, announced nearly three years ago, which commits the United States to using the shuttle to complete the space station, then retiring the shuttle and building a new generation of spacecraft to venture out into the solar system. Congress has ratified that position with an overwhelming bipartisan majority, making the Vision for Space Exploration the law of the land.

 

On the Space Review,

http://www.thespacereview.com/article/791/1

A Moon full of opportunity

 

NASA gave six reasons for going back to the Moon when only one was needed

 

The 2nd Space Exploration Conference held December 2006 in Houston outlined several reasons for a human return to the Moon. Remarkably, some complain that the reason for going to the Moon is still unclear. Possibly the sheer scope of the envisioned surface activities diffuses its impact. Almost 200 activities were described for the Moon, grouped under six major “themes” (as the agency calls them), including settlement, global cooperation, science, and preparation for Mars.