Wikipedia:Reference desk/Archives/Computing/2014 May 31

= May 31 =

Still can't install FSX
I'm trying to install Microsoft Flight Simulator X on my Dell XPS with Windows 8, but it just won't install -- I've tried completely resetting the computer, I've tried using a brand-new DVD, but the installer STILL crashes with a 1722 ("There's a problem with this installation package: A program setup did not run as expected..."), followed by a 1603 ("Fatal system failure during install")! And as I said before, the program used to install and run just fine when the computer was brand-new! What's going on here? How can this be solved? Thanks in advance! 24.5.122.13 (talk) 00:38, 31 May 2014 (UTC)


 * I think you are more likely to get a useful answer to your question if you ask it on a FSX forum. AndyTheGrump (talk) 00:47, 31 May 2014 (UTC)


 * Thanks for nothing -- I've found the technical support page for FSX (which took some searching -- the Microsoft people stuck it way in the back of their site), and I'm now going through their troubleshooting steps one by one. 24.5.122.13 (talk) 03:23, 31 May 2014 (UTC)


 * Be nice to AndyTheGrump; he was trying to help by saving you some time. This problem is probably specific to FSX, and among the people who read this forum it's unlikely that any of them know more about it than you do. You might get lucky, but you're much more likely to find a fellow sufferer on a forum frequented by FSX players. -- BenRG (talk) 18:37, 31 May 2014 (UTC)


 * Yes, it says on the site that the problem is caused by a conflict during the font registration step -- it also outlines several ways of how to solve it, and has a link for downloading a special tool for one of the repair methods. 24.5.122.13 (talk) 00:20, 1 June 2014 (UTC)
 * Odd, I had no issues installing it on Win 8.1 Palmtree5551 (talk) 01:54, 1 June 2014 (UTC)
 * I didn't either at first, but when I installed Java it froze completely and hasn't worked since then despite all my troubleshooting. I'm pretty sure the current problem with installation is because I couldn't uninstall it properly at that time -- I tried removing what was left of it using its native uninstaller, System Mechanic and the Control Panel, but nothing worked (even AFTER I removed Java), so I eventually just put it in the recycle bin. And THAT probably makes it a registry problem of some kind -- but that's odd, considering that System Recovery (which is what I had to do to even get the installer to run again at all) is supposed to completely roll back the registry to factory state, right? 24.5.122.13 (talk) 05:19, 1 June 2014 (UTC)

Rackmount 1U Chassis — without Drive Bays
Hi.

I'm building a cluster of six diskless servers using EATX dual-socket motherboards, which are to boot over iSCSI via onboard 1GbE LAN ports.

The problem is that I need a bay-less 1U chassis for each motherboard, but all the chassis models I've found so far have either 2.5" or 3.5" drive bays. Although the motherboard can be mounted in such a chassis, the presence of unused drive bays adds unnecessary depth and weight to each server.

I had considered getting a chassis such as this, and sawing off the front four inches of the chassis, but this option is not so suitable as it also removes the Front Control Panel — which shows useful information like LAN activity and PSU health.

Any pointers in the direction of a 1U chassis without drive bays will be much appreciated. 183.90.37.251 (talk) 06:26, 31 May 2014 (UTC)


 * I contacted ABMX about a custom rack mount chassis requirement a while ago, and they were happy to quote me on it. We had already found a better solution by then, so we didn't go with it, but they're worth checking out. I notice that they have a line of short-depth servers featured right now, so one of those may meet your needs without a custom chassis.  K ati e R  (talk) 15:08, 3 June 2014 (UTC)

Encrypting and encoding
What's the difference between encrypting and encoding? Surely encoding is just encryption were everyone knows the key Omega30 (talk) 10:59, 31 May 2014 (UTC)


 * Both change data from one form into another; but the intention of encryption is to make the output unreadable to those without credentials. Encoding is shifting form for another reason, such as to allow something to be stored on a given medium or transmitted using a given technology. 87.115.60.38 (talk) 11:23, 31 May 2014 (UTC)


 * Strictly, encoding is converting plain text into codewords (where only the desired recipient has a copy of the codebook, one hopes) whereas encrypting is usually done at bit or byte level using a cipher in classical cryptography, but this distinction is difficult to maintain in modern usage. I think any distinction in current usage will depend on the context.    D b f i r s   11:26, 31 May 2014 (UTC)


 * In computer science, though, "encoding" tends to refer to line coding and channel coding schemes where, as the original poster said, "everyone knows the key". I think that in the usual CS usage, encryption is a special case of encoding. -- BenRG (talk) 16:16, 31 May 2014 (UTC)


 * Yes, I'd forgotten that this is the computing desk, not the language desk.   D b f i r s   21:06, 31 May 2014 (UTC)


 * There's also some distinctions due to what is being encoded, and in what system the encoded words inhabit. Compare e.g. character encoding, which feels a bit more like "encryption with known keys" to the Text_Encoding_Initiative, which is mostly interested in semantic encodings, and feels rather unlike encryption to me. Also, encryption is usually designed to be hard to decipher without the key. However, a skilled cryptanalyst could easily "break" e.g. ASCII encoding, if given a decent corpus of encoded text files but no key. SemanticMantis (talk) 15:28, 2 June 2014 (UTC)


 * I'm not sure but I think that with public-key cryptosystems (e.g. RSA), having a decent corpus of encrypted text won't help. If you don't have the secret key then your only option is to take the public key and spend "enough CPU-power" to find the associated secret key. But even with a lot of CPU it would still take a very long time. My 2 cents. Ekkt0r (talk) 10:44, 5 June 2014 (UTC)

quantum computing
Does quantum computing operate at the subatomic level? Remember The Fallen 1914 (talk) 16:03, 31 May 2014 (UTC)


 * Quantum computing lists a bunch of kinds of quantum computing hardware, some of which store qubits as electron spin states for example. I suppose you could say that they operate at the subatomic level, but you could probably say that about ordinary silicon computers too, since the signals are carried by electrons (and electron holes), not by atoms. -- BenRG (talk) 16:26, 31 May 2014 (UTC)
 * Regarding conventional modern computers: I would say "the electrons are carried by the signal" - not the other way around - because CMOS digital electronics use voltage for signaling, and treat the movement of electrons in the circuit and inside its constituent transistor devices as a parasitic. More specifically, high speed computers transmit signals at the speed of light, which is a lot faster than the speed of electrons in copper wire or semiconductor metal-interconnect routing.
 * Regarding "quantum computers" - reiterating what I wrote in December, I think "quantum computer" is very poor word-choice. Exactly what do you think gets quantized in a "quantum computer"?  Almost everything inside a conventional digital computer is already quantized: the time-domain, the voltage levels... even the conceptual framework that we call boolean algebra are quantized, in that the math works in the realm of discrete operations on integer quantities.  Emergent mathematical properties, like quantum entanglement, are not specific to computers that use electron-spin to store their bits.  Nimur (talk) 19:28, 31 May 2014 (UTC)
 * I'd say that sound is carried by air. I wouldn't say that air is carried by sound, or that movement of the air is an undesirable side effect of the transmission of sound. Voltage is a function of the density of charge carriers, so a change of voltage is a motion of the charge carriers.
 * "Quantum computer" is not a very descriptive name. Neither is "computer", or "quantum". You might as well complain about poor word choice in every question about computers or quantum mechanics. The crucially quantum feature of quantum computers is not quantization of anything, but (nonclassical) entanglement of the bits. This is definitely not found in conventional computers; the reason we have no working quantum computers is that it's virtually impossible to preserve the entanglement for long enough to do a nontrivial computation. -- BenRG (talk) 17:15, 2 June 2014 (UTC)
 * A critical difference between sound-waves and electromagnetic waves is that sound-waves are longitudinal waves representing coherent, bulk movement of atoms. Electromagnetic signals propagate as transverse waves of electric and magnetic fields.  When those signals are guided through copper wire or a similar material, the electromagnetic wave couples with the electrons in the form of an induced electric current.  So, I would not say that air gets carried by sound; but I would say that (in some conditions) electrons are carried by an electromagnetic wave.  Nimur (talk) 19:04, 4 June 2014 (UTC)
 * Electrical signals are carried by longitudinal pressure waves. Electricity doesn't propagate in a vacuum. See hydraulic analogy. -- BenRG (talk) 16:51, 6 June 2014 (UTC)