Wikipedia:Reference desk/Archives/Computing/2010 February 20

= February 20 =

Installing Windows 7 on a HDD with files on it?
I have a HDD with a few hundred gigs of media files on it (it does not contain an OS). I want to put it in my new computer. How can I install Windows 7 on it while keeping the files? F (talk) 04:32, 20 February 2010 (UTC)
 * I've done a couple of Windows 7 reinstallation jobs by just opting for a fresh install, and let the setup wizard move my docs and other stuff to a folder named windows.old. Blake Gripling (talk) 06:48, 20 February 2010 (UTC)
 * Does this work if there's no Windows previously installed on it?F (talk) 08:14, 20 February 2010 (UTC)


 * This is rather lame, but I would trust the Windows 7 installer to not partition or format or otherwise erase the drive without asking. Though I would back up before proceeding.  Comet Tuttle (talk) 17:30, 20 February 2010 (UTC)


 * I trusted Windows 7 installer not to place a pointless 100MB partition on a completely different drive from the system drive without telling me, but it did that anyway. I had to reinstall Windows with all other drives unplugged because that partition contained boot information and I had other plans for that drive. The fact that the other drive was part of a RAID array that had been transferred from another computer didn't bother it either (the array was going to be destroyed anyway, because arrays generally can't be moved like that, but it's not the point. I haven't done precisely what the OP is asking, so I don't know for sure, but I wouldn't trust it. In an ideal situation, I'd suggest buying a smaller drive with great performance to use as the system drive, but perhaps the OP's situation precludes that. I know mine does! Seans Potato Business 20:46, 20 February 2010 (UTC)
 * I thought about buying a SSD but my budget does not stretch that far. I'll partition the hard drive first then. Thanks. F (talk) 00:24, 23 February 2010 (UTC)
 * There's a big difference between the two however. The first one didn't destroy any data, nor did it have any direct ill effects unless you the user did something you weren't supposed to (like disconnect the drive or delete the partition the second of which Windows 7 wouldn't let you do so you'd need to do it with some other OS). And if you did the former, all you need to do is reconnect the drive, if you did the later well you could definitely mostly recover, it would just take a bit of effort (and still no lost data). More importantly, not something your average user is ever likely to encounter (read on).
 * The partition you refer to BTW was created I presume because Windows 7 by default always creates a recovery partition if it can, and the recovery partition is always going to be on the boot drive and I presume the disk (or array) was seen by Windows as the first boot drive. This isn't exactly new to Windows 7 IIRC, while Windows XP didn't create a recovery partition it would use the first boot drive (or current boot drive/partition if it existed) as the main boot drive & partition where it put the bootloader.
 * It isn't too surprising either, when Joeclueless installs Windows they want it to bootup, they don't want their computer to fail to bootup because they installed Windows to a different drive and their mobo is trying to bootup from the first boot drive. Hopefully anyone using Windows is aware MS nearly always prioritises for Joeclueless.
 * For Joenonclueless, it's always wise to disconnect any drive other then the one you want the OS to be installed on to ensure the bootloader etc goes to the right place and nothing else you don't want happening goes on (I've done it with Linux, FreeBSD, Mac OS X & particularly all variants of Windows with as I've already mentioned have always had a tendency to put the bootloader in places I don't want it). Even more so if you plan to remove the drive some time soon. (Actually I'm not sure if I'd ever connect a drive I intended to remove when installing an OS).
 * The Windows installer wasn't of course designed to guess where drives came from nor guess that the user is intending to delete them soon, it sees the first boot drive (or perhaps it was the existing boot drive/partition) and uses it for the boot. Also from memory it was fairly obvious that Windows had created that recovery partition (unless perhaps you use the wizard mode) so it's not like something you only find out after you installed Windows unless you weren't paying attention while installing (which is always a bad idea). In fact in that way it's probably better then the past since from memory it was never that obvious where Windows was going to put the bootloader if you didn't have experience.
 * In other words, annoying but understandable behaviour and something the user with an odd setup needs to and can easily account for.
 * In terms of the ops problem, it's fairly obvious that it's incredible bad design to have your OS delete data without telling you and if it does do that it is something that clueless users could easily do and in some cases would really, really, really piss them off. Therefore something which any installer designer would generally account for. And indeed I have installed Windows 7 (and other versions) on existing partitions before when testing various things and it doesn't generally delete data. Of course it goes without saying if you have important data, you should always have a back up, even more so when carrying out major changes like installing a new OS.
 * Nil Einne (talk) 03:19, 23 February 2010 (UTC)

How to revert (back out) Flash plugin update
I have been experiencing a very annoying problem where the PC (Dell, running Windows XP) will totally freeze up and nothing works except moving the cursor (sometimes right-clicking on the task bar brings up the menu, but clicking on those menu items doesn't work). Even Ctrl-Alt-Delete to bring up Task Manager won't work. I have to hold down on the power button to turn off the PC. Unfortunately, I have had to do this several times already. I did upgrade Adobe Flash today (well, yesterday now) to 10.0.2.45 when I got the auto upgrade request (I see from Adobe's site that this "critical" upgrade was released a week ago; maybe I put it off before?), so I am now working on the premise that it is the culprit. I have now disabled the Flash plugin in Firefox, and so far I am O.K. But this is an unacceptable solution since too many sites I use, like Hulu or Youtube, use Flash. Therefore I was wondering if there was any way to revert or back out this new version of Flash?

P.S. It is true I recently (either the same day or the day before; I forget) also upgraded Firefox to 3.0.18. I actually suspected the Firefox upgrade at first, but then I experienced the problem while using Internet Explorer, so I don't think so.

Thanks in advance! TresÁrboles (talk) 07:35, 20 February 2010 (UTC)
 * http://www.oldversion.com/Macromedia-Flash-Player.html F (talk) 08:14, 20 February 2010 (UTC)
 * Thanks, F!, but is there a way to just go back to the previous version? The latest version on that site is Flash Player 9. There is a message on Hulu saying it requires "Flash Player 10.0.22 or higher."  For all I know Youtube requires 10 also (it doesn't specify).  Also, am I to assume the Flash Player download includes the plugin?  TresÁrboles (talk) 08:41, 20 February 2010 (UTC)
 * Oops, I had the version of my current updated (and suspect) Flash wrong -- it should be 10.0.45.2 . TresÁrboles (talk) 08:44, 20 February 2010 (UTC)


 * You might try using Windows Restore.--Phil Holmes (talk) 10:12, 20 February 2010 (UTC)


 * Here are all of the versions of Flash Player 10.--Drknkn (talk) 21:40, 20 February 2010 (UTC)
 * Thanks, Drknkn! Unfortunately, I have the sinking feeling now that Flash may not be the problem because the PC froze twice when watching a DVD last night without any other apps running. (I had no problems the whole day today though... but then again I didn't watch any DVDs or enable the Flash plugin.)  I really hope it's not hardware starting to fail.  TresÁrboles (talk) 08:33, 22 February 2010 (UTC)

is it possible to remove just one URL from Chrome's autocomplete?
we rely on Chrome's URL auto-completion when we start typing into the URL bar, but wonder if it is possible to selectively remove a single URL that we do not want offered anymore as we type? (ie we don't want it displayed while typing). thank you. 84.153.236.194 (talk) 09:53, 20 February 2010 (UTC)
 * As far as I know - no - not accessible from the front end. (You can delete single history items) - but it only works sometimes.
 * The help forum is here http://www.google.com/support/forum/p/Chrome?hl=en
 * 87.102.67.84 (talk) 11:30, 20 February 2010 (UTC)
 * You can also request it as a feature via the google chrome site.

vu meter
I got a mixer that's got this vu meter which looks like this (http://www.brindeamusica.net/products/Audio/mmistura/Yamaha_MG_16_6_FX.jpg). I'd like to ask for some kind of advice on how to use it or how to understand it rather. I know that the sound should rarely or never get to the red peak bar section, but most of the time should it be like in the middle of it (the green area, usually marked as 0 db) or that is above it which is in my case orange and its marked as +3 +6 db

I also assume that if the signal is far below 0 db around -10 -20 on that meter is bad and the sound wont be strong enough and punchy, so where should i keep it at? —Preceding unsigned comment added by 80.83.238.11 (talk) 10:03, 20 February 2010 (UTC)
 * First of all one important factor is whether or not you are ouputting to a digital medium - if you go over the top threshold with digital the distortion will be unacceptable (clipping) - and you simply must avoid it.
 * For analogue the balance is between a high signal and increasing distortion, and a low signal and a low sound level with increased background noise levels.
 * Typically you need to experiment (or read the manual) to find out what works best.
 * Also OdB is not guaranteed in general to be the cut off level - it depends on the calibration.


 * The usual aim is to record at at a high a level as possible but avoiding the peak levels going far into the 'red zone' since that would cause compression on transients.87.102.67.84 (talk) 11:36, 20 February 2010 (UTC)

Infinite type hierarchy
In my term paper for a course in comparative programming languages that I finished in December, I proposed a C#-descended language in which the Type class would be made generic, and that every type T be an object of type Type. This meant that Type itself would be an object of type Type, which would be an object of type Type>, etc. Would this infinite regress make it possible to write a program whose vtables could not be generated at compile-time? Would it necessarily be possible to compile such a program by using a heap-dynamic vtable that could expand at runtime? Neon Merlin  10:22, 20 February 2010 (UTC)
 * Clearly if you don't do anything to stop multiple analysis of the type of a 'thing' you will can an unworkable compiling model. Is there any reason why type analysis shouldn't stop at Type - ie stop once the type is a sub type of your single built in generic.
 * Anyone can write a program that performs infinite regress - the trick (I think) is to avoid it - not make it a feature ?! :)
 * Dynamic tables (or other dynamic structures) expressing program state and data (of all sorts) expandable at run time are almost certainly the best way to deal with any dynamic language. (Is that too general an answer?)
 * I'm fairly certain that to answer your question about compile time do-ability using vtables you'd have to be more specific about what possibilities there are about operations on the different types. - eg how flexible.
 * If your C# descended language remains sufficiently non-dynamic at to be compiled at runtime then won't it be impossible to generate the infinite regression you describe in an actual piece of code (since you can't write infinite code) ?
 * 87.102.67.84 (talk) 11:41, 20 February 2010 (UTC)
 * By the way what was your base type class going to implement? All I can imagine is a base type of binary, with operations such as XOR, AND, shiftleft, testbit, etc.. and sub types being implementations of Int, Char, Float etc with allowed operations (and castings) derived from the 'binary' operations? Is that anything like what you had in mind?? —Preceding unsigned comment added by 87.102.67.84 (talk) 12:18, 20 February 2010 (UTC)
 * Usually a type of a type is called a kind. There are languages with the features like what you mention.  Wikipedia's coverage of this subject is not very good and I don't know that much about it myself yet, but the book Types and Programming Languages is the usual place to start.  I'm not sure how it would interact with vtables and C#'s features.  You might also look at the wikibooks module on the logical structure of Haskell's type system, which supports higher-kinded types: Haskell/Category theory.  Haskell does not, however, support full-out dependent types which is where elaborate kinding mechanisms tend to come into play.  75.62.109.146 (talk) 02:49, 21 February 2010 (UTC)

How do the public rate the different web domain suffixes in terms of trustworthiness?
There are many web domain suffixes, such as com net biz and so on. Have there been any surveys about how reliable and trustworthy the public would consider websites or web-businesses having the different suffixes? Thanks 78.149.241.220 (talk) 16:13, 20 February 2010 (UTC)


 * Those domain suffixes are technically called Top-level domain names. (This may help you refine your searches).  Our article, Top-level domain, links to this news website, Internet Infrastructure - Top-Level Domains, which has a lot of relevant articles.  There has been much recent debate in the IANA/ICANN community regarding new proposed top-level domains, especially with respect to TLDs encoded in alternative character sets.  The consensus is not clear yet - but as more TLDs are introduced, the hazard of ambiguity becomes larger and larger, given that a DNS name is supposed to be globally unique and easily recognizable as a domain name.  Personally, if I see "wikipedia.org", I know without further context that it is a domain name - because "org" is an easy suffix to recognize.  Country codes also make sense, and they're easily recognizable.  Recently, IANA added a bunch of other TLDs which I consider "less good" because they are a lot more ambiguous.  There has even been talk of adding specific city names - e.g. ".vegas" and so on - because of the potential commercial value.  I think this is a terrible idea.  In the extreme example, imagine seeing a domain name like "orange.juice.is.delicious" or "this.is.a.good.restaurant" - how in the world would you know that's something you should type into your URL bar, and not just artistic use of punctuation?  In that sense, I have a great disdain for domain name hacks - they only serve to obfuscate and disrupt a system that is supposed to be easier than memorizing a 12-digit number.  In the worst case, I'll just go back to memorizing 12-digit numbers - in the old days, that's how we used to telephone our friends, and I'm sure it'll be a fine way to get to our news and information websites too.  Anyway, I should refrain from ranting and just provide some references.  Here's a study called Legitimacy and Authority in Internet Coordination: A Domain Name Case Study - you might find it helpful.  Nimur (talk) 17:03, 20 February 2010 (UTC)
 * Well unless IPv6 takes off... Nil Einne (talk) 21:48, 21 February 2010 (UTC)

What I meant was, how do the public/consumers rank the trustworthyness of business-websites having different web-address suffixes? If you had to put the different suffixes in rank order according to trustworthiness, what would that rank order be? I'm *not* asking about the behind the scenes technical comittee stuff that Nimur has described. I have had a look at the Top-level domain article and its links but could not find anything relevant. Thanks 78.149.241.220 (talk) 17:17, 20 February 2010 (UTC)
 * The last link I provided is non-technical.  Are you specifically looking for summarized surveys of public opinion, as opposed to summarized analysis by internet designers?  Nimur (talk) 17:33, 20 February 2010 (UTC)
 * Yes! (I had mentioned the word "public" three times!) The last link, thanks for all your effort, but its not relevant at all. 78.147.195.219 (talk) 18:53, 20 February 2010 (UTC)
 * Sort of a flip-side is the prevalence of mis-trust. 50% of phishing attacks use .com names.  So, I would suspect that ".com" is both the most publicly trusted and the most publicly mis-trusted TLD.  Nimur (talk) 17:36, 20 February 2010 (UTC)
 * Well, I searched a little more, and couldn't find a survey like you are seeking. The closest I found was this 2009 Industry Brief from Verisign, which makes reference to qualitative surveys about business acceptance for IDN (non-ASCII) domain names.  But again, from the point of view of systems design, there is no validity whatsoever in people's willingness to trust a website any more or less just because it uses a .com or .biz top-level domain.  See for example, Perils of Transitive Trust in the Domain Name System for some technical insight.  Anybody can register a domain-name.  Very few TLDs actually enforce any kind of requirements - except notably the .gov and .mil TLDs, which are tightly controlled.  But beyond that, there's no reason to trust a particular website just because of its name.  In any case, public opinion is not always rational - so there may be some unfounded trend to distrust particular DNS names.  Nimur (talk) 19:10, 20 February 2010 (UTC)


 * The question is interesting, and I'm sure that the registrars would not be anxious to publicize the results of such a study, for fear of informing companies that their .biz addresses for sale are distrusted. (Or even, perhaps, the .net addresses.)  Nimur, you say there's no validity whatsoever in distinguishing between .com and .biz, but I think you have to provide a reference for that assertion.  If we take as a given that young companies fail more often, then as a vast overgeneralization, I think standardinstruments.com is more likely to be around in five years than standardinstruments.biz.  Even if I'm wrong, the original poster is asking about people's perceptions, and I've made mine clear.  The .com and .biz TLDs aside, I don't think it is a stamp of unreason for Americans to distrust e-mail directed at them from the .ng TLD.  Comet Tuttle (talk) 19:21, 20 February 2010 (UTC)
 * An interesting diversion is this SiteAdvisor survey which ranks the TLDs in terms of safety: the worst for malware in order are .info, .ro, .ws, .biz, and .cn, with .info taking the (dubious) top spot with 22% of all tested sites hosting malware or links to malware. The worst TLDs for spam (i.e., websites with "questionable e-mail practices") are .info, .cn, .hk, .net, and .ru, with a staggering amount of .info sites - almost 60% - falling into the questionable category.
 * Although it's clear by this point a survey to directly answer 78.'s question will be hard to find, if one even exists, I can try to answer the question with an anecdote. It seems to me that most non-tech-savvy users, at least in the US, distrust websites that are not .com, .org, .edu, .net, or .gov - probably the five they come across most in their daily usage. I have two e-mail addresses, both in my real name: one is at an internet portal and ends in .com, and the other one is at a German ISP's "Freemail" service and ends in .de. Even though (I think) it's pretty clear that both e-mail addresses are owned by me, occasionally when I send an e-mail from my .de address I will get an e-mail at my .com asking "was this really you?" Similarly when I'm talking to someone and I tell them to go to "acq.osd.mil", "fb.me", "youtu.be", "tagesschau.de" or even "itv.co.uk" - all perfectly trustworthy websites - I might get a funny look. If it's not "cnn.com" they won't want to go there without being prodded. Xenon54 / talk / 23:09, 20 February 2010 (UTC)
 * .edu and .gov are both actually restricted in who can register them to accredited US colleges/universities, and US government agencies, both of which the public has some sort of actual trust it. Websites on those domains are virtually guaranteed to be legitimate. Paul Stansifer 12:52, 21 February 2010 (UTC)
 * .mil is also tightly controlled. Note that "legitimate" here has a very specific meaning—that the organization sponsoring it is who they claim to be. It does not say too much about the accuracy of content, of course. --Mr.98 (talk) 22:19, 21 February 2010 (UTC)

2.5" ATA: 4 additional pins on devices
My Notebook contains a jack for 2×22 pins in a 2mm grid. I assume that this is a female 44 pin parallel ATA connector for 2.5" drives. When searching for 2.5" “IDE” hard drives, on every picture I found there are 2×(22+2) pins, with a gap between the 44 pins and the four additional pins.

What is the purpose of these four additional pins? Are they used to jumper the drive, or do they need to be connected, or do they belong to older or newer versions of ATA, so it doesn't matter if I connect them to anything or not? —Preceding unsigned comment added by Kalkühl (talk • contribs) 19:30, 20 February 2010 (UTC)


 * I assume you are not talking about the pins for the power connector? I which case it is likely they are the pins for selecting master/slave or other HDD configuation options. --220.101.28.25 (talk) 20:16, 20 February 2010 (UTC)


 * I don't think I'm talking about power pins, because it's 40 pins data plus 4 power, so power pins are already included, but for example http://www.reichelt.de/bilder/web/artikel/E600/HDD_25.jpg shows 4 additional pins on the very right, which from their position are not clear to not belong to the connector. I just wondered because on every image of a notebook hard drive with ATA connection these pins seemed to be at exactly that position, so I thougt it might belong to the connector. – Kalkühl (talk) 21:46, 20 February 2010 (UTC)


 * Per http://www.seagate.com/staticfiles/support/disc/manuals/notebook/momentus/5400.5/100468844b.pdf (p. 18) two of those pins are jumpered for 1.5Gb/s operation. ---— Gadget850 (Ed)  talk 23:16, 20 February 2010 (UTC)


 * That's for a SATA drive though, whereas this is a PATA drive. The extra four pins are most likely jumpers, analogous from desktop PATA drives. I do not know their specific functions but it should be explained on the drive itself. --antilivedT 00:20, 21 February 2010 (UTC)


 * Then see http://www.seagate.com/staticfiles/support/disc/manuals/notebook/momentus/5400_3/100398876c.pdf (p.22 or sect. 3.2). nb I got this by Googling for "Momentus PATA specs" --220.101.28.25 (talk) 02:46, 21 February 2010 (UTC)

Installing Linux Genuine Advantage
I'm not that much of a Linux person and I want to install Linux Genuine Advantage on Kubuntu 9.10 running in VirtualBox. I'm not stupid, I know that this thing is a joke but I just want to install it to see how it works and the virtual machine that Kubuntu resides in is a clean install so there nothing special about it that I could lose because of this. So how do I install it, not the source code but the one found under Download? --Melab±1 &#9742; 21:00, 20 February 2010 (UTC)


 * Do you know how to unzip a tar.gz file? You can use tar -xvf linux-genuine-advantage-1.0.0.tar.gz; and then run the script with make install.  Explicit instructions are found in the README file.  While that entire site is humorous, be aware that the install script and perl script do appear to actually functionally disable your computer by overriding your login configuration... so use with caution.  Nimur (talk) 21:15, 20 February 2010 (UTC)
 * (e/c) Download the .tar.gz to the Desktop, then extract it by opening a shell, running "cd /home/[username]/Desktop" and then running "tar -xvzf ./linux-genuine-advantage-1.0.0.tar.gz". This should create a new folder on the desktop called "linux-genuine-advantage-1.0.0". cd into it and run "sudo make install". (As far as I can tell the .tar.gz file retrieved from the "Download" link contains an already-compiled program, so "sudo make" is not required) Xenon54 / talk / 21:16, 20 February 2010 (UTC)
 * There is no compiled program - the entire system is a perl script that runs with root permissions. Nimur (talk) 21:19, 20 February 2010 (UTC)
 * I unzipped it and I changed the directory to the unzipped folder and entered:

sudo linux-genuine-advantage install

It prompted me for my password which I entered and I got this:

Could not open /etc/inittab for reading!

--Melab±1 &#9742; 21:31, 20 February 2010 (UTC)
 * You should be entering "sudo make install". Following this you should enter "linux-genuine-advantage" Xenon54 / talk / 21:36, 20 February 2010 (UTC)
 * But this message came from the Perl script not the Console itself. --Melab±1 &#9742; 22:13, 20 February 2010 (UTC)
 * After some Googling apparently there is no /etc/inittab om Ubuntu/Kubuntu/Xubuntu etc., instead it is replaced by /etc/event.d/rc-default. I don't know if this can be worked around (and if it can it probably isn't worth the effort). Now that I'm actually on an Ubuntu box I also found, in case others are confused like I was, that once you "make install" the script you must run "sudo linux-genuine-advantage install" to "completely install" the "service". This throws up the "could not open..." error above (although I also got it on my initial "sudo make install"). Xenon54 / talk / 22:42, 20 February 2010 (UTC)

How to turn off the pop-up minature image of page in Acrobat while viewing PDFs?
I am using WinXP. Whenever I scroll a PDF slightly, using the default Acrobat reader, I keep getting this very irritating pop-up box which shows a minature image of the page. How can I kill it please? I've looked through what I can find by right-clicking the mouse, but cannot see how to turn it off. I think I have the latest version of Acrobat reader. Thanks. Edit: more specifically, the pop up keeps popping up whenever I move the page up or down by placing the cursor on the solid-looking part of the scroll bar top the right, pressing the left mouse button, and then moving the cursor up or down. 78.146.167.216 (talk) 22:57, 20 February 2010 (UTC)

Routing packets from my computer.
Is there anyway I can route outgoing traffic from my computer, if I know the ip address and the port they are bound for (if that makes any difference)? 66.133.196.152 (talk) 23:14, 20 February 2010 (UTC)


 * Yes. It is trivial.  Open a connection to that IP address/port and send data to it.  It can be done in pretty much any programming language.  Just look for examples of sockets. --  k a i n a w &trade; 01:52, 21 February 2010 (UTC)


 * I suspect you want to re-route packets for some reason, or maybe just watch them. If you route them to an IP that's not expecting them then they'll just be dropped because the receiving machine's not expecting them. You might also be trying to drop those packets, or perhaps view them. I don't think you'd ask this question though if you had already set up a TCP socket able to receive arbitrary connections (unless you're using something like netcat, but then we're back to just viewing the traffic).


 * Could you be a bit more specific about what you're trying to do by rerouting? Shadowjams (talk) 07:55, 21 February 2010 (UTC)