Wikipedia:Reference desk/Archives/Computing/2010 June 23

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

June 23

Just built a computer but it wont turn on

After pressing the power button it doesn't do anything. The fans don't spin. No lights light up. It is absolutely quiet. I am unsure what the problem is. (power supply maybe?) Any help would be appreciated. Thanks. —Preceding unsigned comment added by 72.118.202.92 (talk) 08:04, 23 June 2010 (UTC)

Assuming you are sure mains power is actually getting to the PC (go check the fuses/circuit breakers in your house and the power cable leading to the PC), the first place I would look is the power supply. Many PC power supplies have their own internal fan as shown in this image, and if that isn't running the power supply unit itself is likely to be faulty. Check your power supply has the correct input voltage settings: 100-127V AC, 60Hz in the USA, which is usually selectable with the small red switch on the backplate. The small plugs that supply the power to the components inside your PC, carry voltages typically 3.3, 5 or 12V DC. Check each one with a multimeter to see if the correct voltage is being supplied. Make sure each plug is properly seated, the correct way round in its socket on the motherboard, disk drives, etc. and note that some can be a very tight fit. Astronaut (talk) 09:43, 23 June 2010 (UTC)

(edit) Due to the much better advice provided below, and me forgetting quite how a PC gets switched on, I've struck out my rather simplistic view of how a power supply can be checked. Astronaut (talk) 17:13, 23 June 2010 (UTC)

If you just built a computer and it isn't turning on I would be careful about assuming the PSU is faulty. The PSU will only turn on if the motherboard tells it to turn on so there could be plenty of reasons like the motherboard is faulty, you didn't plug the power switch in properly, the power switch is faulty (presuming the motherboard doesn't have a built in switch which you are using). Okay technically the PSU us always on by the fan doesn't usually run when it's in standby mode (a few keep it running for a few minutes after you turn it back to standby mode but I don't think I've ever seen any that turn it on when you first turn it on in standby mode although probably something does exist), there may be some lights visible on the motherboard, there may not be so without knowing more details about the particular set up, it's difficult to say whether the OP would be aware of the difference between PSU completely not working and motherboard never turns on the PSU. Depending on the PSU if something is shorting you may also get something like this (although usually the PSU will turn on briefly and then go off quickly albeit depending on how much attention you paid and where the computer is you may not notice). Of course if it's the standby that's shorting then I would guess this would usually fuse it fast enough that wouldn't notice.

In terms of testing, you don't really have to worry about disk drives. I would connect only the motherboard and perhaps video card if it has a power connection first (disconnecting anything else plugged into the power supply). In fact, often you'd disconnect everything but the video card (unless it has built in video in which case even that is unnecessary if you have one), RAM and CPU from the motherboard plus the power switch if needed and of course the power. There's also little use trying to measure voltages if the PSU isn't even on (other then the standby lines).

If all else fails, you can disconnect the PSU from the motherboard and video card, connect it to a few disk drives, or fans or something and then short the power on (usually green) and ground (usually black) lines and see if the PSU comes on properly as discussed in our ATX PSU article. If it does, try measuring the voltages if you can. (If it doesn't then yes your PSU may be faulty presuming you've made sure that the PSU is actually getting power and any switch on the PSU is on.)

If you have a good PSU, you shouldn't really need to connect anything, it should come on even without them although the voltage regulation may be poor. (And as Astronaut mentioned the PSU should have a fan which would normally come on when it comes on properly so you should be able to tell if it's at least working somewhat without connecting anything.) For some cheap & crappy PSU however, I wouldn't discount the possibility of them either refusing to turn on, or even dying if you start them without any load so I'd be careful. If the PSU seems fine, I would look more carefully at the motherboard. You could try shorting the pins for the power manually if you're careful.

Nil Einne (talk) 11:15, 23 June 2010 (UTC)

Make sure the PC Speaker header (SPK or PCSPK or something like that) is connected up - the motherboard uses that to signal (by means of several beeps) if there's a problem (and if you can't hear the beeps you can't know what it's trying to tell you). Next make sure the power supply connector to the motherboard is properly connected, including the additional +12V2DC connector that most motherboards have (it's separate from the main connector, and is often located nearer to the CPU). Then make sure that the CPU fan is correctly connected to the appropriate header on the motherboard - most motherboard microcontrollers will refuse to let the system boot if they don't detect the fan (as a missing or bad fan would result in the CPU being very quickly cooked alive). In my experience these two mistakes are the most common causes of newly-built machines not starting at all - if they're okay, the PSU should power up, even if the RAM and CPU are entirely missing. -- Finlay McWalter • Talk 13:40, 23 June 2010 (UTC)

Another simple thing: I assume you're using the front-panel power button to try to turn it on, but first make sure that if there's a hard power switch on the PSU, that it's turned on. -- Coneslayer (talk) 13:51, 23 June 2010 (UTC)

And, of course, that the front-panel power button (PWR) is properly connected to the appropriate header on the motherboard (apologies for stating the obvious, but sometimes one can get caught up in byzantine hypotheses for a failure that's cause by something so trivial one has overlooked it). Ditto for the case reset (RST) switch (if it's accidentally shorted the motherboard may refuse to boot, thinking the machine is being deliberately held in reset). -- Finlay McWalter • Talk 15:00, 23 June 2010 (UTC)

Thanks for all of your help! I thought it was the power supply at first since it was completely dark. (But that's actually because the motherboard just doesn't have an led.) It turns out I didn't plug in the front panel connector for the power switch. —Preceding unsigned comment added by 72.108.156.138 (talk) 04:30, 24 June 2010 (UTC)

Finding and deleting disused DLLs

I've just done a scan of my HD using the freeware program DLL Archive by AnalogX http://www.analogx.com/contents/download/System/dllarch/Freeware.htm which searches for unreferenced DLLs.

I thought it would find either none or just a few. In fact it found over six thousand of them, the biggest of which were over 13MB in size. They must be using up gigabytes of space on my nearly full HD.

a) Are there any paths/folders where it would be particularly unwise to delete them?

b) Would deleting those not in the Windows folder be safe?

c) Is there any other practical means of finding disused DLLs? Thanks 92.15.17.9 (talk) 11:24, 23 June 2010 (UTC)

Here are some more similar programs, except the first one: http://www.softpedia.com/downloadTag/delete+dll 92.15.17.9 (talk) 12:32, 23 June 2010 (UTC)

I don't think such programs are terribly safe, and I wouldn't recommend relying on one. That's because of how DLL dependencies work. There are two mechanisms whereby Windows loads a dll. Firstly is by means of static dependency, where the information in an application's Portable Executable file lists all the dlls it depends upon (and they, being PEs themselves, list those that they in turn are dependent upon). That's how programs like the ones you're talking about work - they find all the executables, and then traverse the tree of dlls indicated (directly or indirectly) by these headers. It's tempting to assume that any DLL file on the disk that isn't reachable by this traversal is therefore "unreferenced", and so can be safely deleted. This isn't the case, because of the second mechanism Windows employs. DLLs can also be explicitly loaded by a program calling the LoadLibrary function; LoadLibrary takes the name of a dll and loads it when it's told to; that name isn't in the PE header of the executable, and so isn't found by the traversal discussed above. This might sound like an obscure way of doing things, but it's actually a very common pattern. Lots of programs will have a "plugins" folder (or several), and when they start they'll look in that folder, read the names of all the dlls they find, and then LoadLibrary each one (and so installing a plugin is as simple as copying its dll into that folder). The traversal misses all of these, classifies them as unreferenced, and so you'd be deleting things that are in fact used (but not statically referenced) and so break stuff. To answer your specific questions: a) clearly the windows directory and its children, but really no-where is safe b) no c) as described above, not really. Instead I'd recommend running a smart uninstaller like Revo Uninstaller Free and aggressively uninstall programs you don't use (smart uninstallers are better than dumb traversal, because they have some knowledge a-priori about the files used by many popular programs). But really I'd question that dlls are really taking up a significant part of your disk; it's almost always media files or the data files associated with applications that's really to blame. -- Finlay McWalter • Talk 13:18, 23 June 2010 (UTC)

There's probably other stuff on your disk that can be deleted first. You can start by clearing your browser cache, emptying your trash/recycle bin, clearing various folders called temp or tmp - c:\temp, c:\windows\temp (80+MB in mine!), and other places too. Astronaut (talk) 17:01, 23 June 2010 (UTC)

I use Ccleaner and CleanUp! all the time, so already doing that. I found three copies of hwxjpn.dll, 13MB each, 20 copies of ieframe.dll about 11MB (although some vary slightly in size), and many more repeated seemingly identical DLLs. I have 6719 unreferenced DLLs, dating back to 1994. BTW, I do frequent virus scans with four or five different scanners so they are unlikely to be malware. Its a pity that XP isnt designed to use just one DLL file rather than having many identical copies of them. I don't suppose there is any software that can do the second thing that FinlayMcWalter described? Thanks 92.15.3.0 (talk) 20:09, 23 June 2010 (UTC)

Just because you have two DLLs that are the same name doesn't mean they're the same, or that they're redundant, or that you can safely remove any of them. DLLs all have versions, and applications sometimes depend on specific versions (it's rare for an application to actually check the minor version, but they do assume that dlls they find in their own search path are the ones they installed, ones they were tested to work okay with). You can see the version of a given DLL by bringing up its "properties" sheet in Windows Explorer and looking in the "details" tab. The second thing I described above is not a solution, it is a reason why no software can exist that accurately find genuinely unreferenced DLLs. -- Finlay McWalter • Talk 21:07, 23 June 2010 (UTC)

Task Scheduler

Resolved

I want to run a .bat script every 10 mins with Windows Task Scheduler. However, it doesn't work, the .bat script just flashes up for half a second and then ends. If I run the .bat script normally it works fine. What am I doing wrong? 82.43.90.93 (talk) 14:34, 23 June 2010 (UTC)

Try adding a "pause" statement to the end of your batch file so you can see the result and hit a key at your leisure, so Windows doesn't just immediately close the cmd.exe window once the .bat file is done. Comet Tuttle (talk) 14:43, 23 June 2010 (UTC)

And you'll probably see a "file not found" error of some nature. That's typically because you're expecting the batch file to run with its working directory set somewhere, and with the PATH set with something in it, and it's providing a different environment where that's not the case. It's generally best to specify the full paths for everything (for the batch file, for all the executables it calls, and for any data files it references). -- Finlay McWalter • Talk 14:55, 23 June 2010 (UTC)

Thanks everyone, that solved it! 82.43.90.93 (talk) 15:15, 23 June 2010 (UTC)

Language Display in Games

I've got a mod for one of my games (Men of War) which adds certain scenarios to it. Unfortunately, the English version of this mod is not available yet (i.e. the localization files for this mod have not been released), and I only have the Russian version. The problem is, the cyrillic is not showing up in the game (I get random symbols), meaning I can't read what the objectives are (the objectives are spoken, too, but it's way too fast for me!). Does anyone know what I should/could do to fix this so that cyrillic is displayed in-game? --KägeTorä - (影虎) (TALK) 17:12, 23 June 2010 (UTC)

Assuming a windowsish OS, try setting your system locale to Russian (Control Panel->Regional and Language Options->Administrative->Change system locale... in the Vista box I'm currently on). You may need the Russian version of Men of War, though. —Korath (Talk) 17:19, 23 June 2010 (UTC)

Did you use Windows Update and make sure the Cyrillic language pack is installed? Comet Tuttle (talk) 18:18, 23 June 2010 (UTC)

I believe the cyrillic language pack is installed, because in 'Regional settings' I can install Russian input (which I have done) and therefore write in Russian. I can view Russian in any other software (Office, browsers, etc.). If this is not what you mean, then how would I go about checking for this (and then installing it if it's not installed)? I don't want to change the system locale just for one single mod, though. --KägeTorä - (影虎) (TALK) 18:45, 23 June 2010 (UTC)

What is the operation done for data or file is created ,deletion ,selection ,cut,copy,past

What is the operation done for data or file is created ,deletion ,selection ,cut,copy,past,drag drop etc —Preceding unsigned comment added by 117.204.80.63 (talk) 17:29, 23 June 2010 (UTC)

Sorry, I don't understand the question, and I don't think other editors are likely to, either. Are you asking how to write a program to do these things, or are you perhaps asking what Windows or Linux do when the user does these things? Comet Tuttle (talk) 18:20, 23 June 2010 (UTC)

I think it's the second one, "what is the OS is doing to the data when those actions are taken". 82.43.90.93 (talk) 20:14, 23 June 2010 (UTC)

Making a page live.

I have edited pages that have already been live in the past, but this is the first time I have created my own page. The page name is "Newman Grace Inc.", and I have finished editing it, I have saved it, however, I cannot find it on wikipedia when searching for it. How can I make sure that my page is indeed live, or is there another step that I must take in order to make the wikipedia page live. Thank you for your help. —Preceding unsigned comment added by OBillyHill (talk • contribs) 19:22, 23 June 2010 (UTC)

You created the article on your user page, User:OBillyHill and it is still there. More common is to create a subpage by entering in the search box at the upper right of a page User:OBillyHill/Newman Grace Inc.. That way you can create and improve the article, adding references and categories, before moving it from a subpage of yours (a sandbox) to the article mainspace. A new article of a sentence or two created in main space is very likely to get speedily deleted, but an article with references and a reasonable claim of notability stands a much better chance. You rename the page by "moving" it as described at Help:Moving a page. In the most recent version of the Wikipedia "skin," you place the cursor on the downward pointing arrow at the top of the page, and you should see the option "Move." Then in "To new title" you enter the name the article should have, Newman Grace Inc. You cannot move a page if your account is too new and is not yet autoconfirmed. If so, you can request at Wikipedia:Requested moves that it be moved from your user page to mainspace. Please do not copy and paste to move a page, since that violates the Wikipedia copyright rules for contribution history. Edison (talk) 21:16, 23 June 2010 (UTC)

To Edison's last point, it's actually OK to copy and paste to move a page if you are the sole author of the page and you originated it on, say, a user sub-page of yours. Comet Tuttle (talk) 00:09, 24 June 2010 (UTC)

You will notice that Wikipedia does not include "Inc." in the title of an article about a company. So you should consider creating your page with the title Newman Grace. Sussexonian (talk) 21:49, 24 June 2010 (UTC)

Letter case in filenames when transferring photos from digital camera to Linux

I've been transferring photos from my Olympus E-520 DSLR to my Fedora 12 Linux computer by connecting the camera to a USB port, mounting it under the Linux filesystem tree (using automatic filesystem detection) and then just copying the files across, because I haven't been able to make any better or easier way work. Previously this has worked all OK, but now I've run into a small issue here.

Previously, Linux saw the files on the camera in all lowercase, such as dcim/100olymp/p1010001.jpg. Now, with no apparent reason, it has suddenly started seeing them in all uppercase, such as DCIM/100OLYMP/P1010001.JPG. This creates a problem, because like all Unix-compatible systems, Linux uses case-sensitive file names, and so the two are actually different files. If I just blindly copy the files across, mixing both conventions, I end up with duplicated photos. I have been able to change the filename case with a simple command: for i in *; do mv ${i} `echo ${i} | tr [:upper:] [:lower:]`; done, but it's awkward having to do it every time.

I don't know whether the fault is in the camera (thinking all the world is Windows, where letter case in filenames doesn't matter) or in Linux (thinking that since Windows is case-insensitive, the default rendition of the filenames doesn't matter and it can blindly force its own version instead), but I want to know how to fix this. I want to force Linux to use a consistent letter case, no matter what letter case the camera uses internally. How can I do this? Is there some option to mount or some filesystem driver parameter I can use for this? JIP | Talk 19:30, 23 June 2010 (UTC)

There are several File Allocation Table filesystem drivers for Linux, as described in FAT filesystem and Linux. It sounds like, for whatever reason, your camera is mounting as msdos rather than vfat. See what it's mounting as now, and try manually mounting it with vfat. -- Finlay McWalter • Talk 19:40, 23 June 2010 (UTC)

No, it's mounting as vfat. I don't have any idea what caused Linux to suddenly start seeing the filenames in a different case. Maybe there is some option to mount that tells it how to see the filenames? If all else fails I'll just have to rename the files afterwards. JIP | Talk 06:46, 24 June 2010 (UTC)

FAT originally didn't store letter case. This changed with extensions added in Win 95. The extensions are called "vfat" because Win95's FAT driver was called vfat (so was Win 3.11's) and people don't understand the difference between drivers and filesystems apparently. Digital cameras (at least the ones I've used) don't implement the extensions. This means there's no correct case for the filenames generated by digital cameras; it's up to the filesystem driver (or higher-level software) to decide. Linux's "vfat" driver supports the shortname=lower mount option to use lowercase, but that's the default, and I don't know why it would suddenly change on your system. -- BenRG (talk) 13:26, 24 June 2010 (UTC)

Thanks for your help. The mount option shortname=lower is exactly what I was wanting. Now I just have to automatise the process. Up until now I've been using the same mount point for every device connecting at /dev/sdc1 (/dev/sda and /dev/sdb are reserved for my internal hard drives), which include both my LaCie Neil Poulton drives and the digital camera, as Linux seems to access device files in /dev in sequential order as it finds the devices in the first place - not probably unlike how Windows assigns drive letters such as C:, D: and so on. I can't just write the option into /etc/fstab because then it would probably affect my LaCie Neil Poulton drives as well. Well, there's nothing forbidding multiple mount points for the same device file, each with different options. I will just have to remember which is which. I could automatise the entire process, but I'm not just storing all my photographs in a single big directory - it would contain tens of thousands of photographs. Instead they're divided into subdirectories based on their sequence number, and I will have to figure out a way to make Linux automatically find out which files it needs to copy (in other words, which it doesn't already have) and into which subdirectories. JIP | Talk 19:28, 24 June 2010 (UTC)

Reasonably new Linux systems use udev to manage devices, particularly dynamic devices like pluggable usb disks. If you want to do special stuff for a specific disk, you'd add a rule to a file in /etc/udev/rules.d (well, that's the directory on Ubuntu, I don't know about other dists) that detects that specific disk and gives it the specific mount options you want. General notes for writing udev rules (and how to figure out how to identify your specific device automatically) are here, and an example that manipulates the mount_options variable is here. Note that udev also allows you to run a given shellscript when a given device is plugged in (or out), so you could have a script that detects your camera, copies its files off, does whatever jiggery-pokery you want with them (e.g. prefix them with the date and time, or put them into a new subfolder named after the date and time, move that to your final picture archive, and delete the copies off the disk). -- Finlay McWalter • Talk 20:41, 24 June 2010 (UTC)

Logo of Diba inc.

Resolved

Can anyone find me a copy of the logo of Diba Inc., a Menlo Park, California internet-appliance startup that was bought by Sun Microsystems in 1997? I only have it on an ancient t-shirt, which I've scanned here. Tineye and my usually proficient Google-fu haven't come up with anything at all. Note that it's not the insurance company, the Iranian bootmaker, or a defunct East-European social-networking site. -- Finlay McWalter • Talk 19:58, 23 June 2010 (UTC)

http://web.archive.org/web/19970120192623/http://www.diba.com/ You will need to remove the drop shadow, and it is a low-res GIF. PleaseStand ^(talk) 22:47, 23 June 2010 (UTC)

More specifically, http://web.archive.org/web/19970505134343/www.diba.com/logo.gif PleaseStand ^(talk) 22:51, 23 June 2010 (UTC)

Gosh, silly me, I entirely forgot about webarchive. That's just the ticket - thanks! -- Finlay McWalter • Talk 23:14, 23 June 2010 (UTC)

How to watch the new Torrent TV Pilot "Pioneer One"

"Pioneer One" is said to be the first TV program pilot specifically for viewing by Torrent. I have never tried this video form before, and when I go via Firefox to the website [1] and click on "Download Torrent" nothing happens. I use Windows XP. I get a box which says "Opening Pioneer.One.S01E01.720p.x264-VODO.torrent Open or save? "Open" leads to "Browse" and a choice of Firefox and other unlikely programs for viewing. I assume "Save" is the right answer. This leads to a list of downloads, with a 90.1kb file, clearly not the 1119 MB file for the show. If I double click that 90.1 kb file, I get a Windows message that "Windows cannot open this file" without knowing what program created it. If I click "Use the web to find the program" I get a Windows page which offers some downloads from sources of unknown trustworthyness which might contain malware. Basically, what computer program should I install to be able to watch a Torrent program, and what are the basic steps to download and watch this particular program? It really should not be this roundabout. Will some widely known media viewer such as Real, Itunes, or Windows Media player play the thing? And how do I avoid cluttering my hard drive with giant files for programs looked at once? If they were in the playlist for a particular viewer, I might be able to download, watch, and delete. If I have to download a "BitTorrent client" then free, widely used, malware free, glitch free and legal would be good criteria. And why is a downloading and viewing program called a "client?" Thanks. Edison (talk) 20:50, 23 June 2010 (UTC)

You need a Torrent client. The torrent file you downloaded (the 90kb one) is just a 'pointer' of sorts to allow a client to then find the file through 'seeds' and then download it. I use Transmission which is on Mac but there's plenty out there. Note - you don't normally watch a torrent until it has completed the download as it doesn't necessarily download the file sequentially. ny156uk (talk) 21:01, 23 June 2010 (UTC)

Per Comparison of BitTorrent clients, Transmission does not run on Windows. What would be a good one for Windows? How about BitTorrent 6 ? Some concern: 1)Is this part of the peer-to peer sharing thing, where people have gotten sued for downloading copyrighted programs, or where material on my computer becomes accessible to others? 2)Are torrent files more likely than others to bring in malware? 30 How do I avoid clogging up the hard drive with downloaded files watched once and of no future interest? I have a fast broadband connection, so I usually just watch TV or movies streaming. Edison (talk) 21:27, 23 June 2010 (UTC)

1) Yes it is a type of peer-2-peer network. The torrent will only share files that you allow it to - which at the very least will always include the file(s) that you are in the process of downloading. As long as what you are downloading/sharing is legally allowed to be downloaded you are fine. 2) A lot of torrent hosting sites have a 'good' and 'bad' indicator (so bad files are voted down) but beyond that i'm not sure - it's certainly accepting files/packages of files from the internet so it's definitely not risk free. 3) Set the downloads to your desktop/a folder on your desktop and then watch the downloads from there - you can then just delete them as soon as you've watched them/whenever (basically keep all the downloaded files in one place then it's easy to go in and delete at your will). ny156uk (talk) 09:28, 24 June 2010 (UTC)

μTorrent is highly regarded. If you want something open source and not just cost-free, try Vuze. -- Finlay McWalter • Talk 21:32, 23 June 2010 (UTC)

Smaller alternatives to OpenOffice with British-English spell checking?

OpenOffice is nearly 400MB with a British English add-on; despite "Never look a gift horse in the mouth" dare I mention the word bloatware? And it has some irritating features I don't like. I've already looked at List_of_word_processors#Freeware and similar articles.

What smaller free alternatives with British English spell checking are there please (that people can recommend from personal experience or knowledge rather than just providing a link to Google)? All I really want is a word-processor, never use the other things in OpenOffice except sometime the spreadsheet which may be available independently. I have Windows XP.

I spend a lot of time writing letters so something that can help me do that quickly would be great. Thanks 92.15.3.0 (talk) 20:54, 23 June 2010 (UTC)

AbiWord does the job if all you want is a word processor. According to their website it has British English language support. --KägeTorä - (影虎) (TALK) 21:52, 23 June 2010 (UTC)

Of course, if staying online to use it is acceptable, there is Google Apps. You may need to download a British spell-check dictionary for your web browser though. PleaseStand ^(talk) 22:37, 23 June 2010 (UTC)

Can the Windows Vista Business be installed in more than one computer within a small business?

Can the Windows Vista Business be installed in more than one computer within a small business (like the Office Small Business)? —Preceding unsigned comment added by 148.137.86.18 (talk) 21:06, 23 June 2010 (UTC)

No, not according to Microsoft's licensing terms. Only through a volume licensing agreement such as Open License does Microsoft permit using the same Windows product key on multiple computers. By the way, the Office Small Business license agreement only allows the second installation to be on a "portable device" that is only used by the primary user of the other computer, so even for that, installing on two desktop computers (rather than on a laptop) would not be permitted. PleaseStand ^(talk) 22:31, 23 June 2010 (UTC)