Wikipedia:Reference desk/Archives/Computing/2008 May 6

Computing desk
< May 5 << Apr | May | Jun >> May 7 >
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.


May 6

edit

The Sound of Thinking

edit

What is the cause of the clicking sound that is made when you ask a computer to do something laborious? I'm not referring to the increase in fan speed but rather the apparently non-mechanical tck. Tks —Preceding unsigned comment added by Stanstaple (talkcontribs) 00:19, 6 May 2008 (UTC)[reply]

It appears that you are referring to the hard drive clicking sound. It makes that when it moves the magnetic heads around to read data from the platters. If you get a lot of it, it is very likely that you don't have enough memory in your computer for what you are trying to do and your computer has opted to use hard drive space as temporary (and extremely slow) memory. -- kainaw 00:48, 6 May 2008 (UTC)[reply]
Also note that this is hard drive specific. My main hard drive barely makes a sound, even under a heavy load. My secondary, on the other hand, makes the clicking noise you mentioned. Leeboyge (talk) 00:56, 6 May 2008 (UTC)[reply]
And if it's a louder than usual click or a very high pitched squeak you should back up your data. See click of death and head crash. --antilivedT | C | G 05:24, 6 May 2008 (UTC)[reply]
Am I the only one here whose skin crawls when they see the photo in the head crash article? Horrifying stuff right there. 206.252.74.48 (talk) 18:58, 6 May 2008 (UTC)[reply]
Once it is seen, it cannot be unseen. Paragon12321 (talk) 20:34, 6 May 2008 (UTC)[reply]

C File writing

edit

I am starting to use C programming language to write into files but I'm having a little trouble finding out what formats do different files take. Is there a good resource for this type of thing?Bastard Soap (talk) 08:02, 6 May 2008 (UTC)[reply]

I mostly use Google. The first hit I get for file formats is www.wotsit.org, followed by Wikipedia's very own file format article (has many links to additional info). If I wanted a specific file format I'd search the web to find the official spec, tutorials, hints, sample implementations, etc. Weregerbil (talk) 08:54, 6 May 2008 (UTC)[reply]
http://filext.com/ is the one I use the most. Or use Wikipedia - everything from .ace to .zip. --h2g2bob (talk) 21:18, 7 May 2008 (UTC)[reply]

Why not have Enter, Backspace and Delete options on mouse or on Toolbar

edit

Why must I have to get my right hand off my keyboard onto the mouse to position the cursor in order to highlight text for formatting (such as copy / paste or deletion) and then have to GO BACK to the board to use the Enter and Delete keys to tidy up the text. If these basic options were available on the mouse, then it would mean I could minimise hand movement between mouse and keyboard. If the hardware is too costly, it should be possible to access these through icons on the toolbar. And make them BIG icons. What say you, techies? And I would like a credit for this if my idea is taken up, plus about a million dollars. American. Myles325a (talk) 08:18, 6 May 2008 (UTC)[reply]

Shift key is your friend, shift+end then delete will remove the whole line from the cursor's position. Or otherwise use your left hand to press backspace/delete. Or you could use cut which you can either use the icon or ctrl+x and not having to move your left hand at all, just don't paste the content. --antilivedT | C | G 10:24, 6 May 2008 (UTC)[reply]
And, higher end mice have additional buttons which I believe you can program to emulate any keyboard button you want. -- Meni Rosenfeld (talk) 12:20, 6 May 2008 (UTC)[reply]
While I don't really thing having all those things on the mouse is a good idea (imagine how easy it would be to accidentally delete files; weigh that possibility against the possible advantages), I do agree that it would be a lot more useful if certain buttons were on the left side of the keyboard in order to facilitate having dedicated hands (left is for typing, right is for mouse things). When I do work on my own application GUIs I have experimented with this a bit, so that for anything other than real typing you can use dedicated hands. It speeds things up tremendously, in my experience. --98.217.8.46 (talk) 12:33, 6 May 2008 (UTC)[reply]
You mean like the keyboards used to be on Sun Microsystems computers? The extra ten keys on the left of the main keyboard were: Stop, Props, Front, Open, Find, Again, Undo, Copy, Paste, and Cut. And Macintosh keyboards used to use F1 through F4 as Undo, Cut, Copy, and Paste (and may still do so). The hegemony of PCs has meant a lot of good ideas have been discarded.
Atlant (talk) 13:09, 6 May 2008 (UTC)[reply]
KDE already had "Clear" and "Go" buttons on toolbars for many years. See the two buttons around the location bar in Image:Konqueror_on_Knoppix_5.11.png. --Juliano (T) 21:34, 6 May 2008 (UTC)[reply]
As a side note, some programs already have icons for cut, copy, and paste. Microsoft Word has had them on its toolbar for quite some time now. Leeboyge (talk) 06:27, 7 May 2008 (UTC)[reply]

OP myles325 responds. Thanks guys, some good points there. I am touch typist (type with all 10 fingers and without having to look at keyboard – cool huh?). All touch typists know that keyboards are designed by peck and stabbers who have no idea of what REAL typists need. The Shift key is duplicated on the left and right sides of the keyboard as a hangover from manual keyboards when it was made the task of capitalising letters on the left and right sides of the keyboard easier. I cannot see how duplicating the Ctrl and Alt keys on both the right and left of the space bar is anything but a waste of valuable keyboard real estate. And the return key should be bigger. And why is Backspace buried on the extreme left of the numbers line – this is one of the most used keys on the board. And why is its twin, the Delete key not even found on the main keyboard space? It would be used 10 000 times for every time a ^ { } ~ are used. Logically, the Delete and Backspace keys, crucial for navigation, should be together. I just used Backspace about 300 times typing this. It should alongside the Spacebar and triggered by the thumb. So you use the Spacebar to move the cursor to the right and then, seeing that you have misspelled the word you just typed, you use Backspace to delete that word. Does that make sense? Myles325a (talk) 08:00, 8 May 2008 (UTC)[reply]

You could redefine the keys with AutoHotkey or something similar - so you could make AltGr send a Backspace, and perhaps change the right-hand Windows/Start key to Delete. For making the keys bigger or smaller, you might want to have a look at the IntelliKeys, for which you can define your own layout. The only problem with that is that it doesn't have keys as such, only a flat sheet, so you may miss the feel of the keys. AJHW (talk) 15:00, 8 May 2008 (UTC)[reply]

Soft spam: a golden business idea?

edit

Why are so many people posting comments for driving a little bit traffic to their websites? Is it so a good business? In almost every forum I see comments like: "Hey, nice idea, but site xyz.com has a better concepts". Normally not following the flow of discussion. 217.168.0.115 (talk) 10:30, 6 May 2008 (UTC)[reply]

It's called comment spam, and is usually done by robots, which is why many forums require you to solve a CAPTCHA before posting. --Sean 12:33, 6 May 2008 (UTC)[reply]
I think Ars reported a few weeks back that Hotmail CAPTCHA was broken into. Kushal (talk) 02:55, 8 May 2008 (UTC)[reply]
While getting people to click on the link is a small part of the plan, the overall plan is to get search engines to rank their pages higher because they show up in so many places. Of course, the big search engines rarely index message boards (for this reason), so it doesn't work. -- kainaw 13:49, 6 May 2008 (UTC)[reply]
There's also the "nofollow" property for links, which is also meant to make this kind of spam unproductive for the spammer. --Sean 19:59, 6 May 2008 (UTC)[reply]
It is probably aimed at a different audience altogether ... maybe they are actually targetting individual visitors to the page rather than the search engines. Kushal (talk) 02:51, 8 May 2008 (UTC)[reply]

wireless passwords

edit

What method would be best for finding out the password to an encrypted wireless internet connection? My computer is running windows and I am a linux idiot.Makey melly (talk) 12:22, 6 May 2008 (UTC)[reply]

Do you mean hacking into someone else's wireless connection? My guess would be to just use a dictionary attack—most people choose really simple, easy passwords for that sort of thing because they are afraid of forgetting them, they plan on sharing the password anyway (with other users in their household, for example), and they are unclear on whether the stakes are high or not. If you mean find out the password on a router that you have set up and forgotten it, it depends on the router, but the easiest thing to do is to just do a factory reset. --98.217.8.46 (talk) 12:30, 6 May 2008 (UTC)[reply]
This might be useful. Algebraist 14:07, 6 May 2008 (UTC)[reply]
Try this never used myself, but hey, its what your looking for --Nick910 (talk) 14:39, 6 May 2008 (UTC)[reply]
I used the aircrack suite to test the security of my own networks. The conclusion I came to was that with really secure passwords I could crack WEP in about 70 hours on a pretty modest laptop, but that WPA is uncrackable assuming you pick a good password (i.e. not one that can be found by dictionary attacks). -- Q Chris (talk) 14:47, 6 May 2008 (UTC)[reply]
As far as I know, WEP should be crackable in far less time. I might not be recalling this exactly, but using a tool that exploits WEP efficiently can result in it being cracked in under 1 minute. WPA, however, requires a plain dictionary/brute-force attack, and with a sufficiently powerful key it is impractical to crack. 206.126.163.20 (talk) 00:16, 7 May 2008 (UTC)[reply]
Aircrack claims to be able to do this using an injection attack (i.e. it sends various packets to the router, which disrupt the handshake, leading to repeated exchanges). My wlan card does not support injection mode, so I had to use passive mode (i.e. just listening to normal conversations). The wlan also had a very low usage (wife and daughter, daughter for a few hours only, wife checking email). Most of the 72 hours was not useful capture but just waiting until sufficient handshake packets were obtained. -- Q Chris (talk) 08:19, 7 May 2008 (UTC)[reply]
IANAL but it might be illegal in some states in the US, including Texas. Please stick to open wireless connections (mine is open too, except for the router config itself, which is password protected). --Kushal (talk) 02:53, 8 May 2008 (UTC)[reply]

Help in checking website fine print updates

edit

Hi, I've signed up for a couple of sites and nearly all of them that happen to have lengthy fine print say that they can change it however they want, anytime. Some provide update dates, but none point out where the changes are. Kind of tired of pasting sections at a time to MS Word's "Find" function, does anyone know of a site where I can copy-paste two bodies of text and will show me the difference? Thanks. --24.76.248.193 (talk) 12:40, 6 May 2008 (UTC)[reply]

This kind of utility is usually called a diff. Here's an online version. — Matt Eason (Talk &#149; Contribs) 12:47, 6 May 2008 (UTC)[reply]

removing my email address from a webpage?

edit

an old newsgroup post i made has somehow been archived on the http://www.lliure.info/ website. the post contains my email address in open text. this website lists no webmaster name or details that i can contact to have my email address removed (i'm not even sure what type of site it is). how can i go about getting my address removed from the site? thanks. 86.31.34.35 (talk) 14:40, 6 May 2008 (UTC)[reply]

Information.. you can't redact from the internet .froth. (talk) 15:58, 6 May 2008 (UTC)[reply]
The WHOIS information for lliure.info does provide some contact information for the guy who owns the domain, which may or may not be correct. But if this is a case of someone archiving something that you or someone else put on the internet, .froth. is correct: once it's out there, it's out there. It ain't ever coming back. Even if you could talk this guy into removing it from his site (and chances are that he'll just ignore you), it's going to be on Google Groups and numerous other archiving services. That's just how it is. -- Captain Disdain (talk) 17:33, 6 May 2008 (UTC)[reply]
I am sorry. I wish I had better news but services like way back machine, which are very useful at other times, are in the way too. However, not all is lost. If you are determined enough, you might be able to get every major involved party into deleting your information. Please try to be polite and mind not to threaten them with legal action or anything silly. Good luck! Kushal (talk) 23:28, 6 May 2008 (UTC)[reply]
I am extremely skeptical that anyone could get the mere mention of an e-mail address removed from all of the various online archives out there, especially if a case couldn't be made for this being exceptionally important. (That is to say, "I'd prefer to not have my e-mail address known to the public" wouldn't qualify, whereas "this leaves my company vulnerable to criminal activities" might -- not that I can think of an instance where a single e-mail address could realistically cause such a situation.) And even if you could pull that off, that doesn't mean that the same information isn't also archived somewhere else that isn't publicly known at this time, but which ends up online later on, so you might very well have the same problem all over again next week, next month, or five years from now. That's just the nature of the internet; once you put something out there, chances are that it's going to stay there for as long as the net exists. -- Captain Disdain (talk) 01:12, 7 May 2008 (UTC)[reply]
Streisand effect - 206.126.163.20 (talk) 02:28, 7 May 2008 (UTC)[reply]

Windows command line question

edit

Today at work, I tried something as simple as deleting every .txt file in every subdirectory, in the Microsoft Windows Vista Business Edition they make me use. I thought a simple del *\*.txt in the Command Prompt would work. But no - Windows said it could not parse the arguments. On Unix, rm */*.txt works perfectly, and I thought it would be simple enough for Windows also to understand. Did I do something wrong, or is it just another sign of the inadequacy of Windows's command line? Is there another way to do it in Windows, short of handling each subdirectory by hand? JIP | Talk 16:59, 6 May 2008 (UTC)[reply]

Cygwin ought to handle it. --Prestidigitator (talk) 17:22, 6 May 2008 (UTC)[reply]
I thought so too, but then thought installing Cygwin would be overkill for such a simple task. But if there's no other way to do it in Windows, then it will have to do. JIP | Talk 17:25, 6 May 2008 (UTC)[reply]
It's got nothing to do with inadequacy (perceived or real), it just isn't how the syntax for the del command works. I'm not a Vista user myself, but the following has always been true for all previous versions of Command Prompt, so I'm going to boldly assume it hasn't changed. What you want is the /s switch, which has the del command also include subdirectories -- so you would type del /s *.txt from the root directory of the hard drive in question, and it should delete every .txt file from that hard drive, provided that they aren't read-only. For more information, del /? should provide you with help. In general, it's probably a good idea to glance through the help file or manual before you do anything involving the delete command. -- Captain Disdain (talk) 17:27, 6 May 2008 (UTC)[reply]


Thanks for the help. This seems like a fundamental difference between the two operating systems - in Unix, wildcards are handled by the shell itself, so what works in one command works in another. But in Windows, the command line passes the wildcards over to the individual commands, which then have the responsibility of handling them. This is also why rm * does not cause an "Are you sure?" question under Unix - but might result in a "Command line too long" error message. JIP | Talk 17:35, 6 May 2008 (UTC)[reply]
There are shells for Unix and Linux, such as zsh, that asks you confirmation for commands such as rm *, among other things. --Juliano (T) 22:32, 6 May 2008 (UTC)[reply]
Oh. I was answering generally to the request of making it more *NIX like and consistent so you can actually get things like this done efficiently on the command line. The exact command you entered wouldn't work in *NIX either for what you want to do (remove all files ending in '.txt' somewhere under the current directory). rm */*.txt would remove all files ending in '.txt' from the current directory's subdirectories. It would even miss files in the current directory (the directory name '.' is not included in expansion of '*', though it would be when expanding '.*'). You'd actually want something like find . -name '*.txt' -exec rm -f {} \;. Note the single-quotes on the name option value; that is because, as you say, the '*' wildcard is interpreted immediately by the shell in the current context. We instead want the find command to be able to expand it for each path it evaluates. The '-f' flag is just in case you are in a shell that will by default ask for a confirmation of each file, or in case you have an alias such as alias rm='rm -i ' defined like I do. --Prestidigitator (talk) 23:29, 6 May 2008 (UTC)[reply]
I don't know what bearing this has on your question, but Windows Powershell seems to be the way ahead for Windows command line stuff, not that it will rid you of the "wish this were Unix" feeling. It's still rubbish, just not so rubbish as cmd.exe. --90.198.200.119 (talk) 18:04, 6 May 2008 (UTC)[reply]
Notwithstanding command line, you could have done a search (right click) of the folder you wanted to find all txt files in. You can then select all and delete from the search result window. Sandman30s (talk) 21:24, 6 May 2008 (UTC)[reply]

GIMP question

edit

I am using the GIMP, and I have a .png file, consisting pretty much entirely of white, black, red, green, and blue. It is not exactly a clean cut between only five individual RGB colours, but pretty near to it. In other words, other than the white, the black, red, green, and blue also have a few "lighter" tones for anti-alias reasons. How can I remove the red, green, and blue from the pictures while keeping the black in? I have tried various methods of colour selections, but they always either fail to remove all of the red, green, and blue tones, or remove some of the black tones too. JIP | Talk 17:40, 6 May 2008 (UTC)[reply]

I would probably start with the Threshold, Levels, and Curves tools (Tools -> Color Tools or Layer -> Colors). Transform anything with a Value below a certain threshold (or do color channels independently) to full value (white). There might be some kind of filter or script that will do it by saturation instead, which I think would be more ideal, but I think this is a good place to try first. Good luck! --Prestidigitator (talk) 21:21, 6 May 2008 (UTC)[reply]
Oh. Or if you don't care WHAT grayscale value each pixel winds up as as long as it is greyscale, you can convert to grayscale (Image -> Mode -> Grayscale) and then possibly back again (Image -> Mode -> RGB/Indexed...). You could then (while the image is grayscale) convert to pure black and white with no gray pixel values in between using the Threshold tool. When you are done, try the Histogram dialog to see if you have what you wanted. --Prestidigitator (talk) 23:41, 6 May 2008 (UTC)[reply]
Another idea to try: Decompose the R/G/B channels into layers (ColorsComponentsDecompose..., select "RGB" and check "Decompose to layers"), then set the blending mode of the two upper layers to "Screen" (or maybe "Lighten only"). Or do the same with "CMY" instead of "RGB" and use the "Multiply" blend mode. (This will give you an inverted image.) The latter may work better if the colors you want to remove are indeed close to the additive primaries. You can play with the channel mixer if you want more control: the basic idea is to eliminate the colors one at a time, by choosing a channel mix that maps the unwanted color to white while preserving the black-white axis unchanged, and then blending the results together. —Ilmari Karonen (talk) 00:42, 11 May 2008 (UTC)[reply]

Taking an image grab of a movie file

edit

Hi all

What's the easiest way of taking a 'movie still' or image grab from a compressed video file?

Simple screen grabs on Windows and Mac come up black.

What d'you think? Joshua.c.j (talk) 18:37, 6 May 2008 (UTC)[reply]

From our Screenshot article, 'One way these images can be captured is to turn off the hardware overlay. Because many computers have no hardware overlay, most programs are built to work without it, just a little slower. In Windows XP, this is disabled by opening the Display Properties menu, then clicking, "Advanced", "Troubleshoot", and moving the Hardware Acceleration Slider to "None."' --LarryMac | Talk 18:42, 6 May 2008 (UTC)[reply]
VLC also has a nifty screenshot feature. 206.252.74.48 (talk) 18:52, 6 May 2008 (UTC)[reply]
I can do them just fine with VLC, just to confirm. asenine say what? 00:21, 7 May 2008 (UTC)[reply]

Thanks everyone! Joshua.c.j (talk) 11:11, 7 May 2008 (UTC)[reply]

Disk Partition

edit

Hey Guys, I am trying to format my hard disk and make a clean reinstallation of Windows XP SP2. However, I have already partitioned my hard disk in 2 halves. I tried to delete one of them using the Computer Management, however, instead of merging it with the system partition, it remains unallocated. Does someone knows how to merge these two partitions? Do I need to use a special software (a free one if possible)? Or will they merge when I reinstall XP? Thanks again Eklipse (talk) 19:42, 6 May 2008 (UTC)[reply]

The Windows XP installer has a built-in partition manager and formatter. Since you intend to completely erase and format your drive, you will have no problem deleting all current partitions, creating a new partition spanning the entire disk, and installing Windows on it. -- Meni Rosenfeld (talk) 20:59, 6 May 2008 (UTC)[reply]
Yes this is a good idea. However, DO partition again when installing XP. This time, choose a smaller partition for XP (for example 20 gig) as Windows folders tend to get fragmented very fast. Install your smaller or integrated programs into C and your larger, data-intensive programs into D. Sandman30s (talk) 21:22, 6 May 2008 (UTC)[reply]
I tried to install XP, but when I got to this step and tried to delete Partition1, I got this message:

Setup is unable to perform the requested operation on the selected partition. This partition contains temporary Setup files that are required to complete the installation.

and when I choose to set up Windows XP on this partition (without deleting it), it warns me that I'm installing multiple operating systems. Strange... Eklipse (talk) 21:51, 6 May 2008 (UTC)[reply]
I think I solved it. I should've changed the BIOS settings to boot from the CD. It is now in the process of formatting and installing. I hope it will go with no problems. Thank you guys. Eklipse (talk) 22:11, 6 May 2008 (UTC)[reply]

WMV to 3g2 converter

edit

I am looking for a wmv to 3gp converter, but I cant seem to find a free one that works. Does anyone have any suggestions? --Omnipotence407 (talk) 21:56, 6 May 2008 (UTC)[reply]

Try mediacoder it is available here[1]. It has got a very wide range of video/audio conversion capabilities and is under active development. Bobatnet (talk) 21:46, 7 May 2008 (UTC)[reply]

One the Mac, Quicktime Player Pro (with WMV Quicktime Components) can convert to 3gp. There's even a preset for it. Two clicks and you are done! Not sure how QT Player handles WMV on the PC side, though. --70.167.58.6 (talk) 17:13, 8 May 2008 (UTC)[reply]