Wikipedia:Reference desk/Archives/Computing/2006 July 26

Humanities Science Mathematics Computing/IT Language Miscellaneous 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 at one of the pages linked to above.

< July 25 Computing/IT desk archive July 27 >

Best way to study for Comp sci GRE

Can anybody offer any advice as to how to study for the computer science GRE? I have a few books with practice tests, but none of them offer any subject material to study. I've gone over most of the material in previous classes (besides algorithms, architecture, networking) but going over everything seems a bit overwhelming.

I remember two types of questions in the CSCI GRE:

One type was just math word problems. I didn't get the purpose. They were similar to: "If a harddrive has an average transfer rate of 120b/ms and an average seek time of 8ms, how long would it take to read a 12kb file assuming that the file is not fragmented?" They aren't complicating it. They are simplifying it - 8ms seek to find the start of the file and then 120b/ms after that until you read all 12kb.

The other type of problem was random definitions - again, I didn't get the purpose. They were similar to: "(insert some name you've never heard of)'s Theory of (insert some word you've never heard used in Computer Science) states that...(insert possible answers that are all very similar)." You have to take the practice tests and, if you get one wrong, Google (or Wikipedia) for it. --Kainaw ^(talk) 16:05, 26 July 2006 (UTC)

It was a while ago that I took it, but I recall there being some questions on databases and computer graphics which were never covered in any classes I took (luckily I'd learned a lot of that stuff on my own). There were definitely things like given two nested for loops, what would the output of a given block of code be, and some simple time complexity analysis problems. I also seem to recall a question about finite state machines. Basically they cover all sorts of ground, and more likely that not, there will be questions on there your classes haven't prepared you for. The best thing to do is just take practice tests and see what you don't know, as Kainaw suggested. Digfarenough 16:29, 26 July 2006 (UTC)

Running out of memory storage

hi all, someone has told me that when I delete data (file, folder etc) stored on my computer, it is erased from the hard disk( physical area in CPU) leaving that area(certain blocks) useless i.e connot be used anymore for the purpose of the data storage, so my question is how can I shift those blocks at the end (In other words, put the nonempty blocks together) in order to fill the gaps in between. -- FOZ

I think you might be a little confused here but I'll see what we can do:

• First of all, when you delete something off your computer, it doesn't make those sections of the hard disk (sectors) useless. Otherwise the hard drive manufacturers would be ten times as rich and we'd be ten times as poor. When disk errors occur, some file systems, like NTFS render the sector unuseable to prevent more data corruption. But normally, when you delete something, you can use the space again.
• Second, the CPU and hard disk are not joined together "erased from the hard disk( physical area in CPU)". The two communicate to each other, with each doing their separate tasks but working together along with other parts to form the computer.
• Third, ("how can I shift those blocks together in order to fill the gaps in between") when hard disks add data, they put in the first available spot. Even if it's to small to fit, they do. The remaining data gets put in the next available place. Over time, bits of files get scattered all over the disk making the hard disk work harder (no pun intended). This is called fragmentation. Defragmentation is where the hard disk rearranges the data back into order, like when a librarian puts all the scattered books into their correct place on the shelf. Windows comes with a free defragmentor under Start, Programs, Accessories, System Tools, Disk Defragmentor. There are commercially available defragmentors which do a much better job.

Hope this helps. Harryboyles 07:36, 26 July 2006 (UTC)

Thanx Harryboyles, it realy helped me alot.-- FOZ

There are some other commercial utilities that will defragment. --Proficient 12:00, 26 July 2006 (UTC)

With Microsoft Windows operating systems, the default behavior when deleting something using the GUI is to place it in the "Recycle Bin," so indeed, the disk space is not immediately freed up. This scheme allows one to recover files that have been accidentally deleted, but it's usually a good idea to empty the recycle bin regularly so that the space is, indeed, freed up. --LarryMac 13:47, 26 July 2006 (UTC)

Music software

Are there any softwares that allow you to edit music? For example i have a music file and I want to edit it so taht the vocals could be cut out and the piano part could be cut out? (more importantly are there any open source softwares like that?) Thanks!

Audacity is a great music editing open source program. If the piano part and vocal part are on the same track as the other parts, they cannot be edited out, but no program can do that once all of the parts are combined. —Mets501 (talk) 09:13, 26 July 2006 (UTC)

There's a list at Digital audio editor. But as Mets501 pointed out, what you want specifically isn't easy. Hoever, if the quality doesn't matter too much (eg you want to play and sing along for practise) and the voice is in a specific frequency range (it usually is) you might filter out that range, hoping not too much of the other instruments will be removed in the process. With a piano this would be a lot more difficult because a piano had a much wider range. DirkvdM 09:58, 26 July 2006 (UTC)

Often (but not always) the main vocals are mixed right in the middle of the stereo field, so if you invert one channel and add it to the other, the vocals will disappear (most programs have an option to do this, in Adobe Audition it's under Amplify, as I recall). However, if stereo reverb was used on them, you'll still get some leftover reverb of the vocals. For a given song, there's an easy way to check this: play the song on some system with headphones plugged in, then slowly pull the headphone plug out.. at some point the sound will get weird because the channels will be adding in the way described above and anything in the center of the stereo field will disappear (try it with Nine Inch Nails - Piggy for a very clear demonstration). DirkvdM is right about filtering though, especially with piano which has a lot of higher harmonics, it'll be hard to get rid of it all. One other possibility: if the producer of the album or the musicians were really lazy, they might have just recorded the piano part for the verse and chorus once and then repeated it in the song, so if you could find a section of the music with just the piano playing that part, there's a small chance you could invert both channels and add it on top of the same part elsewhere in the song to cancel some of it out. The chances of that working are very slim, though. Digfarenough 16:38, 26 July 2006 (UTC)

With regards to removing individual instrumental/vocal parts from a stereo mix - a good analogy would be a little like trying to remove individual ingredients from a baked cake, unless of course you do have stems (an industry term for tracks or groups of tracks, for example backing vocals of a song as one stem) available which are sample-accurate to the final mix so you can phase-cancel them like Digfarenough quite rightly said. I'll second Audacity as a nice audio editor and it's free too! Neilius 19:31, 31 July 2006 (UTC)

Circumventing Blocking

When I'm at school I sometimes want to use my free time on the school's computers to play online games or browse pages, however these are often blocked. Our school's blocking software (provided by the NSW education dept) is quite smart and won't allow me to circumvent it by looking at Google's Cache or translating the page from english to english. Are there any other simple ways that allow someone to circumvent system blocks? --AMorris (talk)●(contribs) 12:19, 26 July 2006 (UTC) (though it's probably not the wisest idea to put my name to this q.)

My friend searched the internet and looked for anonymous proxies and managed to get onto eBay, despite Websense blocking it as "internet auctions". Anonymous proxies could be one way to go, but these are normally also blocked as "proxy avoidance". Worth a try though. Iolakana|^T 12:59, 26 July 2006 (UTC)

Indeed. Proxies seem the easiest solution. --Proficient 15:54, 27 July 2006 (UTC)

I keep a list of links for a situation like this at my 204 page. Encrypted ones are prefered. And a lot of times all you have to do is stick the "s" on the end of "http" to use a site. If you need help with something like this again, just ask on my talk page. — The Mac Davis] ญƛ. 21:29, 27 July 2006 (UTC)

If you what want is Wikipedia, the TLS server is at https://secure.wikimedia.org/wikipedia/en/wiki/Main_Page. --cesarb 15:15, 28 July 2006 (UTC)

If your school doesn't block executable files, (unfortunately mine does) rather than use a public proxy (Which I might add, is largely insecure) you might want to try using tor. - Ridge Racer 18:54, 30 July 2006 (UTC)

WikipediA font?

What font and settings are used for the "WikipediA" in the logo??--Sonjaaa 15:12, 26 July 2006 (UTC)

Not sure. Looks like a custom font to me (I don't think I know any with a crossed W like is shown in the picture). It is similar to "CAC Camelot" in small caps, however.—Mets501 (talk) 17:51, 26 July 2006 (UTC)

I think it's Garamond. This question has been asked before. Neat; I've never noticed that the 'A' was capitalized.--epf 22:02, 26 July 2006 (UTC)

It's Hoefler Text. It's a beautiful font, incidentally, particularly for long documents. Print out a school paper in Hoefler Text and it looks wonderfully professional - like a novel, even. Incidentally, I believe Hoefler Text is available on Apples but nowhere else (unless you pay an exorbitant fee.) zafiroblue05 | Talk 23:01, 26 July 2006 (UTC)

I use the Colongne skin and its all capitals in Times New Roman, white. — The Mac Davis] ⌇☢ ญƛ. 20:21, 27 July 2006 (UTC)

Sunni vs Shia

I have somewhat of an idea about the difference between these two groups, but from what I know I cannot understand why the hate each other. Anyone have an answer?

You might want to ask at the Humanities or Miscellaneous Ref Desks, this is hardly a Computing question. --LarryMac 16:49, 26 July 2006 (UTC)

Dunno, I think I've heard the Sunni like KDE but the Shia prefer GNOME... but it's some ancient dispute that I don't know the details of. When will they learn you can run both at the same time? Digfarenough 16:58, 26 July 2006 (UTC)

No, the Sunni prefer Unix, the Shia MS-Windows. User:Zoe|^(talk) 17:48, 26 July 2006 (UTC)

Isn't there a radical wing of Sunnis that prefer Macs? --Kainaw ^(talk) 22:34, 26 July 2006 (UTC)

First of all who have told you that these two parties hate each other??? they realy dont...-- FOZ

And second of all .... ? DirkvdM 06:24, 27 July 2006 (UTC)

Sync sent items on PDA

I have Windows Mobile 5.0. Is there any way I can sync emails sent from my PDA over Wi-Fi with the 'Sent Items' folder on Outlook on my PC when I syuc the two devices. Many thanks, --

How might I create a Dimensional analysis program?

I'm very impressed how Dimensional analysis and Buckingham's pi theorem can be used to create a mathematical formula (or set of formulas) just from specifying the variables you think might be relevant. (The common example used is the formula for a pendulum).

I think it would be fun to have a software program that could take as input the user supplied list of variables, and output a formula (or more probably, a list of formulas) obtained by using Dimensional Analyis and the Pi theorem.

I am only an amateur programmer now and then, and I cannot think how I would begin to program such a thing.

Could anyone suggest how this might be done please? I suppose the problem can be broken down into how to represent the data and constraints, and how to transform them in accordance with Dimensional analysis and/or the pi theorem.

Further constraints that could be added are making the formulas consistent with levels of meaurment used (eg. ordinal etc), and with intensive and extensive measures (eg an extensive measure divided by an extensive measure gives an intensive measurement (I think)).

Thanks. --81.104.12.4 19:11, 26 July 2006 (UTC)

You need to specify the problem a bit more, I think, and then the algorithm should be more obvious. If what you're doing is taking a set of physical dimensions (such as "power" or "speed") and attempting to express one solely in terms of a product of powers of the others (which is the usual task in dimensional analysis: it is assumed that the one quantity is proportional to some power of each other one), you can express it mathematically by defining each dimension in terms of some basis of dimensions (e.g., the SI base units) and making each a vector of the powers of the basis dimensions. For example, using the basis of just {mass, length, time}, we can express speed as $\{0,1,-1\}$ and power as $\{1,2,-2\}$. Multiplying dimensions corresponds to adding vectors (since the vector components are exponents, and multiplying adds exponents), and so raising dimensions to powers corresponds to scalar multiplication of the vectors. Then we can state our problem as ${\vec {v}}_{0}=\sum _{i}p_{i}{\vec {v}}_{i}$, or $v_{0j}=v_{ij}p_{i}$ (with Einstein summation), where ${\vec {v}}_{0}$ is the desired result dimension, ${\vec {v}}_{i}\;(i>0)$ is one of the "source" dimensions, and $p_{i}$ is the power (possibly 0) to which the ith dimension should be raised. Since all the v-quantities are known, this is a standard linear system that can be solved with basic linear algebra.

For your pendulum example, we take the dependent dimension as time ($\{0,0,1\}$), and the candidate dimensions as length ($\{0,1,0\}$), mass ($\{1,0,0\}$) and the acceleration of gravity ($\{0,1,-2\}$). The system then looks like ${\begin{bmatrix}0&1&0\\1&0&1\\0&0&-2\end{bmatrix}}{\vec {p}}={\begin{bmatrix}0\\0\\1\end{bmatrix}}$, which has the solution ${\vec {p}}=\{{\frac {1}{2}},0,-{\frac {1}{2}}\}$. This corresponds to $T=L^{\frac {1}{2}}m^{0}g^{-{\frac {1}{2}}}={\sqrt {\frac {L}{g}}}$, which is of course the correct result to within a constant coefficient that we were not hoping to discover.

I hope I haven't ended up destroying your desire to write such a program by reducing the problem to well-understood mathematics! I'm afraid there just isn't much of an interesting algorithm here unless you consider more complicated analysis that may or may not be feasible to automate. Perhaps you could hew closer to the π theorem and look for dimensionless quantities in terms of all the variables (but then you're computing the null space of a slightly larger matrix, which is a somewhat harder problem). In any case, hope this helps. --Tardis 06:50, 27 July 2006 (UTC)

What pocket-sized thing best to read ebooks on?

There are lots of books listed in Project Gutenberg that I would like to read. However it is not very comfortable sitting up by a desktop computer reading them, nor would I want to try to use a laptop to try to read them in bed.

I would like something that could fit into a large pocket, that I could comfortably read in bed or on the train. Something that can cope with .txt files. Ideally, something not too expensive.

What specific brand of device would people recommend please?

(By the way, I did read one book by printing out the text in tiny type in two columns onto sheets of paper, but it was too tedious a method to do again.)

Thanks. --81.104.12.66 20:33, 26 July 2006 (UTC)

Some sort of PDA would be what you are looking for. Some sort of cheap Palm based device would probably be cheapest, but every model should cope with something as simple as a text file. --Robmods 21:16, 26 July 2006 (UTC)

I suggest a PDA with a black-and-white screen and no backlight. A backlight causes eyestrain after a while, and color displays require a backlight. As a side benefit, the lack of a color screen and backlight will triple the expected battery life. For reading in the dark, get a headlamp.

Personally, I use a Palm Zire 21, which cost me $100 new. The model's been discontinued, but you should still be able to find it in stores. Alternatively, you could look for one of the older Palm "m" models which uses standard AAA batteries rather than the integrated lithium battery of the Zire 21. --Serie 22:25, 26 July 2006 (UTC)

With PDAs, see if you can try them before you buy - I have a Palm Zire (very old), and its display resolution is a bit too low for comfortable reading, as I can't get many words on one line. Also, check if it supports longer text files - most PDAs ought to, but better to check (for instance, mine only supports a few kB of text in a single file, but, then, it is somewhat ancient). However, I echo the advice about black and white screen and no backlight - you will save a lot on battery power, and unlike fancier PDAs, my Zire will easily run for a month or more on one battery charge. — QuantumEleven 06:30, 27 July 2006 (UTC)

Perhaps you can copy it to your cell phone. --Proficient 15:55, 27 July 2006 (UTC)

I use a tablet PC. Its about the size of a etch and sketch. Many so called tablets are just laptops with a fancy hinge called "convertables" but some, like mine have no keyboard ( Mine does in its little notebook case) I carry a mouse and a rubber keyboard for when I can't reach my desktop. This brand was cheap; under a thousand bucks, but discontinued like the day after I bought mine so I'm not telling you its name. The pen is a bit dicey sometimes but thats what usb mice are for. The digital card reader is great cause I can clean out my cameras without going home. The sound with mine is tinny even with earphones. It understands my voice recognition software well. 31 July 2006 (UTC) ProtoGeek

swapping DIMMs

My Dell Inspiron 5100 with 384K memory crashed and then mostly it wouldn't start up again, or it would start and fail to completely boot, or it would boot and fail quickly in some application. I swapped the 128K and 256K DIMMs in the two sockets, and it has been working now for two days. What made the difference? Can I have any confidence that this "solution" will last? Thanks, Halcatalyst 22:36, 26 July 2006 (UTC)

You might have bad memory (I assume you mean 128MB and 256MB DIMMs). To test, I'd recommend using memtest86. To isolate which memory stick is bad: if you get errors, run memtest with only one stick in, and then only the other stick. This will let you isolate which one is bad. Don't assume that if the problem shows up in the first 128MB of memory that it's in the first bank; memory can have funny ways of being mapped from physical DIMMs into logical address space. Good luck! grendel|khan 03:42, 28 July 2006 (UTC)

question about network admins..

Do network admins of major US corporations have tools that will alert them of strange behavior on their intranets, such as a user slowly siphoning off data from the intranet onto a hard-drive and from there, uploading it onto a random ftp site?

I can say that those tools exist, but that doesn't mean that they are being used. I have a real-time monitor for the status of my servers (if they die, I get an email every 5 minutes until I get it fixed). As for usage statistics, I get a summary email at the end of every day. --Kainaw ^(talk) 01:00, 27 July 2006 (UTC)

hmm..interesting. Thanks1

Depends on the security of the network. Sites with extremely sensitive data will alert on users behaving in *any* unusual way, including connecting to ftp sites, or spending unusual amounts of time on internal sites, or even using an IM client. See: [1] for one example of compliance monitorng. Ronabop 21:16, 30 July 2006 (UTC)

not showing gmail

I run XP SP 2. When I visit gmail via either ie 6 or the most recent firefox, I can log in, but after page says "loading," all I can see is a blank page. I suspect the problem may be a setting in my firewall (Kerio personal ed.) Thank you. Amanaplanacanalpanama 23:51, 26 July 2006 (UTC)

for me it takes a bit of time, how long do you wait? Jon513 00:29, 27 July 2006 (UTC)

I would say longer than should be expected for the website to load. I've probably waited for 10 minutes. --amanaplanacanalpanama 02:33, 27 July 2006 (UTC)

I sometimes have that problem when I try to access gmail from work (all kinds of firewalls and proxies in the way). I'll see "loading" in the upper right (in red I think). But if I just do a refresh, then I see Loading in the upper left, in black, and it works. Not very scientific, and definitely original research, but I thought I'd mention it. --LarryMac 14:46, 27 July 2006 (UTC)

Perhaps you are using Firefox with the NoScript extension. Allow Google. Or perhaps a firewall is disallowing access. Make sure you have the correct plugins. --Proficient 15:57, 27 July 2006 (UTC)

Thank you very much for your input, LarryMac and Proficient. I have tried refreshing specifically and it doesn't help. I don't think it is a firefox problem because ie does the same thing. I'll have to try some of Proficient's other suggestions.--amanaplanacanalpanama 17:41, 27 July 2006 (UTC)

Thank you also Jon513.--Amanaplanacanalpanama 17:43, 27 July 2006 (UTC)

Ubervisor

I'm trying to find any source for article Ubervisor, but failed to find any. Article in IBM Systems Journal (linked in article) do not contain such term. Is this original research? --89.104.121.156 23:55, 26 July 2006 (UTC)

Seems like a protologism coined by some blogger. -- Koffieyahoo 01:33, 27 July 2006 (UTC)