Wikipedia:Reference desk/Archives/Computing/2010 March 14

= March 14 =

latex multicols
I'm using the multicols package in latex but am having some difficulties with certain things. Firstly trying to get floats to sit inside the column, currently i have this:

In my preamble then i simply declare floats as \begin{figurehere}...\end{figurehere}, However this seems to make latex not consider the image and the caption as one entity, so at times i'll have the image on a separate page to the caption. Is there a better way?

Secondly, i have some wide tables that i want to put on a separate page, thats easy just do \begin{table}[p] but this makes the tables appear on a page at the back of the document, is there a way to put the page in the middle of the document, closer to where its being called?Thanks--82.26.227.101 (talk) 11:52, 14 March 2010 (UTC)

My monitor won’t work with a splitter cable
I have recently changed my monitor which had a VGA connector. My old monitor was a Viewsonic TFT. This was linked to my LCD television using a VGA splitter cable and worked perfectly, giving a display on both the monitor and the TV at the same time. My new monitor has a D.Sub connector. It is connected to my LCD TV exactly the same way and with the same splitter cable as my old monitor. Now though, when I connect the splitter to the TV my monitor’s display doesn't come on, the TV does. Unplug the TV and the monitor comes back on. Has anyone got a clue as to what is causing this. —Preceding unsigned comment added by Shsharri (talk • contribs) 12:35, 14 March 2010 (UTC)


 * There is no need to SHOUT. I have taken the liberty of converting your message to use lowercase where appropriate - it is now so much easier to read.
 * I think more information is needed. What kind of D.Sub connector does it have? - a white DVI connector perhaps?  Take a look at List of video connectors and see if you can tell which it is.  I'm also a little puzzled - you say the monitor is connected to your TV.  Is that to view your PC on both TV and monitor screens, or to view TV pictures on both TV and monitor screens? Astronaut (talk) 13:07, 14 March 2010 (UTC)


 * Yea, we need you to fill in a chart like this:

OLD SETUP NEW SETUP ========= ========= Output from computer:   VGA        D-SUB Input to TV:           VGA        VGA Output from TV:        None       None Input to monitor:      VGA        D-SUB Output from monitor:   None       None


 * If this is correct, the confusing part is how a D-SUB signal is split into a D-SUB and VGA signal, under the new setup. StuRat (talk) 14:20, 14 March 2010 (UTC)


 * What brand and model is the new monitor? ---— Gadget850 (Ed)  talk 14:43, 14 March 2010 (UTC)


 * I'm rather confused by the responses above and the question in general. The only D-subminiature connector commonly used in modern computer monitors AFAIK is the DE-15. This is also what people usually mean when they refer to a VGA connector. The DVI connector is not a D-sub connector, and I've never even heard it referred to as one. While I guess you could technically call it a VGA connector since you can send a VGA signal thorough it, again I've never heard it referred to as one. Ditto on both counts for the Displayport or HDMI. Nil Einne (talk) 14:18, 15 March 2010 (UTC)
 * Ditto— that is why I asked for the model so we could could find the documentation. I have given up on explaining connector nomenclature and just use pictures. ---— Gadget850 (Ed)  talk 14:33, 15 March 2010 (UTC)


 * Many people use D-Sub to refer to all kinds of connector with a roughly D-shaped outer shield/ground (see File:DSubminiatures.svg for some examples), but I also considered it quite unlikely that they would have got a new, modern monitor with a really unusual connector. That is why I suggested perhaps a DVI and also suggested the OP take a look at our list.   Astronaut (talk) 01:14, 16 March 2010 (UTC)

Handwritten Character database
I'm doing a software project (for college) that tries to recognize handwritten characters(English alphabet). So far I've been working with scanned handwriting samples provided by some friends. But for good training and testing 10,000's of characters might be necessary. I tried googling for such ready-made labelled databases. The only useful one I found was that of NIST and I can't spend (any) money to purchase it.
 * 1) Is there a free handwritten character database out there?
 * 2) Where can I find a list of good features that can help identify a character. Currently my features are not good enough (for characters like D and O etc.)  —Preceding unsigned comment added by 59.93.2.126 (talk) 17:05, 14 March 2010 (UTC)


 * I'd distinguish between two types of writing:
 * Handwritten with ballpoint pen, as many people do now. That's probably the closest to your project, but I'm not sure if there is such a database available.
 * Calligraphy - written with brush. I would start with kaishu, and maybe progress to caoshu later. There should be many pages available online, but there's no mechanical way to assign the encoded character to the picture. &mdash; Sebastian 19:32, 14 March 2010 (UTC)
 * Please disregard my above post. I just realized that by "character", you did not mean Chinese characters. &mdash; Sebastian 02:36, 15 March 2010 (UTC)
 * This has some links (but not all are freely available), including this one, which looks promising. And see "zip code", the last entry on this page, which is apparently a classic database of digits from the USPS.  You'll find digits and capital letters in the third dataset on this page, though there aren't too many examples. and it seems to have dirty data.  Paul Stansifer 21:25, 14 March 2010 (UTC)
 * Thank you the links were helpful. 59.93.34.1 (talk) 01:08, 16 March 2010 (UTC)

"Gothic" fonts
Why are fonts such as Bank Gothic called "Gothic"? ╟─ Treasury Tag ►  Chancellor of the Duchy of Lancaster  ─╢ 17:09, 14 March 2010 (UTC)


 * From our Sans-serif article: "Before the term “sans-serif” became standard in English typography, a number of other terms had been used.  One of these outmoded terms for sans serif was gothic, which is still used in East Asian typography and sometimes seen in font names like Century Gothic."  Comet Tuttle (talk) 17:49, 14 March 2010 (UTC)


 * East Asian sans-serif typeface says "Gothic is an outmoded typographic term for sans-serif. It was so named because the type color of early sans serif typefaces were thought to be similar to that of the blackletter or 'gothic' script."; however there is no reference to support that claim. --Normansmithy (talk) 11:30, 15 March 2010 (UTC)

Rotten Tomatoes webpage problem
When I click on the appropriate buttons here http://www.rottentomatoes.com/features/special/2007/botbp/ I just get taken to the same webpage again. What is wrong please? I am using Firefox, all "cookies" enabled. The same thing happens when I use Internet Explorer 8. Is there anyone for whom the webpage does not do this? Is it some sort of pop-up which my browsers are banning? Perhaps it is being banned for malware reasons. Thanks 89.242.120.116 (talk) 18:52, 14 March 2010 (UTC)


 * That thing is from 2007, and has a stated cutoff date. ¦ Reisio (talk) 05:59, 15 March 2010 (UTC)
 * The smallprint says the cutoff data for gathering the statistics was in February 2008, it does not say anything about the webpage being unusable after that or any other date. So it still a mystery. Very bad web design by RT to have glossy duff webpages that do not do what they say they will. I have found the current URL and will update the RT wikipedia article link, which is where I got it from. 89.242.243.82 (talk) 11:47, 15 March 2010 (UTC)

Banned torrent
I was downloading a torrent from a private torrent tracker and it got banned by the tracker. Apparently the uploader forgot to make the torrent private so it's still being seeded over DHT and Peer exchange. Will they stop working when I have to update my tracker or will they continue to work forever? Thank you 93.136.219.166 (talk) 20:16, 14 March 2010 (UTC)
 * I am not quite sure but, They will peter out, but updating the tracker will not affect things. The tracker is a way for different peers to find each other, so without the tracker it is difficult for new peers to join the swarm Taemyr (talk) 11:38, 15 March 2010 (UTC)

Photoshop problem
Why is it that sometimes, in Photoshop, no matter what I do or if I restart the program, or create a new layer, or switch layers, or whatever, it only lets me draw a straight line? How do I fix it? Thanks,  Shannon   talk   contribs  22:19, 14 March 2010 (UTC)


 * With what tool(s)? Some really basic things: Try tapping the "Shift" and "Control" keys a few times—it might just think they are being pressed (and that can restrict its line-drawing functions). Last if not least, try rebooting the machine, see if it persists. --Mr.98 (talk) 00:16, 15 March 2010 (UTC)

Which files are the most difficult to compress?
This question has frequently arisen in my mind. In file compression, it is obvious that the files that are the easiest to compress (i.e. shrink the most when compressed) are those which consist entirely of the same byte value, perhaps even entirely of 0-bits or entirely of 1-bits. But which files are the most difficult to compress (i.e. shrink the least when compressed)? This obviously depends on the compression algorithm, but has there been any definite research results about this on any popular compression algorithm? This question might sound like the answer cannot be known, but upon further thought, the problem is often indeed solvable. If we assume a given file size (such as an n-pixel digital photograph), there are a finite number of possible files, so there must exist a specific solution (if not necessarily a unique one). It's just a matter of finding out what that is. Unfortunately, for pretty much all non-trivial cases, the number of possible files is so huge that a brute-force approach to solving the problem would not complete until the Sun goes supernova. So finding the solution requires a more clever approach. Has any been tried? J I P | Talk 22:35, 14 March 2010 (UTC)


 * Generally, files that compress have predictable patterns of bytes (not just all the same). Files that compress poorly have a random byte structure.  --Phil Holmes (talk) 22:46, 14 March 2010 (UTC)


 * That's right &mdash; a file with randomly generated content is not going to be compressible. (I mean, there might be a little, like if your file randomly contains the letters "AAAAAAAAAAAAAAAAAA" at one point; but that's unlikely indeed.)  See data compression, which is a surprisingly small article but has many links to related interesting articles.  Comet Tuttle (talk) 22:55, 14 March 2010 (UTC)


 * It was a much larger article, before it was compressed. :-) StuRat (talk) 02:53, 15 March 2010 (UTC)


 * Though, just to be pedantic, and useless to the spirit of the question: depending on the nature of the random data, there might be simple ways to "compress" it. If your random data is the first billion digits of pi, it would be easier to just write a short program that generated the data, than it would to compress the data itself. This is related to Kolmogorov complexity—technically you want data that is not only random, but complex, or something along those line (I am not a mathematician... or a computer scientist!). Or put another way, it depends on our definition of "random". But has nothing to do with real-world file compression as far as I know. And should thus be ignored by the poster unless they are just curious! --Mr.98 (talk) 00:39, 15 March 2010 (UTC)


 * Let me introduce you to the work of a chap named Claude E. Shannon on entropy (information theory). Marnanel (talk) 23:27, 14 March 2010 (UTC)


 * Besides "stuff that was chaotic to start with" and "stuff that's already been compressed", another class of files that are hard to compress is encrypted files. If an encrypted file contains any patterns that a compressor can take advantage of, those patterns may also be useful to an attacker trying to break the encryption. This is why whenever encryption and compression are used together, the compression must be done first. 98.226.122.10 (talk) 01:44, 15 March 2010 (UTC)


 * The main problem with the question is that it is ill-posed. You've made an incorrect assumption to start with, which will affect the entire approach you take. As per sun, the sun will NOT go supernova but will instead evolve into a red giant star, following the main-sequence. You may want to read up on stellar evolution and formation and evolution of the Solar System and come back if you have any further questions. By waiting for the sun to go supernova you actually afford yourself an infinite amount of time, thus a brute force method would be guaranteed to work. Zunaid 12:15, 15 March 2010 (UTC)


 * Your observation is completely correct, but as it depends only on the assumption that the Sun will go supernova instead of becoming a red giant, which is a matter of astronomy and not mathematics or computing, I am fairly sure you are being overly pedantic here for the purpose of amusement. =) J I P  | Talk 17:44, 15 March 2010 (UTC)


 * And our Sun going red giant will kill us just as surely as a supernova would, provided we stay on Earth. Unless we develop some really good A/C... :-) StuRat (talk) 12:48, 16 March 2010 (UTC)
 * Yes yes of course that too, but I thought I'd leave that part out...I was also going to link "infinite amount of time" to heat death of the universe but I didn't want to be accused of being TOO pedantic. One needs to maintain decorum, there are strict RefDesk standards for the level of pedantry deemed acceptable. ;) Zunaid 17:45, 16 March 2010 (UTC)

Intro to Computer Science
I'm trying to learn more about computer science and possibly work on some projects. I have a decent grasp of Java but I don't know where to go from here. Should I just keep working in them? So far I've just been doing the Java tutorials and experimenting with some of the articles they have. Any recommendations about what I should do? Much appreciated. 66.133.196.152 (talk) 23:36, 14 March 2010 (UTC)


 * If you've self-taught yourself to have a decent grasp of Java, congratulations. Take on more ambitious and bigger projects that will force you to learn more about particular areas that you aren't comfortable with yet:  Networking, graphics, user interface, performance optimization....  There are plenty of software engineers in the first rank of their profession who are self-taught.  Comet Tuttle (talk) 04:54, 15 March 2010 (UTC)


 * Studying computers is like studying medicine. You can become an eye doctor, a cardiologist, a plastic surgeon, etc. Likewise, you can work as a network technician, a programmer, a computer repairman, etc. That's why I don't like the phrase "computer science" very much. It's too vague. At a university, if you major in computer science, you end up learning mostly programming. But there's so much more to computers than writing programs. Many people earn one or more certifications in networking, security, web design, server administration, hardware repair, Windows, Linux, Mac OS X, etc. You could specialize in any one of those fields (among others) and have more than enough to learn. I'd recommend learning a little bit of everything. Whatever you find interesting, you should specialize in.--Chmod 777 (talk) 05:10, 15 March 2010 (UTC)


 * Some quotable guy once said: Computer science is no more about computers than astronomy is about telescopes. 98.226.122.10 (talk) 06:25, 15 March 2010 (UTC)
 * It's a good idea to see some kind of project through to completion, because finishing something and making it work is surprisingly tough. But you might want to choose your next project to try out a field of CS that interests you.  Here are some fields that haven't been mentioned:
 * programming languages - learn a functional language (actually, every programmer should know a few radically different languages)
 * algorithms - write something that needs to access a huge amount of data in an interesting way, and make it perform well
 * machine learning - there are lots of data sets available for free, like the netflix prize dataset. Try to analyze something about them.  Or for some quick fun, implement dissociated press.
 * theory of computation
 * Good luck! Paul Stansifer 13:14, 15 March 2010 (UTC)


 * Try solving some Project Euler problems if you want to program some mathematically-oriented exercises. If you want to get started in functional programming, I like http://learnyouahaskell.com .  If you want to contribute to Wikipedia's back-end software (it's collaboratively written, just like the encyclopedia), visit http://bugzilla.wikimedia.org and find some items you think you can help with, and join in. Note it is mostly written in PHP, a pretty easy language, but there are some parts in Java if you prefer that.  For computer science, hmm, try CLRS.  For a bit more academic approach to programming, try SICP or CTMCP. 66.127.52.47 (talk) 10:40, 16 March 2010 (UTC)


 * Hey guys, thanks alot! I'll try some of the topics you suggested like networking and GUI. I'm kind of familiar with awt and swing components. I'll also try and look over the Haskall programs. I got bored just learning Java because I wasn't doing anything with it, but next year I'll enter college so I could continue it there I guess. Thanks again. 66.133.196.152 (talk) 03:13, 17 March 2010 (UTC)

Recommendations for a wireless router
I have a Linksys WRT54G router which has for the most part been good to me, but I am looking for a better one. I'd like one with great range and speed, a connection that doesn't drop/need to be restarted, and preferably dual band. I looked up reviews for routers, and one that stands out is the Netgear WNDR3700. Anyone else have suggestions? I'm not averse to spending a lot of money if it means I get a quality router that lasts me long time, so expensive suggestions are welcome. 24.189.90.68 (talk) 23:50, 14 March 2010 (UTC)


 * Have you tried swapping out the firmware for something like Tomato (firmware) or one of the other HyperWRT firmwares? It's apparently compatible with your router and I have heard that it helps improve stability and similar. Gunrun (talk) 09:11, 15 March 2010 (UTC)

Neither are compatible with mine, since it's v.8. 24.189.90.68 (talk) 10:03, 15 March 2010 (UTC)
 * +1 for Tomato. Stable, nice interface, lots of options (if you need them). I'm running it on my WRT54GL. 121.72.169.25 (talk) 10:06, 15 March 2010 (UTC)
 * Consider VPN capability. - Kittybrewster  &#9742;  10:15, 15 March 2010 (UTC)