Wikipedia:Reference desk/Archives/Computing/2010 December 22

= December 22 =

FCC's new Internet traffic rules: What does that mean for us?
First, I hope someone's made an article on that by now. If so, but that still redlinks, please redirect.

http://www.theglobeandmail.com/news/technology/us-regulator-adopts-internet-traffic-rules/article1846054/

What will it mean for you and me now that this FCC regulation is in effect???

My ISP is Cox Communications. Will being part of Cox restrict me on anything now? --70.179.178.5 (talk) 01:50, 22 December 2010 (UTC)


 * From what I heard on NPR today, your ISP cannot block or limit what you see on the net. Dismas |(talk) 04:30, 22 December 2010 (UTC)


 * Listened to the same broadcast. They said except for illegal things. Which will eventually be most things. --71.240.162.87 (talk) 06:21, 22 December 2010 (UTC)
 * ╟─ Treasury Tag ►  duumvirate  ─╢ 12:05, 23 December 2010 (UTC)
 * THE GOOD: They can't block or slow down access to sites they don't like, such as their competitors. They also can't use that ability to blackmail sites to pay them money so they will have full speed restored.  This would have been bad for free sites (like Wikipedia) and especially those with a need for a lot of bandwidth (like YouTube), which couldn't afford to pay for high speed, and thus would slow to a crawl and lose users.


 * THE BAD: This might tend to overload the entire Internet, if some sites irresponsibly send large amounts of uncompressed data, knowing they won't be punished for sending such a big load down the "pipes". Thus sites that could have paid for top speed, if allowed, will now be slowed down, too.  This would include sites that sell products, especially large-ticket items, like cars.


 * THE UGLY: Sites that send nasty, but legal, stuff, like lots of porn, may also be protected from discrimination by ISPs.


 * See net neutrality for more info (although the article hasn't yet been fully updated with info on the new law). StuRat (talk) 07:26, 22 December 2010 (UTC)


 * It's not clear to me that your claim they can't 'punish' irresponsible sites is accurate. The source linked above says they can slow down traffic for traffic management (provided they don't practice "unresonable discrimination"). It's not clear and likely to be highly disputed what this is but it doesn't seem that unlikely they would be allowed to slow down services they argue are wasteful in their data usage because of excessive usage of uncompressed data. Of course ultimately one would assume they can rely on whatever peering agreements they have and also what they charge their customers to ensure their are getting fair payment for traffic they route, whether it's large amounts of uncompressed data or the same size but highly compressed is arguably moot. Nil Einne (talk) 11:59, 23 December 2010 (UTC)

Apple Apps that are removed from the AppStore
When an app is removed from the iTunes store, such as this one from WikiLeaks, what happens to those who have already downloaded the app? Does it cease functioning? Does it continue to work? Dismas |(talk) 04:27, 22 December 2010 (UTC)


 * afaik Apple does not remove apps from someone's phone if they already downloaded it even though I am pretty sure they have the capability to do so. I can't remember for sure but I think I remember them doing that at some point and they got a lot of flak about it so they stopped. So afaik it should still work if you already downloaded it. For what it's worth I have downloaded a few apps on my Android phone that got pulled from the market later and they still work but I realize Android is an entirely different situation than the iPhone so I don't know for sure. Thingg &#8853; &#8855;  05:12, 22 December 2010 (UTC)


 * Apple could indeed delete apps installed in iPhones. And this is possibly irrespective whether the app was withdrawn, removed, censored, DMCA-ed, etc. from the App Store. 118.96.167.217 (talk) 05:18, 22 December 2010 (UTC)

Tables in HTML and wiki text
Due to the order of the hierarchy (Table > Row > Cell) of HTML and wiki text, it's very easy to add a new row anywhere in the table, but it is very hard to add a column because the code line for each row has to be modified... so what's the solution? What's the easiest way to add a new column when writing out HTML manually? I already have a hypothesis... maybe tables can be written out with a Table > Column > Cell hierarchy instead? Although... that would just create the other problem of difficulty adding new rows! Input your solutions please! Roberto75780 (talk) 09:26, 22 December 2010 (UTC)


 * Format the text nicely and/or use an editor that supports columns. Emacs "rectangle" commands are very useful here. --Stephan Schulz (talk) 13:14, 22 December 2010 (UTC)


 * Agreed. Reversing the order would just reverse the problem, so that adding, deleting or changing a row would be more difficult.  Since this is normally what happens most often, this would be a problem.  The other advantage of laying it out as a table in the source code is that it's easier to view there:


 * BAD:

Washington,District of Columbia,United States Toronto,Ontario,Canada Lansing,Michigan,United States Sacramento,California,United States Yellowknife,Northwest Territories,Canada


 * GOOD:

Washington    ,District of Columbia     ,United States Toronto       ,Ontario                  ,Canada Lansing       ,Michigan                 ,United States Sacramento    ,California               ,United States Yellowknife   ,Northwest Territories    ,Canada


 * Be sure to put plenty of room between the columns, so you don't have to increase it when a longer cell is added. Tabs can be used to line up columns nicely, but some editors may have problems with those or interpret them differently.  Sorting things in a more logical order in the source code would also make sense. StuRat (talk) 16:16, 22 December 2010 (UTC)

Why so many different computer languages?
Most computers use the Von Neumann architecture I think, so instead of constantly re-inventing the wheel, why are there not just a small number of languages that are written for specific purposes, instead of lots of similar ones? 92.29.126.195 (talk) 12:17, 22 December 2010 (UTC)
 * There are many reasons for programming languages to exist. They may have been created for hobbyist, educational, or commercial purposes; they may have intended to include language features that optimize some particular part of a software cycle (like compiler-performance, software-maintenance, or human-readable code).  In fact, almost all computer languages are functionally equivalent to each other; this is a fact of being Turing complete (meaning that any algorithm could be expressed in any Turing-complete language, perhaps with more characters needed).  So, even something nutty like Haskell is essentially a functional equivalent of something very different like Java.  (However, expressing the same statement will look very different in the two languages, and each will be more compact at expressing certain types of ideas).  On the other hand, other languages really are virtually identical, aside from some syntactic sugar - so, C# and Java are essentially the same language (albeit with different keywords, a different system library, and so forth).  Some languages go to great effort to hide machine characteristics (like Java); others make huge efforts to expose them (like pure C, or even more so in machine-specific assembly languages and macro languages for specific hardwares).  Any time a language has been created, and a user community of at least one person exists, there's a potential for useful tools and software to spring up in that language; so legacy software and community momentum keep the language alive, even if it is flawed or has a suitable replacement.  But in general, even though many thousands of computer languages do exist, it's safe to say that no more than a dozen or two are relevant.  Furthermore, except for those pesky esoteric programming languages, if you understand the logical program-flow of computer programming, the difficulty of hopping to a totally new language is usually pretty low (provided that the language is reasonably-well-designed).  Nimur (talk) 13:30, 22 December 2010 (UTC)


 * For the most part, there is only one computer language - assembly. Everything else is an abstraction of assembly to make it much easier for programmers to do their job.  Different programmers working on different projects have different requirements.  So, you end up with many different languages.  Because they get compiled down to the same base language for basically the same hardware, the languages all appear to be very similar. --  k a i n a w &trade; 13:22, 22 December 2010 (UTC)
 * Assembly itself is somewhat of an abstraction of machine language.Smallman12q (talk) 14:18, 22 December 2010 (UTC)


 * An additional reason not mentioned is politics and commercial interests. At one point Microsoft felt threatened by Java, which was not under their control, so they released a very similar language called C.  C# has since been made available for systems besides Windows, but many people decline to use it because they don't trust Microsoft.  Java itself might get that treatment due to recent actions of its current owner, Oracle.  --Sean 19:39, 22 December 2010 (UTC)


 * In short, in random order: problem, pride, popularity, pay, profession, politics, product (positioning and penetration), profits, and patents. :-) 118.96.163.93 (talk) 07:07, 24 December 2010 (UTC)

I take it there is no theory which tells you what the optimum language design for a particular use would be. 92.24.178.208 (talk) 13:05, 24 December 2010 (UTC)
 * Not really; mostly, because "optimum" is not well defined in that context. However, an experienced software engineer or computer systems designer could probably help narrow down to one or two "most practical" languages, just like an experienced woodworker could help you select a good saw or chisel for a specific task.Java or C# are generally used for applications programming; C is usually used for systems programming and embedded computers; MATLAB and FORTRAN are generally used for scientific and numerical algorithms; PHP and Ruby are best-suited to light-weight server-side web applications.  Do you have a specific problem for which you would like help evaluating different programming languages / tools?  Nimur (talk) 16:30, 25 December 2010 (UTC)

Operation of redirects on other websites
When I go to http://www.bn.com, the old website for Barnes & Noble, I'm immediately redirected to http://www.barnesandnoble.com. Is the HTML for this similar to what's used to have http://en.wikipedia.org/wiki/WP:RD/C redirect to http://en.wikipedia.org/wiki/Wikipedia:Reference_desk/Computing? Obviously I don't get a little "(Redirected from www.bn.com)" message at the top of the Barnes and Noble page, and the URL of the page immediately changes to barnesandnoble.com (instead of staying "WP:RD/C" until I click the "project page" tab), but otherwise I can't see any differences. Nyttend (talk) 12:21, 22 December 2010 (UTC)


 * It's mostly done in DNS configuration or the web server settings rather than in HTTP, but it can be done with a "meta" tag in HTML. URL redirection explains. -- Finlay McWalter ☻ Talk 12:31, 22 December 2010 (UTC)


 * Indeed. And redirections in wikis (like Wikipedia) are configured in and processed by the wiki software (which, in Wikipedia's case, is MediaWiki), not the DNS server, HTTP server, or your web browser. 118.96.160.94 (talk) 13:53, 22 December 2010 (UTC)


 * The http://www.bn.com web site issues a 301 response (permanently moved) which redirects to http://www.barnesandnoble.com/. Technically, http://en.wikipedia.org/wiki/WP:RD/C issues a 301 redirect to http://en.wikipedia.org/wiki/Wikipedia:RD/C, and that page returns a 200 status, but there's no redirection that I see after that. --  JSBillings  14:22, 22 December 2010 (UTC)

Usb drive in the mail
Hi, I'm wanting to send a usb stick with some large files on it to a friend in dubai (I'm in the uk), the drive is very small and light (sandisk cruzer blade), would the files on it remain safe?, or is there a chance any data on it may be wiped, or someone would look through the files to check it? Thanks--77.44.28.237 (talk) 12:49, 22 December 2010 (UTC)


 * If you don't want it read, use something like Pretty Good Privacy to encrypt it (providing this is not illegal out there). USB sticks are about the best way to send files through the post that I can think of -as they are so tough. Keep a backup copy too. Don't send the key until it safely arrives at the other end. Use a very long key. Better still use digital_steganography to conceal the files.--Aspro (talk) 13:18, 22 December 2010 (UTC)


 * Regarding x-rays and other technologies: it is unlikely that any x-ray or other scanner device would accidentally wipe the data off of a USB flash-drive. (ref. Tips for Caring for Your Flash Memory).  It may be legal for customs, border control, or other security agents to view or copy your flash-drive contents; it is rarely legal for anyone to delete them; but you can encrypt your data (as Aspro advised above).  This will make the data irrelevant and unreadable to any agent who does not have the decryption key.   Nimur (talk) 13:51, 22 December 2010 (UTC)


 * However, sending large amounts of data there, in an unreadable form, may get you put on a terrorism watch list, since that's the type of thing a terrorist might do. StuRat (talk) 15:56, 22 December 2010 (UTC)

I think BlackBerry was recently prohibited in Dubai due to its encryption. Quest09 (talk) 17:47, 22 December 2010 (UTC)

VLC
I have a WinTV-NOVA-T USB. I read that VLC can be used to watch tv through it, but I can't work out how. I followed several guides on setting the capture device to direct show dvb but it doesn't seem to work. What am I doing wrong? 82.44.55.25 (talk) 14:56, 22 December 2010 (UTC)


 * Yes, this is pretty easy to do actually. Just open VLC player, and from the Media menu, select Open Capture Device, and choose your TV tuner card in that dialog box. You will also notice your webcam is in the drop down list, as it's also a capture device and can be played by VLC. Another interesting thing is that when you use Yahoo, Skype etc you can select any capture device, including your TV tuner card, as your "webcam" from their respective options menu. Roberto75780 (talk) 21:48, 22 December 2010 (UTC)
 * I got that far, but just clicking play doesn't work; no picture or anything is displayed. There are "transponder/multiplex frequency" and "terrestrial bandwidth" settings, but I don't know what to do with them 82.44.55.25 (talk) 22:06, 22 December 2010 (UTC)

Pictures Gone
Since a few days ago no pictures appear on my WP pages. I see just a little sign instead. Anything wrong with my ie8 settings? I have this same problem in Google Chrome as well. Thanks for comments. Omidinist (talk) 16:40, 22 December 2010 (UTC)


 * Pics that no longer display usually mean my computer is out of memory, and a reboot fixes the problem. Have you tried that ? StuRat (talk) 02:08, 23 December 2010 (UTC)


 * I did it and it worked. Thanks a lot. Omidinist (talk) 05:18, 23 December 2010 (UTC)


 * You're quite welcome. StuRat (talk) 06:31, 23 December 2010 (UTC)

Analyzing patterns in general
I know that tools like BLAST can be used for different purposes, but can you use it for analyzing non-sequential structures? (like a network). Are there other pattern analyzing software tools available (not to be confused with Software_analysis_patterns) that are so developed as BLAST? (specially for analyzing graph-like structures, for example a network).Quest09 (talk) 17:40, 22 December 2010 (UTC)


 * BLAST (and FASTA) are simplification/estimates of performing a full Wagner-Fisher analysis for either a local or global alignment. You can ignore the simplications and perform a full analysis.  What you are doing is trying to align (locally or globally) two sequences.  For example, if you have 12345 and you try to align 153, you will align 1 to 1 on both.  If it is a local alignment, you will align the 3's together.  If it is a global alignment, you will align the 5's together.  Then, you also end up with a Levenshtein-like distance between the two alignments.  If you don't have a background in sequence alignment, I can expand on this further. --  k a i n a w &trade; 17:45, 22 December 2010 (UTC)


 * No, I don't know anything about sequence alignment, but I am not interested in bioinformatics directly, but in analyzing patterns in general. Can you analyze a road-map or city street grid with such sequential alignment tools? Or would the similar patters be irrelevant? 18:01, 22 December 2010 (UTC) — Preceding unsigned comment added by Quest09 (talk • contribs)


 * You can use sequence alignment for a lot of things, such as comparing a Wikipedia editor's contributions to a large number of existing websites to see if the editor's contribution is pretty much copied directly from a small part of another website. You'd be performing a local alignment of the editor's contribution with each website.  After alignment, you check the distance.  If the distance for the local alignment to one of the websites is negligible, the editor's contribution is a copy of a subsection of the website.  The key here is that you need a sequence, not a pattern.  In DNA, the sequence of letters is obvious.  In an editor's contribution, the sequence is also the sequence of letters that form each word and eventually each sentence.  I've seen this used to compare the order in which people read books - the order in which a person reads a book is a sequence of books.  I've seen it used to analyze roads that people take - the order of roads and intersections becomes a sequence.  So, you need to decide how you will encode your sequences.  Then, you can compare two sequences to one another for alignment and/or difference. --  k a i n a w &trade; 18:08, 22 December 2010 (UTC)


 * But, imagine that you have a city map (infinite number of sequences between point A on the West and B on the East), and you want to find the shortest route between point A and B. Would you generate a number n of sequences and compare them? In the examples that you gave, you already have the sequences (DNA, lists, whatever) and are comparing them. In this example, it is not clear which sequences exist, you just have the layout of the graph. How to find the shortest path (sequence) here? Quest09 (talk) 18:21, 22 December 2010 (UTC)


 * You need to solve the Graph_theory with an algorithm for solving the Shortest_path_problem80.58.205.34 (talk) 18:57, 22 December 2010 (UTC)


 * You began with BLAST, which is a sequence alignment tool. To align sequences, you must know the sequences. If you want to look for shortest paths, you shouldn't be using a sequence alignment algorithm. --  k a i n a w &trade; 19:10, 22 December 2010 (UTC)

Apple products
Why doesn't Apple append a simple model number to the name of each version of the iPod, iPhone, iPad, and even the Mac computers? The "generation numbers" people tack onto them are a little inconsistent and confusing, especially since the introduction of iPhone 3G, in which its 3G phone technology not the third generation of iPhone. It would make referring to (or searching for!) a specific model much easier. A simple (and unique!) 4 digit model number after the name like BlackBerry and Nokia phones would be so convenient. In the first few years of iPod it was even worse, as the flagship model was always officially called "iPod" with no further qualification whatsoever. Names like "touch" "3G" "3GS" etc have helped a lot but there is still ambiguity about which generation of "touch", "nano", or "shuffle" one is referring to. So what's up with the over-simplified and ambiguous naming? I know they have some sort of long model numbers but why not prominently displayed in their names, (like blackberry and nokia many other electonics) Roberto75780 (talk) 22:54, 22 December 2010 (UTC)


 * Apple does a fantastic job of presenting a simple product line to normal people who don't spend much time worrying about the specifications of their phone, or its features, or its generation. And they're making billions of dollars, mostly at Blackberry's and Nokia's expense.  Do you really think that by adding little numbers to the names of their phones they'd sell more of them?  If not, "do as Nokia does" is a silly move - Nokia is dying. 87.115.159.188 (talk) 02:57, 23 December 2010 (UTC)


 * Actually, Apple does assign model numbers to their products. I know MacBooks print the model number on the back of the computer. To get the model number of an iPhone or iPod, you go to Setttings --> General --> About. The model number of an iPhone might be MB528C, for example. Most phones print the model number underneath the battery, but Apple doesn't let you remove the battery from your own phone. That way, like you said, consumers don't have to worry about replacing their batteries. They can just buy a brand new phone. Real simple.--Best Dog Ever (talk) 03:22, 23 December 2010 (UTC)

Usage statistics
I am looking for the following statistics, with references if possible:


 * What fraction of the full CPU capabilities of a computer is actually exploited, on average, in a given year?
 * What fraction of the full internet bandwidth capacity does an average person use, on average, in a given year?
 * How much does a person spend every year, on average, on computer hardware and software?

I am also interested in general references which provide similar information.

Cheers, 92.156.151.218 (talk) 23:20, 22 December 2010 (UTC)


 * Don't know the answers to these but I'd like to point out, for the bandwidth question, that wouldn't reflect the percentage capacity of the system being used, because if everyone tried to use up all of there bandwidth, the speeds would drop way down... so even if a typical person uses 10% of their bandwidth, the system (the internet service provider in that region) could be (and probably is) running at well over 50%.

You can calculate this value for yourself easily by getting a program that keeps track of your network traffic and tells you how much data you have used this month (actually most smartphones already have such a counter), and then devide that by the maximum theoretical ammount based on your ISP's claimed speed. As for CPU capacity.. there's so many programs that display the current load as a percentage and as a graph over the last hour, day, etc. im sure there would be more advanced ones that can average it out for any given time frame. Roberto75780 (talk) 07:47, 23 December 2010 (UTC)