Computing desk
< December 23	<< Nov \| December \| Jan >>	December 25 >

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.

December 24

Asymmetric and symmetric ciphers

I have a couple of cryptography questions, specifically about asymmetric and symmetric ciphers:

Why are there fewer asymmetric ciphers than symmetric ciphers?
Why are there fewer cryptanalytic attacks discussed (available?) for asymmetric ciphers than for symmetric ciphers (according to the ciphers' articles anyway)?
Why are some asymmetric ciphers (ElGamal and RSA, and probably more) still considered secure (still in wide use, that is) even after more than 25 years in existence, while no symmetric cipher (except the one-time pad) is still considered secure after such a long time in existence?
Can we say that, in general, asymmetric ciphers are more secure than symmetric ciphers?
How fast (slow) is a typical asymmetric cipher, in terms of encryption speed (MB/s) on a modern computer?

Thanks for any answer. 118.96.163.93 (talk) 03:07, 24 December 2010 (UTC)[reply]

I'm not sure I know the answers to these questions, but here are some semi-educated guesses. First, DES (which is symmetric and about as old as the oldest asymmetric ciphers) is still considered strong after all these years. Its only major problems are the short key length and slow speed in software—both of which were designed into it with the goal of giving an edge to well-funded agencies who could afford custom hardware to break it (specifically the NSA). Without that deliberate sabotage it would be competitive with AES (though AES would still beat it on speed). So I don't think it's really true that asymmetric ciphers last longer.

Second, asymmetric ciphers are generally many orders of magnitude slower than symmetric ciphers, but it doesn't matter because you can encrypt your message with a random symmetric key and encrypt only that key with the asymmetric cipher, giving you the benefits of the asymmetric cipher with the speed of the symmetric cipher (plus a fixed per-connection overhead). So there's more pressure to come up with new, faster symmetric ciphers, since they give you a bigger practical benefit.

If anything, asymmetric ciphers may be less secure than symmetric ciphers. All popular asymmetric ciphers are breakable by quantum algorithms, while no symmetric cipher seems to be, which suggests that there may be a poorly understood weakness in the whole concept of asymmetric cryptography. -- BenRG (talk) 06:00, 24 December 2010 (UTC)[reply]

Thank you for your information. 118.96.158.21 (talk) 02:38, 30 December 2010 (UTC)[reply]

Inkjet printers

One more question: If I used an inkjet printer's "print head cleaning" function, which wastes ink, where did the wasted ink go? Thanks. 118.96.163.93 (talk) 03:07, 24 December 2010 (UTC)[reply]

Probably onto the internal components in the printer, that are around the cartridge. Sumsum2010·T·C 04:23, 24 December 2010 (UTC)[reply]

There is a waste station— basically a sponge that absorbs the waste ink. Some printers track the cleaning count and will shut the printer down when it figures the wast station is full. ---— Gadget850 (Ed) ^talk 05:13, 24 December 2010 (UTC)[reply]

See Inkjet printer#Cleaning mechanisms. Although in 15 years in the printer biz, I never heard it called a spittoon. ---— Gadget850 (Ed) ^talk 05:20, 24 December 2010 (UTC)[reply]

Thank you for your explanation. 118.96.163.93 (talk) 05:50, 24 December 2010 (UTC)[reply]

Oh, and since you've been in the printer business for 15 years, let me ask you another question, if you don't mind: Do you think that newer printers are less reliable (that is, fail more often and more quickly) than older ones?

To tell you the truth, I (and many people I've met) have a belief that "newer consumer electronics are optimized for production cost and not for longevity and reliability, thus fail more often and more quickly." Based on your experience, do you think this belief have some truth on it? 118.96.163.93 (talk) 07:08, 24 December 2010 (UTC)[reply]

Depends on the product— production, mid-range or consumer. Let's use Xerox as an example (not my company, but they have been a customer and an OEM). Their production equipment should last at least ten years, and I know of products that have been in place for fifteen or more.[1] Their mid-range office equipment is pretty robust, and has at least a ten year life span.[2] Line printers and serial matrix printers can be even more robust. Consumer products are a very different story. Production and mid-range products ten to have a long production span, with new models released every year or two; consumer products seem to come and go by month. A $59 ink jet printer that or a $150 laser printer is simply not going to last as long as a $2000 laser printer. Those low-end ink jets have a much higher TCOP. We don't have an article on total cost of printing (TCOP), but here are a few sources.[3] [4] [5] ---— Gadget850 (Ed) ^talk 15:33, 24 December 2010 (UTC)[reply]

Excellent explanation, thanks. 118.96.158.111 (talk) 01:00, 26 December 2010 (UTC)[reply]

Publishing summer research project in machine learning

Dear Wikipedians:

I recently finished a summer research project in the area of kernel trick, machine learning, artificial intelligence and was hoping to get it published. I do not care about the impact factor of the journal that I can publish in, as long as I can have my project published somewhere. I am wondering if there is a registry out there that sorts the journals in the areas of kernel trick, machine learning and artificial intelligence by their "ease" of publication? Are open-access journals easier to publish to than traditional journals?

If you have any recommendations of specific "easy" journal titles I would greatly welcome them too!

Thanks for all your help!

174.95.169.52 (talk) 16:48, 24 December 2010 (UTC)[reply]

You mean disreputable, non-peer reviewed journals?Smallman12q (talk) 22:27, 28 December 2010 (UTC)[reply]

html (query string)

I've noticed that can you add something like "?whatever=123" at the end of a .html url, similar to php, and the page will display as normal. For example http://example.com/index.html?whatever=123 I tested this on a local .html file with no php software running and it also worked. What exactly is the purpose of this? Does it have any practical applications (like supplying a variable to html or something)? —Preceding unsigned comment added by 82.44.55.25 (talk) 20:04, 24 December 2010 (UTC)[reply]

It is the Query string of the HTTP request. It supplies variables to the server. The server doesn't have to do anything with them, though. The variables can be accessed either with server-side scripting (like PHP) or client-side (with Javascript). Pure HTML by itself cannot access the variables, because it doesn't do any scripting. (I added (query string) to the header so that it doesn't conflict with the other question titled "html" on this page.) --Mr.98 (talk) 20:10, 24 December 2010 (UTC)[reply]

You're right, but I think it confuses the issue to say "pure HTML by itself" rather than "static HTML pages". After all, server-side code which does understand the query should still be producing pure HTML. Marnanel (talk) 20:25, 24 December 2010 (UTC)[reply]

The page author can access the query string via JavaScript. Thisisnotatest (talk) 05:23, 27 December 2010 (UTC)[reply]

I would like to try a different explanation; I think yours assumes some knowledge that the OP may not have.

A URL is, at its basic level, a string sent from a computer out into the internet in order to get a response from a specific computer (or set of computers). The main portion (www.google.com) specifies the web site from which the response is desired; other parts of it specify other things. By convention, characters following ? are information for the computer responding to the URL. Also by convention, things like "name=value" indicate a thing's name and a thing's value, so that you can enter things like "country=US" or "zip=10101"; these are data the server can use to help figure out how to respond to your request.

A server that is not looking for any such data will respond to your request the same way regardless of what you put after the "?"; a server that is expecting things is likely to expect specific things in specific formats, so you can't just make up names and values for those URLs.

rc (talk) 04:38, 25 December 2010 (UTC)[reply]

Learning C++

I am taking C++ lesson from here with the aim of making different kind of software and games in near future. Is that a good start? I am already in "command line arguments" lesson. What else I should follow along with these lessons? Any book that might help ? thanks--NAHID 21:36, 24 December 2010 (UTC)[reply]

I recommend the C++ Programming wikibook. If you are compiling on a Linux or Mac OS X system, I recommend that you use and learn gcc, a free and open-source compiler/software build toolchain. The official GCC website - http://gcc.gnu.org - also has excellent documentation (though it is not intended as a "tutorial.") You can also use GCC on Windows (using MinGW or Dev C++), but I do not recommend this for novices, because you will spend a lot of extra effort setting it up. Instead, if you are developing on Windows, I recommend you get the free-to-use Visual Studio Express 2010, available from Microsoft. (You want the C/C++ version). And, I also recommend the Microsoft Technet website - http://technet.microsoft.com - which has many free-to-use online tutorials, books, programming documentation, and other resources specific to the Windows C++ platform and the APIs provided through Windows (and the newer Windows APIs provided via .NET / CLR environment). I have found that during the early stages of learning a new programming environment using the "official" resources, instead of blindly searching the Internet at large, is a more disciplined approach, and is generally more efficient. It also exposes you to exclusively-correct-information - which the Internet at large cannot guarantee (this is especially critical when you are still too novice to evaluate technical advice and/or facts for yourself). Sadly, there is a plethora of bad advice and (unintentionally) incorrect technical information spread across the internet, so save yourself a headache and search the best resources first. Nimur (talk) 16:51, 25 December 2010 (UTC)[reply]

Aside from learning C++, depending on which OS you plan to devolp for, you'll need to learn the relevant API's such as the Win32 API. For games, you'd also need to learn the graphics rendering libraries such as DirectX or SDL.Smallman12q (talk) 22:24, 28 December 2010 (UTC)[reply]

Updating Android OS

hi.. I have micromax A60 mobile phone and it came with android 2.1 Eclipse.. can I update it to android 2.3..... how can i do so...? —Preceding unsigned comment added by 220.225.96.217 (talk) 21:54, 24 December 2010 (UTC)[reply]

I believe you will need to talk to your retailer and ask them to do it, depending on your carrier (In Australia, it's not always that easy) General Rommel (talk) 06:17, 25 December 2010 (UTC)[reply]

According to [6] and [7] the Micromax A60 might get an upgrade to 2.2 in 2011. Needless to say, 2.3 is likely a long way a way if it ever comes. There are sometimes independent efforts to port a newer Android to a phone but from what I've heard (not something I've looked in to myself) these are usually stymied by driver difficulties. [8] suggests such efforts are still in a very early stage in any case, perhaps not surprising since the phone is fairly new. More likely would be leaked official test or beta firmware (although that will only likely be 2.2). Nil Einne (talk) 20:02, 25 December 2010 (UTC)[reply]

EEE PC hostname

I got an EEE PC and when I connected it to IRC (local daemon) its hostname resolved to 'EEEPC-M'? How is that possible? It's on my local network, but I didn't manually change the IRCd configuration or anything. 62.255.129.19 (talk) 22:17, 24 December 2010 (UTC)[reply]