Wikipedia:Reference desk/Archives/Computing/2007 April 22

From Wikipedia, the free encyclopedia
Computing desk
< April 21 << Mar | April | May >> April 23 >
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.


April 22[edit]

Smiley Face[edit]

When people post that they hang out at the smiley face, where is that? Thanks 72.84.30.159 02:30, 22 April 2007 (UTC)[reply]

":-)", ":)", ":D", "", etc. just means someone is happy; or that the remark is not intended to offend. These faces are called emoticons. I've not heard of someone hanging out at :), but someone may be happy to be hanging out. --h2g2bob 03:04, 22 April 2007 (UTC)[reply]

Do you mean they said "I'm hanging out @:)" ? If so, perhaps the at sign represents hair on the smiley face. StuRat 07:27, 22 April 2007 (UTC)[reply]

CSS instead of tables: left align, center, and right-align, all on one line[edit]

I want to have one line of text with some text at the far left, some text at the center, some text at the far right. In the old days, I'd do this by making a table row with three columns and being done with it. But tables for layout are evil, I'm told, so I thought I'd try to do it with CSS-- thus far, I haven't stumbled upon the answer. What am I missing? --Wouldbewebmaster 03:43, 22 April 2007 (UTC)[reply]


Something like this? --h2g2bob 03:53, 22 April 2007 (UTC)[reply]

<div id="main" style="text-align: center">
  <div style="float: left"> LEFT </div>
  <div style="float: right"> RIGHT </div>
  CENTER
  <!-- make sure it all stays inside the main div -->
  <div style="clear: both"></div>
 </div>


You could use display: table, display: table-row, display: table-cell, etc. While using tables for layout is evil, I believe using CSS table layout isn't (the main problem with HTML tables is mixing content with formatting; tables are formatting, so they should be done in CSS). --cesarb 16:38, 22 April 2007 (UTC)[reply]

display: table isn't IE6 friendly (and I wasn't sure it was IE7 friendly either)

You should use css classes to set up the columns, not inline stuff (unless you're on wikipedia... bleh)

I hacked this up real quick with a Real-time HTML editor. I didn't try it in IE but I can't think of anything here that wouldn't take care of it in IE. you may want to pad the middle column, but there may be problems of IE doubling the margin. I can't remember the fix. (positioniseverything.net is a good reference)

<style>
.container{
   _height: 1%;  /*IE hack */
   width:400px;
}
.container:after{  /* float container hack */
  clear:both;
  content: "."; 
  display: block; 
  height: 0; 
  clear: both; 
  visibility: hidden;
}

.leftcolumn{
float:left;
width:20%;
text-align:right;
}
.middlecolumn{
float:left;
width:40%;
text-align:center;
}
.rightcolumn{
float:left;
width:20%;
text-align:left;
}
</style>
<div class="container">
<div class="leftcolumn">This is a right aligned left colun</div>
<div class="middlecolumn">This is a center aligned center column</div>
<div class="rightcolumn">This is a left aligned right column</div>
</div>

Root4(one) 22:46, 22 April 2007 (UTC)[reply]

Image question[edit]

How do you scan an image from TV show or movie? I'd like to scan a few fair use screenshots for some film/TB articles. Incidentally, I looked into the rationales needed and policies. Aaron Bowen 07:12, 22 April 2007 (UTC)[reply]

Most of those are taken from the computer via print screen. For Windows, you can take a screenshot from Windows Media Player, but first you need to change some settings, see: [1]. - Akamad 08:48, 22 April 2007 (UTC)[reply]
Thanks! Aaron Bowen 13:18, 22 April 2007 (UTC)[reply]
Or you could just use VLC Media Player and use its built-in screencapture tool. Aetherfukz 11:19, 23 April 2007 (UTC)[reply]

Request for details on Error message "Skein 0.0.3 error code 2"[edit]

I have Broad band internet connection & using WinXP...while browsing continously for about 15 minutes, the connection is cut off showing the message "Skein 0.0.3 error code 2..contact the administrator etc". I need to log out and then log in again to resume browsing. I have contacted the administrator but they say it is a problem with the system. The hardware engineers who have checked the system have also not been able to identify the problem. What does this error pertain to? I have checked the internet for Skein 0.0.3...all I am getting is details of baseball scores(?). Please help. rimenon—The preceding unsigned comment was added by 202.164.146.103 (talk) 07:45, 22 April 2007 (UTC).[reply]


i think its a os.i also get this messege. it runs on my isp's computer the logout i think happens because whatever software you are using for logging in(my isp has given me 24 online client") logs you out after a set time of inactivity . maybe you can set it using the "logging in" software or you could ask your admin to increase your inactivity timeout... or if you log in using a web pageit is because your web browser does not have the ability to "keep-alive" a connection ... you can probably set it to keep alive ..

Joost invitation[edit]

Does anyone here have a Joost account? Could I, please, receive an invitation? Thanks. --Taraborn 12:27, 22 April 2007 (UTC)[reply]

I have a Joost account, and so might have some invitations. I'm away from home atm, so I can't find out, or send you one 'till the weekend. Remind me on my talk page :) --saxsux 16:23, 26 April 2007 (UTC)[reply]

Optimal encoding of (not with) a Huffman tree[edit]

I have a scheme which uses (2 + m)n + 1 bits to store a Huffman tree (where n is the number of symbols, and m is the number of bits per unencoded symbol). Is there a more compact scheme? (I have the beginnings of a (1 + m)n + constant scheme, but I'm not sure if it's workable). --User:Taejo|대조 13:07, 22 April 2007 (UTC)[reply]

If I am not mistaken, every binary tree with unencoded words at the nodes is a possible and distinct Huffman code. This means that you will need to store at least a binary tree (plus your codes). According to the binary tree article, this takes 2 bits per node. Every binary tree with n leaves has 2n - 1 nodes total, creating a total size of 4n - 2 bits for the tree. After this, you can simply store your codes in post-order traversal (size m*n bits, the very minimum, since you have to store the unencoded codes anyway). Assuming the unencoded words are stored in a fixed length representation, you'd need some constant amount of bits c to delimit the list of words and the tree representation. The total works out to mn + 4n - 2 + c , or n(m+4) - 2 + c. I may have made a mistake somewhere (I'm no expert on this stuff), but I don't think you can get any smaller than this. Every possible bitstring in this format is a distinct huffman tree and vice versa, any smaller and, you would't have enough room to assign a representation to every possible tree. Could you tell us your (first) scheme ? risk 17:27, 22 April 2007 (UTC)[reply]
With pleasure. The trick is that not all binary trees are Huffman trees. Huffman tree nodes are either leaves or have two children. Which means you only need one bit per node: 0=leaf, 1=nonleaf. And you can eliminate the constant for the separator by simply storing the m-bit words immediately after the 1 bits for nonleaves. The (1 + m)n scheme (which I know realise is workable, and the constant is -1) relies on the two-queue construction algorithm. You store a bit for Q1 or Q2, and since leaves only come out of Q1, you can store the m-bit words immediately afterwards. Since (as far as I can see) the algorithm takes (n-1) iterations, you store (1 + m)n-1 --User:Taejo|대조 07:53, 23 April 2007 (UTC)[reply]
Oops, that scheme needs 2 bits per iteration, since two items are taken each iteration. Since the codes are equally optimal if a tree is left-right mirrored, only one of (Q1,Q2) and (Q2,Q1) is needed. As a terminator we can just take an extra node from Q2, since as soon as it has less than 1 element, the algorithm is complete. So we can use n trits or n log23 bits to encode the tree shape plus mn to encode the values. Still not sure if this is optimal. --User:Taejo|대조 16:04, 23 April 2007 (UTC)[reply]

Computer Spec, difference between processors/etc[edit]

I'm in the middle of trying to find a suitable upgrade for my parent's PC. They have a Dell 8250 (http://support.dell.com/support/edocs/systems/dim8250/specs.htm#1084976) and we want to upgrade (ideally) RAM to 1gb, add a reasonably decent Nvidia graphics card, upgrade the processor and maybe add a second hard-drive. From looking at the spec i'm not sure what things I can add that are 'compatible' (the ram Is DRAM? Can't figure the max processor it can accept and it has AGP 4x/PCI so not sure which graphics card is best to upgrade to). Any help/points in the right direction welcomed. Sorry I know this isn't 'exactly' what reference desk is for. ny156uk 16:14, 22 April 2007 (UTC)[reply]

Nyan? Not what the reference desk is for? I don't see the problem with this question. It's not homework and it's general help. Anyways, looking at the page you linked, you want to upgrade RAM to 1GB, but what does it have right now? It seems to take "PC800 or PC1066 RDRAM (non-ECC)". The PC800 looks to be about USD$200 for 512 megs. You can get up to a 3.06GHz Pentium 4 running with a 533MHz bus. The chances are, you can upgrade the harddrive with any ATA type because it mentions 2 3.5" bays. I say that because it might not have more molex power connectors for another device (or only has one IDE channel, which isn't too likely, but does happen on mass produced computers), but then you can split it, although be careful about the power supply being only 250 watts. As for video cards, even most new cards have AGP8X/4X versions. But like before, remember that you only have a 250 watt power supply, so a lot of the high end ones can't be used in your parents' computer without upgrading the power supply too. My advice? Buy a cheap refurbished computer or an older new one on ebay or something, because that will end up cheaper. --Wirbelwindヴィルヴェルヴィント (talk) 16:36, 22 April 2007 (UTC)[reply]
Thanks for the response Wirbelwind. I found the RAM to be extremely costly for some reason, compared to DDR/DDR2 variants. I would like to end up with a computer something like...3ghz, 1gb RAM, 128mb Nvidia Graphics Card (or better!) at minimal price. Already have 120gb HD, DVD-ROM, CDR+W, Monitor, Keyboard/Mouse/Printer/Wireless Card/Network Card/Licensed Copy of Windows. I'm trying to get it for around £200 ($400). I have found the components required (processor/ram/card for about £100 but need to get a motherboard/box to host it all - that's where I am struggling). Thanks for the details, hadn't even considered 'power' as an issue. ny156uk 17:32, 22 April 2007 (UTC)[reply]
Out of curiousity, why do your parents want to upgrade? Unless they're trying to run Vista, or they run intensive things like games, you shouldn't need to upgrade RAM or the graphics card- HDD I can understand though. If they run XP, and it's running slow, you might have better luck backing up their media and reinstalling windows. Windows has a bad habit of getting bogged down with a lot of unneeded stuff. -- Phoeba WrightOBJECTION! 02:16, 26 April 2007 (UTC)[reply]

Computer thinks its dual core[edit]

I've got a single processor Intel P4 650 3.4GHz machine. Lately, I've noticed that it seems to think it's dual core... whilst system properties correctly identifies the processor, device manager lists the computer as "ACPI Multiprocessor PC" and under processors, lists the P4 twice. Task manager shows the two different graphs for my "two different" CPUs. The net result of this seems to be that single threaded processor intensive tasks (such as mencoder) are only ever given 50% of the available processing power. Right-clicking and choosing to "uninstall" one of the processors doesn't seem to work, so I'm hoping for some alternative suggestions... Thanks in advance UkPaolo/talk 16:44, 22 April 2007 (UTC)[reply]

It shouldn't really affect it that much, but what you have is called hyperthreading, where it has a pseudo dual core. You disable it in BIOS to enable/disable hyperthreading. --Wirbelwindヴィルヴェルヴィント (talk) 16:49, 22 April 2007 (UTC)[reply]
Though there is very little reason to disable it. SMT doesn't limit each thread to "50% of the available processing power". -- mattb 16:56, 22 April 2007 (UTC)[reply]

Drawing/sketching applications for UIQ 3[edit]

What are the drawing/sketching applications available for UIQ 3? --Masatran 19:18, 22 April 2007 (UTC)[reply]

I just found the pre-installed Notes application, but I need a drawing area with zoom, scroll, etc. And in-built rectangles, circles, etc. Like a mini-version of Inkscape, or Gimp. And preferably free software. --Masatran 00:14, 23 April 2007 (UTC)[reply]

Ubuntu 7.04[edit]

I just downloaded Ubuntu 7.04 for my Macbook Pro(wanted to try it out). I burned it to a CD and put it in the computer. I got past the Ubuntu loading screen, then said that it couldn't start my windowing system. It then gave me a terminal to type into. Parallels Desktop didn't even get this far. I'm re-downloading, but why is this happening? Old versions of Ubuntu worked fine, thanks. —The preceding unsigned comment was added by 71.195.124.101 (talk) 19:23, 22 April 2007 (UTC).[reply]

Can you post the exact error, please? -- ReyBrujo 02:10, 23 April 2007 (UTC)[reply]
You're doing it wrong. For Parallels, didn't you go
  1. New virtual Machine
  2. OS Installatoin
  3. Typical OS
  4. Linux/Other Linux
  5. <name of virtual machine>
  6. Put disc in
  7. Finish?

Everything should work fine, I have Ubuntu Linux on mine. Looks awesome too! Except it doesn't have widescreen support as far as I know. [Mαc Δαvιs] ❖ 02:03, 24 April 2007 (UTC)[reply]

It should have widescreen support, just do a "sudo dpkg-reconfigue xserver-xorg" and select your resolution. --antilivedT | C | G 05:53, 24 April 2007 (UTC)[reply]

Ubuntu/Linux Software[edit]

I am buying a new laptop soon, and think that I'll get myself a non-Mac notebook and then set up some sort of Linux/Ubuntu/Kubuntu OS. I know most of these don't support quite the range of software as does a Mac and not nearly as much as a PC, but I've had trouble finding specifics. For instance, GIMP aside, can I buy and install Adobe Photoshop on a Ubuntu computer? Linux? What about Itunes? Can I get most non-game Windows-friendly programs to install on Ubuntu or Linux or whatnot? How about the regular Microsoft Office package? Do I have to buy a special type of MS Office? Will it not work at all? Are the alternatives really up to snuff, and will regular (ie mac or windows) using computers be able to open files I've made in, say Spreadsheet or OpenOffice? Will fairly obscure Windows-friendly programs such as Xilisoft, MacDrive, Daemontools, RosettaStone, and other work on any Linux/Unix-based OS's? And, final question, I sometimes go to download freeware and am given the option of whether to download it for Macs, Windows, or Linux. Could I not install this freeware on a Ubuntu computer? I know that's a lot of separate questions, but I think it all boils down to whether if I use Ubuntu or Linux I wont be able to use all the programs I've come to know and love. Thanks for any help, 70.108.191.59 19:53, 22 April 2007 (UTC)[reply]

  • There are no versions of Adobe Photoshop and Microsoft Office that are natively supported by Ubuntu. However, using CrossOver or WINE, those programs will work.
  • Don't know if iTunes is supported by CrossOver or Wine, but Kubuntu comes with Amarok which you can use to upload music to your iPod.
  • OpenOffice will open spreadsheets and word documents from MS Office, but some of the more complicated formatting and such might be lost. But in general, for most documents/spreadsheets, this isn't a problem.
  • You can definitely open OpenOffice docs in MS Office.
  • For the more obscure software you described, there is usually a (K)Ubuntu alternative.
  • Don't really know what you mean by that last question. If the freeware has an option to install in Linux, then you can install it in (K)Ubuntu. But there are easier ways to install software on (K)Ubuntu that browsing for it on the web: Package management system.
Hope this all helps. - Akamad 20:36, 22 April 2007 (UTC)[reply]
I expect someone will beat me to it (and I was right), but here goes. First, I should explain what the difference between Linux and Ubuntu is. In the early nineties, a guy named Linus Torvalds started work on an operating system. This was just the kernel of the operating system, which means it handles all the basics of various components of your computer talking to each other, starting the computer up and so on, but on its own doesn't make a computer usable yet. It's basically the software that's needed to start writing specific programs (like a window manager that draws windows on the screen, or an office application that uses the window manager). What's more, Torvalds made his software open source, so that anybody could look at, and use the source code behind his kernel. He named his kernel Linux. The Linux kernel was quickly combined with other open source software (mostly from the GNU project) to form an entire operating system, the set of software that is needed for a user to be able operate a computer. This operating system is known as Linux too, or as GNU/Linux. Since this GNU/Linux is basically a bag of different software, which you can install in many different ways, and which you can distribute freely, people started installing GNU/Linux in their own way, and then distributing it with things like friendly installation procedures and a set of preconfigured software packages. These are the different distributions of GNU/Linux, and Ubuntu is one of them. In short, Ubuntu is Linux, but Linux is not necessarily Ubuntu.
As for the available software, the linux experience is slightly different than that of windows. The open source ideology has a very active community of programmers behind it, and most types computer programs have some open source variant. For instance, one open source office suite is OpenOffice (which can indeed open MS Office files without a problem in most cases). The great joy of distributions like Ubuntu is that they make use of the fact that open source software allows you to redistribute it. This means that you don't have to go to the OpenOffice site to download the program and install it yourself, because the ubuntu guys have done that for you already. They've taken the OpenOffice software, and adapted it slightly to make sure it runs smoothly on any Ubuntu system. All you need to do is tell a program within ubuntu (called synaptic, or simply "add/remove software") that you want openoffice, and it downloads and installs it for you. Pretty much all open source software written for linux can be installed this way.
Not all software (even open source) is written for linux, and you can't install software written for one platform, on some other platform (with the notable exception of cross platform software, like Firefox, but even that has different versions for different operating systems). Photoshop has versions for Windows and for the Mac, but not for Linux. The same goes for MS Office. There is the possibility to install Wine, which allows you to run windows software in linux, but this doesn't work perfectly. For instance, this allows you to run photoshop 5.5, but not CS2. The wine project has a big database on what will work, and what won't. So no, you won't be able to use all the programs you've come to know and love (some, perhaps), but you might grow to know and love a lot of new ones. risk 20:45, 22 April 2007 (UTC)[reply]

Clock speed and FLOPS[edit]

I was about to delete a sentence of the FLOPS page , but, for fear of making an idiot (as I'm not too familar with hardware) of myself, I thought I'd ask here first.

According to that page, the CPU of a "common AMD A64 or Intel Pentium 4 general-purpose PC" (I'm guessing aroung 3Ghz), would have a FLOPS rating a ten. Which in turn leads to the conclusion that these processors perform over three calculations per clock cycle. This seems quite absurd to me, how is it possible for a CPU to perform three calculations in one cycle? —The preceding unsigned comment was added by 217.42.242.45 (talk) 20:04, 22 April 2007 (UTC).[reply]

Ten GFLOPS, you mean. The answer lies with SIMD. The Cell microprocessor's vector units (SPUs in IBM-lingo) are designed to perform eight single-precision floating point operations per cycle. -- mattb 20:45, 22 April 2007 (UTC)[reply]
Many of today's CPU cores execute single execution path instructions in parallel. Even 20+ years ago, instructions were executed in "parallel" so to speak as the execution time for each instruction was significantly overlapped for fetching the instruction, processing, storing results, etc. In another instance, (my memory may fail me here, so take this with a "huge" grain of salt, but) I've heard of instances where the first few instructions of all three "if","then","else" paths of an if/then/else block are executed and when the results if the "if" operation actually completes, the results of the correct "then","else" path are saved and used for further instruction processing. I can't remember what that's called. Or maybe that's just something I imagined. Root4(one) 01:43, 24 April 2007 (UTC)[reply]

Nintendo Wi-Fi[edit]

I am trying to get my WiFi started for the DS and I need to access my routers settings but everytime I put in my routers IP address, the router settings page never shows!!! Please can someone help me out thank you alot

You are putting the IP address in to the browser of the computer with which you are accessing the internet and it isn't coming up? Are you sure you have the right IP? Do a tracert to make sure you're putting in the right address. Vespine 23:03, 22 April 2007 (UTC)[reply]
Use ipconfig to determine the router's IP address. Splintercellguy 00:09, 23 April 2007 (UTC)[reply]

The Wall Street Journal mug shots[edit]

How did they create these engravings? How did they do it before computer graphics? -- Toytoy 21:05, 22 April 2007 (UTC)[reply]

They arent engravings, they're the normal dot-printing style of newspapers, just the resolution has been drastically reduced. The same general concept was used to make illustrations in books in the 1800's. -Mask? 21:43, 22 April 2007 (UTC)[reply]
There is an article on the WSJ's version of this at Hedcut. -Mask? 23:22, 22 April 2007 (UTC)[reply]
And we note from that article that these pictures are not "the normal dot-printing style of newspapers"; they're hand-stippled illustrations which mimic old-fashioned engravings. —Steve Summit (talk) 01:00, 23 April 2007 (UTC)[reply]
I meant the style is fashioned on them. I'll be clearer. -Mask? 02:22, 23 April 2007 (UTC)[reply]