Wikipedia:Reference desk/Archives/Computing/2008 December 13

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 13

Yahoo! Mail Chat

Hey y'all. I have a Verizon Yahoo! account(ISP) and I sometimes use Yahoo! Mail. I have never used the classic interface sticking to the new one instead (Which is still boring). I sometimes also use the chat feature. (Not very well integrated, boring interface, etc, I prefer Gmail, :P) but I recently set some Custom Status Messages. How in the hell do you clear them? I only have three at the moment, (Lunch, Dinner, Breakfast[Out of order as you can see, but that's the way it happens, except for breakfast, I just added it for later on]) In Gmail I can clear my statuses in two clicks. But there doesn't appear to be any way clear your status messages in Yahoo! Mail chat???--Xp54321 ^{(Hello! • Contribs)} 03:40, 13 December 2008 (UTC)

Nvidia GeForce 9800 vs ATI Radeon HD 4830 graphics processors

On my laptop (Vista, 4 gigs ram), I'm using a GeForce 9300 and want to upgrade it. I use this laptop as a gaming pc. I've read [this review]and the 9800 slightly better, but I'm not totally convinced that the nvidia is the better deal. The over I'm willing to stray from the nvidia line. So should I give in to nvidia and buy the 9800? or buy the radeon 4830? Parker2334 04:41, 13 December 2008 (UTC)

Uh while I can't answer your question I would strongly suggest you look at more reviews (especially reviews with different sets of games and configurations of games and different hardware configs (that uses a quad core which I doubt your laptop has). For example in that review the Crysis Warhead FPS is a little on the low side on all cards and would probably be too low for many. I'm guessing that the game is playable you just need to change the config and testing that would probably be more meaningful. Also it's quite common you'd find different games give different results. (Actually even that review shows that although the 4830 never really 'wins') Also the difference shown by the review isn't great so if there is a substantial price difference I would strongly recommend the cheaper one. However a big question might be do you know what you're getting into trying to upgrade the laptop video card? It's not unusual an easy process and sometimes may not be possible. (If your getting someone else to upgrade it then this consideration doesn't come into it). P.S. Remember to make sure the laptop versions of the card are the same configs (mem, GPU and shader frequency, memory config [256 bit or whatever], and number of shader) as the desktop versions otherwise the reviews are meaningless Nil Einne (talk) 08:40, 13 December 2008 (UTC)

Among other reasons for choosing nVidia, is the higher percentage of games that are designed specifically for an nVidia chipset. --EvilEdDead (talk) 15:20, 14 December 2008 (UTC)

No games are specifically designed for Nvidia cards. Both use standards such as DirectX, and OpenGL, and should look practically identical at the same settings. While it was the case in the early days of 3d cards games were designed with a specific brand in mind (3dfx mostly) this no longer happens. Nvidia pays games companies quite a lot to put the Nvidia logo on their games though. However if these two cards are the same price the 9800 is probably a bit better, but if you can stretch your budget a tiny bit the ATI HD4850 would be quite a bit better. However if you plan on playing a lot of OpenGL games Nvidia drivers are faster at that. If you don't know if your game uses OpenGL then chances are it doesn't. edit: Um I just realised you are using a laptop. Are you totally certain you can upgrade your graphics card. I am aware of a couple of models that can use special graphics cards, but you won't be able to use standard onesGunrun (talk) 10:10, 15 December 2008 (UTC)

Which is the best operating system?

Hello, I'm looking into buying a new laptop and want to seek your unbiased recommendation (no fan allegiances, and thanks for recommending Firefox!). Disregarding the cost of the operating system, which would you recommend based on reliability, user experience, and overall performance? I've used Windows for my entire life and know that I do not like it, because of many reasons, and would prefer not to get another Windows. But would really like my new computer to be fairly compatible with Windows and .exe files even if it means possibly getting another Windows. I've used a Mac for half a year at my old work and really like it and all the thousands of features it had except that wasn't compatible with my home computers even though the Parallels application was good, I hated switching OS's. I haven't tried Linux or any other operating systems so I can't say much about them. Can they come with as many features as a Mac and be compatible with windows applications? Which OS would you recommend disregarding cost? -- penubag  (talk) 07:27, 13 December 2008 (UTC)

It really depends on what you're looking for. Look at this article for a basic chart of information for quite a few operating systems. If you like games ... go with Windows. If you want it for anything else (music, video, or just the basics), I'd personally go with Mac. Linux Ubuntu is nice, if you know what you're doing. –  LATICS  talk  07:39, 13 December 2008 (UTC)

What else is Windows better than a mac for besides games? Can the newest Mac read .exe files? -- penubag  (talk) 08:42, 13 December 2008 (UTC)

No, macs can't directly use .exe-files, because those are windows-programs, and mac isn't windows. However, they do run fine under an application like Parallels, which virtualizes windows (think of it as sort of a simulator, the program thinks it's really running under windows, even though it's on a mac. It's like running a computer inside a computer). Linux can run a lot of windows-programs using something called Wine (it will, for instance, run Word 2003 fine). However, even though different Linux-distributions have made great strides recently to become more user (and newbie) friendly, they are significanly more challenging to use if you don't know what you're doing. There is a fairly significant learning curve.

I do think you are thinking about your choice in a slightly wrong way. You shouldn't ask yourself "what operating systems can run .exes", because Windows is the only one that's built to use them. What you should be asking instead is "Do other operating systems have good substitutes for the things I need in Windows?". And the answer for Mac is essentially yes. Besides games (which is completely dominant on Windows), you can pretty much do anything on a Mac (or Linux) desktop that you can do on a Windows desktop. You can't run Office for Windows in a Mac, but you can run a version specifically made for Macs. Not to mention the excellent OpenOffice which is pretty much just as good as Microsoft Office, but with the added benefit of being free (it is also interoperable, you can both save and open a file in Office's .doc format).

Is there any specific app you just can't live without on Windows? Do you have a list of apps that you would like to know if there are replacements for on windows? I'm sure we can help you out. And in the (very unlikely) event that there is an app for which there is absolutely no good replacement on an Apple machine, you can always run Parallels, and it will take care of that as well. I strongly recommend you go with a Mac. 83.250.202.208 (talk) 14:43, 13 December 2008 (UTC)

Some programs that I'd like to run on a Mac are just little tiny programs, stuff like these and such, I know Parallels pretty much nulls that problem, even though it still is a pain to have to run that thing. Another issue I have with macs is that I have 2 desktop Microsofts in my house all connected to each other via wireless network. I fear that getting a mac would mean that I wouldn't be able to wirelessly transfer files between the three. Is there an application or can a mac transfer with Windows?-- penubag  (talk) 08:50, 15 December 2008 (UTC)

As to games you may find you are better off with a games machine linked to your telly for games and just not bother with that requirement for your computing needs, i.e. have two machines. Dmcq (talk) 17:26, 13 December 2008 (UTC)

I would strongly advocate for Linux - but that's not the best way to answer this question because plenty of other people will advocate something else. So I'm going to say this: Why decide now? The good thing about choosing Linux as your alternative to Windows is that you can set up the PC to 'dual boot'. In this way you get a little menu pop up when you reboot your computer that asks you whether you want to run Linux or Windows - you could go nuts and install BSD Unix too - have a tri-boot machine! The only downside to doing that is the disk space it consumes - but you could install Linux (or whatever) on an external USB drive. This is a really low risk approach. When you wake up one morning and realise that you haven't booted into Windows for a month or two - simply wipe out that disk partition and reformat for Linux...if you find that Linux is too much of a pain to deal with - then erase that partition and hand it over to Windows. SteveBaker (talk) 05:48, 14 December 2008 (UTC)

I am intrested in Linux but hardly any programs are written for it. Is wine realiable enough where I shouldn't care? I don't think having a dual boot will be good on my laptop as I'm not getting a really expesive one with a lot of disk space. Can you tell me another reason as to why I should go linux?-- penubag  (talk) 08:50, 15 December 2008 (UTC)

As I stated earlier, though ... Linux is only good if you know what you're doing. For me, it wouldn't be ideal because of compatibility. But if someone's using just for internet (that doesn't know a ton about it), Linux would be fine. For me ... with Photoshop, Zune, several games, etc... Windows is really the only option. –  LATICS  talk  06:21, 14 December 2008 (UTC)

To answer the more hardware-oriented question: I think that the ThinkPads are awesome computers. They're not manufactured by IBM anymore, but they still do a lot of the background support for them. I've got a T61p which performs beautifully, and Lenovo tends to have good sales (with plenty of ways to find extra % off through coupon codes or contractor logins). Washii (talk) 07:25, 14 December 2008 (UTC)

PC World ran a review several months ago and said basically the best PC laptop was a MacBook Pro. You can use it exclusively in PC mode and install Windows, Linux or whatever. So you get one box that can run all three major OSs. No other laptop can do that. Apple is also well known for their high build quality. Nearly the entire shell is milled out of metal, which is far more durable than plastic laptops. --71.158.216.23 (talk) 05:41, 15 December 2008 (UTC)

Thanks for that info. -- penubag  (talk) 08:50, 15 December 2008 (UTC)

Thanks a lot everyone for the replies. I have responded above. -- penubag  (talk) 08:46, 15 December 2008 (UTC)

Another response, Ubuntu Linux (and linux itself for that matter) is pretty awful and overcomplicates a lot of things just for the sake of it. It is good, don't get me wrong, but BSD bases are much better. — neuro^(talk) 02:14, 16 December 2008 (UTC)

Command Prompt unable to delete folder

Everytime I try to open a certain folder in my F:\ drive, Windows Explorer displays the following message:

Windows Explorer has encountered a problem and needs to close. We are sorry for the inconvenience.

I tried deleting the folder (it contains subfolders as well) using Command Prompt using rd, rmdir and del, but to no avail. Each time I hit Enter, Command Prompt simply closes down. I can't delete the folder using GUI either. Please help. (Windows XP, SP2) 117.194.226.34 (talk) 07:38, 13 December 2008 (UTC)

You could try copying everything else off to another drive and then reformat F: and restore. Think about this carefully before you try it. -- SGBailey (talk) 09:19, 13 December 2008 (UTC)

Sounds like filesystem corruption or something. I'd try a chkdsk and some similar kind of tools. Indeterminate (talk) 09:35, 13 December 2008 (UTC)

Filesystem corruption wouldn't cause a user-mode program to crash. It would result in a BSOD. I would say the folder contains files/directories with weird filenames, or that there is some sort of user-mode hook on the API functions used to open directories. Try using a Windows "Live CD" like BartPE or a GNU/Linux Live CD with NTFS-3G. --wj32 ^t/c 10:50, 13 December 2008 (UTC)

Could you browse to the directory in a prompt and type DIR, paste the outcome at http://pastebin.com/, and link to it here? — neuro^(talk) 02:12, 16 December 2008 (UTC)

• Your operation system DOES have computer viruses in it. Save any valuable information in a safe place and clean your computer ASAP! 194.99.216.134 (talk) 12:13, 20 December 2008 (UTC)

Question

Where can I find computers that run old operating systems (such as Windows 3.1)? I'd like to have my own personal computer for old stuff. Yes, I know I can use Ebay, but I'd also like to know other places. (Country: Australia.) 58.165.14.208 (talk) 08:15, 13 December 2008 (UTC)

You should be able to run Windows 3.1 in VMware or other virtualisation program. The only potential problem I can see is midi support. You can also run Windows 3.1 on DOSbox [1]. The hardest thing may be finding a legal copy of Windows 3.1 Nil Einne (talk) 09:36, 13 December 2008 (UTC)

That's exactly what I'm trying to ask - sort of. I would like to know where, not counting Ebay, I can buy computers that run Windows 3.1. Though I have only used Windows 3.1 about 10 times, it's one of my favourite Windows operating systems - and it's only fair that I can find a computer that has Win3.1 on it. 58.165.14.208 (talk) 10:30, 13 December 2008 (UTC)

If you actually want a physical computer to run Windows 3.1 on then your current computer should do, just install Windows 3.1 on a separate hard drive and format it as FAT, or if you're willing to do a bit of work then you could even partition your current hard drive. It shouldn't be a problem if your computer uses Serial ATA as most BIOS can emulate for older hardware. If you want the full nostalgic experience complete with noisy fans, high power consumption and unbelievably slow loading times then I'd imagine you can pick up old computers at second hand shops, garage sales, flea markets etc or ask around your friends as many old pcs still lurk in their owner's attics and basements collecting dust. My advice though would be to use Qemu Manager to make virtual computers. You can set it up to emulate older hardware and Windows 3.1 works well on this. SN0WKITT3N 15:07, 13 December 2008 (UTC)

Sorry but was with SNOWKITT3N I'm not sure if you're quite getting me or perhaps I'm not understanding you. I don't see why you need a computer with Windows 3.1 when you can just run it in a virtualisation program or emulator (you can also probably run it in your computer natively but as mentioned that might be more problematic for a variety of reason). Sure acquiring a computer may be one way to get a license, but it's unlikely to be the only way (and if it's an OEM license you may or may not depending on where you live and the license agreement be able to install it in another computer anyway). In any case, it's probably going to be difficult to verify the license is even legit since there's a good chance whatever material that came with the computer to prove the license is legit is lost Nil Einne (talk) 07:35, 14 December 2008 (UTC)

I think we need to back up and explain that Windows 3.1 will run on either the old computers it originally was designed for or on newer computers. It will, however, be much faster on newer computers. In some cases, it can actually be too fast, if the programmer depended on the slow speed to give you time to view the display before it updated. One example of this is the information that scrolls by when booting a computer. It was originally displayed slowly enough that you could read it, but even a speed reader would be out of luck now, with today's faster computers.

So, if you want an old computer, then you may be able to get it with Windows 3.1 still on it. If you want a new computer, then you only need to find a copy of Windows 3.1 to install on it (I suppose there might be a few new computers that already have Windows 3.1 on them, but good luck finding them). What don't you like about buying things via eBay or other web sites ? If you don't like buying things over the Internet, then you're limited to computers near your home, and there aren't likely to be old computers and operating systems for sale in your town. If there are, they would be at garage sales, I suspect. StuRat (talk) 18:06, 13 December 2008 (UTC)

On a related note: for newer computers, are everything still compatible with those old DOS / Win3.1 drivers? Are sound cards, for example, still Sound Blaster compatible, and would one be able to get decent screen resolutions? It is my impression that the relatively decent plug-and-play capabilities of Windows XP has made people forget about the hassle this used to be. And the original poster should be aware that (as far as I know) networking on Windows 3.1 might not be easy; you would need some additional programs, right? Jørgen (talk) 18:46, 13 December 2008 (UTC)

Good point. The last time I tried Windows 3.1 on a new computer I could only get sound to come from the motherboard's internal speaker and a screen resolution of 640x480. Also whenever sound played the system would lock up until the sound had finished. That's why I think its best to install 3.1 on a virtual pc as it will emulate older hardware which the drivers will work with. If you do install on a newer computer check out this site which has a collection of drivers for newer hardware, even DVD and USB drivers for Windows 3.1 and DOS. SN0WKITT3N 19:34, 13 December 2008 (UTC)

• You can obtain old PC here 194.99.216.134 (talk) 12:17, 20 December 2008 (UTC)

Blue screen of death error report

Can anyone decipher for me the error report I got after I restarted from a blue screen of death. The error report said I had had a "serious error" and when I asked for detail on the report it was going to send to Microsoft (which I didn't bother doing), it gave me:

C:\DOCUME~1\(redacted)~1\LOCALS~1\Temp\WERccdc.dir00\Mini121308-01.dmp
C:\DOCUME~1\(redacted)~1\LOCALS~1\Temp\WERccdc.dir00\sysdata.xml

Thanks.--71.247.123.9 (talk) 11:33, 13 December 2008 (UTC)

I don't know if anyone here can decipher the report, but I'm sure they'd have to see it to even try. You need to upload those files so we can see them. StuRat (talk) 17:50, 13 December 2008 (UTC)

By chance, do you happen to have Kaspersky or Zonealarm installed and what are your systems specs? Zonealarm and Kaspersky cause errors like that. Also, outdated drivers could cause this problem. But without those two files. Knowing exactly what is wrong is not truly possible. It will all be guess work. Rgoodermote  08:01, 14 December 2008 (UTC)

Upload the files, I have experience working with minidumps. — neuro^(talk) 02:10, 16 December 2008 (UTC)

...as does anyone who's changed a baby's diaper. StuRat (talk) 14:44, 17 December 2008 (UTC)

(outdent)LMFAO ... took a second on the diaper comment ... loved it. Anyway ... what they are saying is you need the contents on those files. The C:\DOCUME~1 .... ya da ya da refers to two files that can now be found on your hard drive. Using Explorer (not INTERNET Explorer) go to your C drive, then you will see a folder called Documents and Settings. (Now may need to change your folder options to view hidden files and folders) ... in there you'll find your user name, inside that Local Settings, and inside that a folder called Temp. that's where you'll find those 2 files. you can read the xml one by double-clicking on it (Internet Explorer) ... and if you right click the .dmp one and pick open with ... choose notepad. viewing these files will tell you error codes (usually in Hex) ... it'll probably tell you a filename (explorer.exe) caused a fault: 00xF00 something. Doing a google search will often tell what type of error was caused, and sometimes offer a solution. If you're still not sure ... contact Neurolysis (hey ... you offered) and he'll try to help. Ched Davis (talk) 04:10, 19 December 2008 (UTC)

You may want to check out this article on debugging STOP (blue screen) Errors With WinDBG. The results of that analysis (what caused the fault) should help point you in the right direction. --Martinship (talk) 06:04, 20 December 2008 (UTC)

My problem with a C program

As a part of a college project, I have been asked to write a C program which accepts an employee's username and password. The difficult part is that the output screen is not supposed to show the password while the user types it in the console. I have a limited knowledge about C (since my major in college is electrical engineering) and do not know how to proceed with this problem. Please explain what to do. I will be very thankful for your help. —Preceding unsigned comment added by 220.227.68.5 (talk) 12:40, 13 December 2008 (UTC)

You don't say what platform your program will run on, and that makes a difference. On a unix platform, you'd call tcsetattr to temporarily unset the terminal's ECHO bit, read the data with gets or whatever, and then restore the terminal with another call to tcsetattr. Almost all modern Unix implementations, including OS-X and Linux, implement the getpass() call, which does this for you. I don't know the specific call for C on windows, but I'm sure it's much the same. 87.114.128.88 (talk) 14:41, 13 December 2008 (UTC)

#include <termios.h>

int get_a_char_from_kbd_without_showing_what_it_is(void)

{ /* blocking function to wait for a keystroke, then get it without also echoing it */

struct termios stored;

struct termios newios;

int fd, ch;

 

fd = fileno(stdin);

if (!isatty(fd)) {

ch = fgetc( stdin );

if (ch == EOF)

ch = 0;

return ch;

}

fflush(stdout);

tcgetattr(fd,&stored); /* Get the original termios configuration */

memcpy(&newios,&stored,sizeof(struct termios));

newios.c_lflag &= ~(ECHO|ECHONL);

#if defined(ECHOPRT) /* not POSIX */

newios.c_lflag &= ~(ECHOPRT);

#endif

#if defined(ECHOCTL) /* not POSIX */

newios.c_lflag &= ~(ECHOCTL);

#endif

newios.c_lflag &= ~(ICANON); /* not linemode and no translation */

newios.c_cc[VTIME] = 0; /* tsecs inter-char gap */

newios.c_cc[VMIN] = 1; /* number of chars to block for */

tcsetattr(0,TCSANOW,&newios); /* Activate the new settings */

ch = getchar(); /* Read a single character */

tcsetattr(fd,TCSAFLUSH,&stored); /* Restore the original settings */

if (ch == EOF)

ch = 0;

return ch;

}

Enjoy. -- Fullstop (talk) 16:11, 13 December 2008 (UTC)

As an alternative, couldn't you use the getch() function from the standard C library - I don't believe getch() echoes the key pressed. It also has the advantage of being portable across many platforms. However, getch() only returns a single character, so you will need a loop to construct the entered password string from the individual characters entered until it gets a <CR>, <LF> or <ctrl-D> or some other end character. Astronaut (talk) 22:19, 13 December 2008 (UTC)

getch() doesn't "echo" anything. Input functions don't echo anything either. It's the terminal that displays the characters the user types, not the input functions. tcsetattr with the right arguments tells the terminal not to display the characters. --wj32 ^t/c 00:53, 14 December 2008 (UTC)

Yeah - that's really the heart of the problem. In a Unix system, the device driver is the beast that's responsible for echoing typed input back to the user - so there is nothing your program can do directly to change whether there is echo or not. Instead, you have to send a command to the device driver to tell it to turn off the echo (and to turn it back on again). Doing this correctly is a total pain in the ass. For example - you have to consider what happens if someone types a ^C while the echo is turned off...your program exits - but since nobody told the device driver to turn the echo back on again, you suddenly find that you don't seem to be able to type anything (although, in truth, you are typing - you just don't see the echo). So you should really be prepared to catch the various signals that could cause your program to exit and make sure you turn the echo back on again before the program exits. To be UTTERLY correct, you should read the state of the terminal settings when your program starts - then restore that setting when you exit.

This may seem at first sight to be an ugly way to proceed - but there are (as usual) very solid reasons why Unix works this way. Recall that you can take a program that reads from the keyboard and writes to the screen and redirect it's input to come from a file on disk. It's fundamental to Unix command-line usage that this be allowed by EVERY program...so you don't want the program to be echoing input that it thinks is typed in by the users because then every character that it read from the file would appear on the screen. Hence, the need to echo or not echo is an attribute of the DEVICE not of the PROGRAM. More importantly in the early days of Unix was that there were dumb terminals where the echo was done in the terminal itself - hence the computer would have to know not to echo things that were typed or else your keyboard input wwoouulldd ccoommee oouutt ttwwiiccee!! In those cases, the device driver would have to send commands to the terminal to tell it to turn echo on or off - and that absolutely, utterly, had to be something that the device driver did.

SteveBaker (talk) 05:35, 14 December 2008 (UTC)

How can I implement this feature in a database

How could I best implement the following in a (relational) database:
To keep things simple let's suppose I have table "Cars" which keeps records of information relating to various cars. It may have fields such as "Engine" "Drive" etc

Now say there's a field for the "Colour" of the car:
I want to be able to select a colour from a list in a manner which I could best be described by showing you the list below:
Red

Light red

Dark red

...

Blue

Light blue

Dark blue

Green

...

Yellow

...

...

From a list such as the one above, I want to be able to select "Red" or one of the types of red. It wouldn't necessarily be limited to 2 types of each main colour and there may be more levels (i.e. there may be different types of "Dark red")

Finally, I want to be able to search cars by colour so that if I search for "Red", the result will be any car which is any type of red. If I search for "dark red", the result will be any car which is any type of dark red etc

--212.120.246.219 (talk) 12:42, 13 December 2008 (UTC)

That idea, where "light red" is-a-kind-of "red" is one that natively sits well with an object database; if you wish to implement this in a relational database (where such familial relationships aren't captured" you could either:

1. have an additional field in your car table, say called "colour_family", where a maroon coloured car could have a colour of maroon and a colour_family of red
2. if doing stuff by colour was important (that you'd be doing a lot of queries and changes specifically in terms of colour, then you might chose to have a seperate colour index where you'd store cars in terms of their colours; you can have any number of entries for this, so a given car can be (if you choose) any number of colours. Obviously this is considerably more complex, and you need to keep everything up to date when making changes.

87.114.128.88 (talk) 14:35, 13 December 2008 (UTC)

---

You're basically trying to have a hierarchical structure in a relational database, which is a common problem. I'd suggest a structure like this for the color table:

    ID       PARENT     NAME      
+---------+-----------+-------------------+
| 100,000 |  0        | Red               |
| 110,000 |  100,000  | Light red         |
| 120,000 |  100,000  | Dark red          |
| 121,000 |  120,000  | Dark red metallic |
| 200,000 |  0        | Blue              |
| 210,000 |  200,000  | Light Blue        |
| 300,000 |  0        | Green             |
| 400,000 |  0        | Yellow            |
| 500,000 |  0        | White             |
+---------+-----------+-------------------+

There would then be some code to select a car color from this COLOR table which would look something like this:

ID = 0
LOOP:
 select COUNT from COLOR where PARENT = ID; ! Get number of sub-colors for this color 
                                            !  (or number of top level colors if ID=0).
 if COUNT = 0 goto END;                     ! Quit program if there are no more color levels.
 select *     from COLOR where PARENT = ID; ! Get sub-colors for this color 
                                            !  (or top level colors if ID=0).
 ...                                        ! Display list and let user select color (not shown).
 CONTINUE_SEARCH ?                          ! Once color is selected, user should be prompted 
                                            !  to either use that color or search for subcolors.                                
if (CONTINUE_SEARCH=YES) goto LOOP

Obviously there's quite a bit more needed to the code, like assigning local program variables, but that's the general idea.

Now, once a final color is selected, that color's ID would be added to the CAR table (as COLOR_ID, for example). To search for cars of a given color, you must first select a color from the color table. In this case, all colors should be listed. Note that you could make the list look more presentable, when displayed, by using some extra spaces in the NAMEs in the COLOR table:

    ID       PARENT     NAME      
+---------+-----------+---------------------+
| 100,000 |  0        | Red                 |
| 110,000 |  100,000  |  Light red          |
| 120,000 |  100,000  |  Dark red           |
| 121,000 |  120,000  |   Dark red metallic |
| 200,000 |  0        | Blue                |
| 210,000 |  200,000  |  Light Blue         |
| 300,000 |  0        | Green               |
| 400,000 |  0        | Yellow              |
| 500,000 |  0        | White               |
+---------+-----------+---------------------+

In this case a simple select would work:

select * from COLOR sort by ID;

You could also skip adding the extra spaces in the table and write some extra code to display the list that way, instead. Now, once you had the color selected, you'd need to prompt the user to either search for cars with that exact color or else that color and all subcolors of it. If they say they want that exact color, just do this:

select * from CARS where COLOR_ID = ID;

If they also want subcolors, here's where you could use the naming convention I've used for the ID:

select * from CARS where COLOR_ID like '1%%%%%%'; ! All cars any color under red.
select * from CARS where COLOR_ID like '11%%%%%'; ! All cars any color under light red.
select * from CARS where COLOR_ID like '12%%%%%'; ! All cars any color under dark red.  
select * from CARS where COLOR_ID like '121%%%%'; ! All cars any color under dark red metallic.

Note that this assumes that the ID and PARENT fields in the COLOR table and the COLOR_ID field in the CAR table are character strings. If they're numeric, some conversion between character and numeric fields would be needed in the program. StuRat (talk) 14:54, 13 December 2008 (UTC)

One other comment on the naming convention I've used for the COLOR table ID field: I've used numbers in the character strings in my example, but you aren't limited to just numbers. If you have more than 10 top-level colors, you could use uppercase letters for 26 possibilities, or uppercase and lowercase for 52, or add in numbers for 62. The same would be true at each level of subcolors. You could even use all ASCII codes for 128 (7-bit) or 256 (8-bit) possible colors at each level. However, beware that some of those characters are unprintable, so require more work when debugging (you have to print out the ASCII codes instead of the characters).

One other comment on the sort order of the COLOR table: I've sorted based on the ID, which works well for a static table. However, if you need to change the colors frequently, you can't change their IDs easily, as those are in the CAR table too. So, you might want to have an additional SORT_ORDER field in the COLOR table which controls how colors are displayed for selection. StuRat (talk) 17:25, 13 December 2008 (UTC)

---

However, if your set of colours and their relationships is fixed when you write your program, you can just do the work in your SQL query:

SELECT * FROM cars 
  WHERE colour="british racing green" 
     OR colour="green" 
     OR colour="moss"  ;

That's the very opposite of StuRat's suggestion; which is appropriate depends on your problem. 87.114.128.88 (talk) 15:36, 13 December 2008 (UTC)

www.a.com

Apparently, single letter domain names (e.g. www.a.com) are not allowed. Why ever not? I wanted need one. --Seans Potato Business 17:46, 13 December 2008 (UTC)

We have an article at single-letter second-level domains. According to [2], in the early days of DNS they considered splitting each TLD into 26 subdomains in order to split the load, so the English Wikipedia, for example, would be at en.wikipedia.w.org. They decided against it (obviously) and considered releasing them in 2005, but apparently nothing came of it. If they did release them, I imagine you'd have a hard time getting one - there'd only be 26 per TLD, and I could see some companies paying substantial sums for them (Google would probably want to snap up g.com, for example). There are some, actually - q.com and x.org are two. They were registered before it was blocked. — Matt Eason ^{(Talk &#149; Contribs)} 21:01, 13 December 2008 (UTC)

Google sites

Hello,

I recently wanted a [email protected] email for a website i'm in the process of building, so I went to 'gmail for organisations' and got one.. The bad news is that, becuase i bought the domain name through google, I can't see a way of uploading the website i've done in dreamweaver -i can only use their stupid templates. Does any one know how to get around this? I feel that as i've bought the site, its not for google to decide how it should look....82.22.4.63 (talk) 19:13, 13 December 2008 (UTC)

Sorry guys, I've just found out that the answer seems to NO. Is there anyway, i can therefore take the domain name off google and host it elsewhere? thanks..82.22.4.63 (talk) 19:55, 13 December 2008 (UTC)

Oh god, I've answered this one too! am hideous, ignore me..82.22.4.63 (talk) 20:00, 13 December 2008 (UTC)

Homer Simpson's big blue pants

I've noticed that cartoons are still using the old cheat of having each character wear the same clothes all the time, so scenes can be used interchangeably, without concern for clothing. Couldn't the coloring be done by computer, with each item, say "Homer's pants", assigned an index number, so that they could be filled in with a different color just by changing the mapping in a look-up table ?:

Index  Description     RGB Color  
-----  -------------   ---------
  1    Homer's pants   00 00 FF 
  2    Homer's shirt   FF FF FF

I realize that some cartoonists may always choose the same outfits as an "artistic choice", but this would give them the option to change the color of a clothing item in every panel, when they render the images, if they wanted. Also, couldn't a variable texture be assigned in the same way ? You might end up with something like this in a large table:

Frame#  XY_Location  Item Index
------  -----------  ----------
200000   100,100       1           
200000   100,200       2
 

StuRat (talk) 20:09, 13 December 2008 (UTC)

They could, but characters wearing the same clothes and having the same hairstyle isn't merely a matter of not having to recolour or redraw character frames (indeed, colouring and tweening are, for mass-production cartoons like The Simpsons, mostly done using automated or mostly-automated systems). Characters in comics are typically drawn looking much the same, even when the whole strip is hand-drawn (and thus there's no copy-paste economy to be had). Homer actually wears all kinds of things as the situation demands (a suit to dinner and to church, a muʻumuʻu when he's super-obese, a spacesuit when he's an astronaut. The normal Homer outfit merely means "the clothes don't matter, don't pay attention to them"; a different outfit means "something special is happening; the clothes are the clue as to what". And given the heavily-stylised art style of The Simpsons, they need to do that to properly communicate - if Homer wore a cheap suit to work and a fancy suit for a special occasion, they really couldn't easily draw the difference. All the Simpsons characters are the same: Apu wears the same outfit all the time, but when they want to show he's out on a date (shirt with ruffled collar) or getting married (white outfit with sash) they change it. 87.114.128.88 (talk) 21:47, 13 December 2008 (UTC)

It appears that you are under the impression that no humans are drawing Homer's pants. Why? The first 13 seasons used traditional cell animation. Since season 14, they use digital ink and paint - which means the humans draw it and scan it into a computer for arranging and syncing the animation. -- kainaw™ 04:04, 14 December 2008 (UTC)

Cartoons aside - the thing that our OP suggests is certainly used in computer games. In my time with Midway Games, I wrote software to take a single model of a person and switch colors textures and some interchangeable 'parts' (hats, hairdo's, cellphones, bags, etc) to create extra variety in crowd scenes. But there are a lot of traditional aspects to making cartoons - conventions that have come up and seem to work. Having four-fingered hands for example. I suspect it's just simpler to maintain continuity and to re-use animation from one part of the cartoon to another. If sections of the cartoon are produced in different locations and out of sequence - it's much easier if everyone can simply rely on the character's appearance never changing. In "The Simpsons" they frequently make jokes about their character's overly-standardized appearance - when Marge's iconic pearl necklace gets stolen - she briefly agonizes over it - then pulls open a drawer containing dozens and dozens of identical necklaces. SteveBaker (talk) 05:17, 14 December 2008 (UTC)

This is indeed mostly used in old games, the trick is called a palette swap, the most famous creation of which is of course Luigi (orignially just Mario with green pants). A character sprite takes up valuable space on those old-school cartridges (what with all the running and jumping and fifty other animated things it has to do) so it's a dead easy way of creating additional people. This is easy to do when your character is just a few pixels big, but you can't do that with Homer. He's far to hi-res and detailed, there's all sorts of shading and things you have to look out for. You can't just do a search and replace. Besides, I like his blue pants! 83.250.202.208 (talk) 15:12, 14 December 2008 (UTC)

I'd think you could handle shading. Just have the pants be shaded beforehand in the usual way, but with a shade of gray, then add the color on top of that until you hit a black border. So, for example, an RGB hex code for a pixel of gray, like 90 90 90, could be shaded blue to get 90 90 FF, or shaded purple to get C0 90 C0. At times it might be necessary to specify an interior point on two or more discontinuous regions of the pants in a single frame, such as if he has something on his lap while sitting and each leg needed to be colored in separately. StuRat (talk) 17:21, 14 December 2008 (UTC)

I don't doubt that this kind of thing already happens in digital animation, in one way or another. With things as complicated as Procedural generation of hair and cloth, I think swapping the colour of Homer's pants would be rather arbitrary wouldn't it? Vespine (talk) 00:48, 15 December 2008 (UTC)

Uncheck "Read-Only" Globally; Windows98

You remember Windows98, don't you?

I've got several DVD's, loaded with thousands of photos, in hundreds of folders.

I want to copy them to a Win98 computer, and massage the data. However, the files all have Attributes of "Read-only". Is there any way to change that attribute on all of them in one swell foop?

Right now, the only way I know of to change them is to open up Window Explorer, open up the particular sub-(sub-sub-)folder, select all the files in that particular folder, and change them. If I change that attribute on the folder itself, the change is not reflected on the individual files. I'd hate to have to do this hundreds of time, and would like do it with one command.

While we're at it, what the heck is an "archive" attribute?

Thanks. Bunthorne (talk) 23:20, 13 December 2008 (UTC)

From a dos prompt (in XP do "Start | Run | Cmd") enter (without the quotes and using the correct drive and path) "ATTRIB -R X:\this\that\theother\*.* /S" -- SGBailey (talk) 00:21, 14 December 2008 (UTC)

Fantastic! Just what I needed, and it works like a champ. Thanks a bunch Bunthorne (talk) 01:04, 14 December 2008 (UTC)

The archive attribute is designed to be turned off during a backup, and is turned on any time a file is modified. This way, an incremental backup program can tell if the file has been modified since the last backup. --Bavi H (talk) 02:38, 14 December 2008 (UTC)

Resolved

Note on any version of Windows NT since at least I believe 2k you should be able to just change the folder setting from Explorer to change all files and subfolders Nil Einne (talk) 07:21, 14 December 2008 (UTC)