Wikipedia:Reference desk/Archives/Computing/2008 July 19

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.

July 19

Drilling holes in a circuit board

This is going to sound like a stupid question, but how safe is it to drill a few holes in a circuit board? I have a wireless access point that gets very hot (as in, painful-to-the-fingers hot), so I bought a tiny little fan for it that I'm going to rig up on the inside. There are some air vents in the top and the bottom of the case, but the main circuit board seems to completely block any airflow from bottom to top, so I'm thinking of drilling a few small holes in it. It's a pretty sparse circuit board, so I've found some rather large areas with no components or connections on either side (a square centimeter here, a square centimeter there), and obviously that would be where I would want to drill. Is there a significant danger of fracturing the board or something? This is a pretty cheap WAP, so if I do break it life goes on, but I'd like to know before I begin if everyone thinks this is the worst idea ever and guaranteed to fail spectacularly. —Bkell (talk) 05:23, 19 July 2008 (UTC)

Holes won't do too much to cool it down. Find a small heatsink and glue it onto the chip with epoxy. That'll lower the temperature right away. --antilived^{T | C | G} 05:29, 19 July 2008 (UTC)

Hmm, I hadn't thought of that. Will a heatsink be enough by itself, or should I include the fan too? —Bkell (talk) 06:01, 19 July 2008 (UTC)

Well a fan without heatsink is basically useless. It's designed to stand that temperature (~70°C?) so I wouldn't worry about it that much, but a heatsink should increase the convection efficiency so much that you wouldn't need a fan to increase the airflow. --antilived^{T | C | G} 06:05, 19 July 2008 (UTC)

I wouldn't necessarily say that a fan alone is useless, if it's positioned properly. In this instance, if the circuit board completely blocks the airflow, that sounds like a situation that could be improved by a bit of ventilation -- it's probably not that it runs so hot, just that it kind of "cooks" in its own casing because the air doesn't move properly -- the idea isn't so much to cool it down as it is to move the hot air out so it doesn't get as hot in the first place. That said, though, I'm not sure that just a couple of tiny holes drilled in the circuit board (which should be safe to do, by the way, as long as you avoid breaking anything on the board itself) is enough to insure ventilation. Then again, depending on where they are and what the device itself is like, it might be! If you're willing to give it a shot, you probably don't have a whole lot to lose -- it's not as if buying replacement is going to be terribly expensive, and it sounds to me like you're resigned to doing so anyway if you can't get this thing to cool down a little. -- Captain Disdain (talk) 10:38, 19 July 2008 (UTC)

Not convinced? Do a little experiment: remove the fan on your relatively modern CPU and see how long before it freezes, then remove the heatsink but maintain the fan and see how long before it freezes. A bare chip has like 50°C/W for thermal resistance, and even if you put a fan on top it'll still be around 30°C/W, whereas a heat sink will instantly lower it to less than 10°C/W. --antilived^{T | C | G} 11:00, 19 July 2008 (UTC)

Yeah, but but we're not talking about modern CPUs, which -- unless they shut themselves down in time -- get so hot that they'll scorch themselves and the circuit board they're on pretty quickly unless equipped with a heat sink and a fan. They run much, much hotter than the itty bitty processors used in small devices (like the wireless access point in question), which don't even have any fans or heat sinks at all. I'm not saying that a heat sink isn't more efficient -- sure it is! I'm just saying that he may not need one to cool it down, any more than I needed a heat sink or a processor fan in my old 386. They generate heat, but not that much of it, and a bit of ventilation may well be enough to cool it down. -- Captain Disdain (talk) 11:23, 19 July 2008 (UTC)

It's probably safe for this device, but note that many circuit boards have traces sandwiched in multiple layers, not just on the surface. --Sean —Preceding unsigned comment added by 76.182.119.200 (talk) 11:45, 19 July 2008 (UTC)

Why not just open up some holes in the case? ---J.S (T/C/WRE) 02:23, 20 July 2008 (UTC)

Or you can do what I did, and remove the computer case entirely. This allows it to run nice and cool. (In theory, it could run cooler with the case on, if it was designed with a proper "wind tunnel", but they seem to have put no thought into this and completely blocked the air flow, just like yours.) It's a bit ugly, though, so I put a "computer cover" (really a small table cloth) on it when I put it to bed at night or when company comes over. StuRat (talk) 18:58, 20 July 2008 (UTC)

Hmmm, that gives me an idea for a new home industry - computer cosies. Matt Deres (talk) 19:03, 21 July 2008 (UTC)

Yep, my idea exactly. The table cloth is my first prototype. StuRat (talk) 22:36, 21 July 2008 (UTC)

code execution by kernel

if we provide various whitespaces in the code and its passed to the kernel will it execute the code incuding those spaces or will change it?Prachi 08 (talk) 06:23, 19 July 2008 (UTC) Ρrachi

If by "whitespaces in the code" you mean spacing out your source code to make it easier to read, most compilers will ignore the spacing and produce exactly the same object code (and ultimately the exact same executable) as the source without any additional spaces.

If instead you mean using some kind of patch program to insert space characters into an executable program, it will almost certainly break the program. Astronaut (talk) 07:47, 19 July 2008 (UTC)

Check the documentation for your compiler. Some really don't do well with white-space formatting and others dont' care. ---J.S (T/C/WRE) 02:25, 20 July 2008 (UTC)

Buying a new Computer

My Sony Computer is now 5years old and after ram upgrades is still running slow. I have looked at HP and Apple and am completely confused regarding the processors. The Apple has a Intel 2 core dual processor up to 3.06 Ghz. The HP has a Intel 2 Quad processor Q9300 2.5 GHz. Is four 2.5 GHz going to process work faster than two 3.06 GHz? —Preceding unsigned comment added by 89.241.241.73 (talk) 11:35, 19 July 2008 (UTC)

In general, yes.87.102.86.73 (talk) 12:31, 19 July 2008 (UTC)

But as the apple and HP use different operating systems (apple OS and microsoft OS) that might complicate things.87.102.86.73 (talk) 12:33, 19 July 2008 (UTC)

Is your computer "running slow" compared to other computers, or compared to past performance of the same computer? If it's the former, go for a new PC. The 2.5 GHz processor is not necessarily going to run "faster" than the 3.06 GHz. It depends on everything else, specifically the FSB, L2 cache, RAM size and speed, and hard drive speed. If it's the latter, simply reformat your hard drive and start anew. --Russoc4 (talk) 14:57, 19 July 2008 (UTC)

Could you tell us the specs of your current computer as well as what you will use the new computer, if you buy one? Russoc4 made a valid point. It might be possible to salvage your older computer. If its possible, it would be the best of both worlds — you can save some money and there is one fewer working computer in the dumpster. Kushal (talk) 15:41, 19 July 2008 (UTC)

I'd start with this and go from there. But that's only if you were planning on building it yourself. Useight (talk) 17:30, 19 July 2008 (UTC)

On the question of quad v. dual cores, it depends what you are going to do with it. Many programs cannot take advantage of quad cores effectively—the program has to be written in a way that it can deal with such things, and it's not a trivial matter to write a program that can successfully split its processing tasks into four pieces. Here's a nice blog post showing how a few sample programs deal with four cores. Some get you 2X the performance of a dual core; some give you no performance increase, and in the case of many games, they will run faster on your dual core of a higher clockrate than on the quad core of the lower clockrate. --98.217.8.46 (talk) 17:54, 19 July 2008 (UTC)

I don't do anything particularly exciting on my computer, and it gets away with two cores easily. If want to crunch numbers, you could get a Mac Pro with eight cores.78.151.50.55 (talk) 15:03, 21 July 2008 (UTC)

Processor with 4 cores is faster only if you are simultaneously running at least four CPU intensive programs, or if you are using one of those few programs that support multi core. For example, I have a dual core processor, but I don't have a single program that would take advantage of multiple cores, so a single core processor would be just as fast. Some 3D rendering programs do support multiple cores. For example Poser 7 would support up to 4 cores (but I have only Poser 6 which only supports one core). I believe there are some games that support multiple cores, but most games do not. --PauliKL (talk) 03:13, 22 July 2008 (UTC)

Firefox 3 & preview errors / edit conflicts

My FF3 appears to behave differently from FF2 in one point crucial to Wikipedia: In FF2, when I got edit conflicted, or some error occured during preview, I could go back, and the text I typed in would still be there. Now in FF3, when I get an ec or some other error and return to retrieve my posting or other edits, they are lost. The edit box I see when I go back looks like the one I originally opened, everything I typed is gone. I've tried to find something about this in Mozilla forums, but it's probably either an issue with my installation/config or something that only really matters on a wiki. It's kind of annoying, because despite having adopted the habit of copying anything before hitting preview or save, I still lose edits all the time, and it's bloody annoying. Do I have to go back to FF2 or is there some way the behaviour of (my?) FF3 could be changed? user:Everyme 13:46, 19 July 2008 (UTC)

That's odd. I recently upgraded from FF2 to FF3 (on WinXP Pro 2002 SP3, no apparently relevant add-ons) and got no change in this functionality. Algebraist 15:53, 19 July 2008 (UTC)

Oh crap. I was half-hoping it was something to do with FF3. Hm. Ok, thanks for crushing my hopes for a quick fix or at least the reassurance that it's a general issue ... :D — Guess I have to start looking deeper into those forums then. user:Everyme 17:41, 19 July 2008 (UTC)

Second-life rip off PS home

Questions:

1. In PShome will combat be allowed either melee eg swords,cudgels etc or ranged eg archery,pistols etc.

2. Also will it be possible to steal from other people houses?

3. Is it possible to play in 1st person perspective?

87.102.86.73 (talk) 14:08, 19 July 2008 (UTC)

Why does your choice of title make me think that this is not an entirely serious enquiry? - X201 (talk) 14:16, 19 July 2008 (UTC)

I believe the "rip-off" refers to the second question, stealing from other's homes within the game. StuRat (talk) 18:48, 20 July 2008 (UTC)

WEP Key

Hi I'm using an EchoLife HG520s wireless router and have forgotten my wep key. Well my dad has. It's not a problem cos the computer has stored it automatically, but I want to use my new DS wirelessley. there is an option to enter 3 more WEP keys but we are frightened we will lose the original and mess it up or something. Can we enter a new wep wihtout losing the old one, and then go back to the old one when necessary? 84.13.111.201 (talk) 16:23, 19 July 2008 (UTC)

That would depend on the specifics of that router, I don't know for sure. But if you've got admin access to the router itself (i.e. you can add, edit, or remove WEP keys), so what if you lose the old one? All your dad needs to do is re-enter the new WEP key if that happens, and then his wireless connection will work just fine, right? It's not exactly a big hassle. -- Captain Disdain (talk) 17:13, 19 July 2008 (UTC)

WEP is such low security that if you really need to find out what it was, you could probably crack it. There are lots of pages on how to crack WEP keys on the internet if you google around (e.g., here's a nice one for Windows XP). --98.217.8.46 (talk) 17:45, 19 July 2008 (UTC)

Another option is to use a "password revealer" to figure out what that password is. I won't provide a link, but you can find them if you look around. Have you tried logging into the router and just setting a new one? ---J.S (T/C/WRE) 02:26, 20 July 2008 (UTC)

Also, depending on the router, it may store it in plaintext in the admin screens. My father forgot his router password awhile back but his own machine was connected to it. I fired up the router admin, assumed he had not changed the default admin password (which was true), and then just viewed the source of the password page, and there it was in a hidden INPUT field. (I think this was a Linksys.) --98.217.8.46 (talk) 21:28, 20 July 2008 (UTC)

Depending on the router (mine, for example,is a Belkin) you could "hard reset" the router by pressing the tiny button on the back with a pin, sharp pencil, or other pointed object. This will restore the factory settings and allow you to enter a new WEP key. -=# Amos E Wolfe ^talk #=- 22:17, 20 July 2008 (UTC)

File organization on disk

I have purchased a new mp3 player. It is a very cheap mp3 player: when I place mp3s on the disk, it organizes them in a random fashion, supposedly by the order at which they're stored on the disk, not by filename (like most mp3 players). Is there any sort of software to verify that files are stored in a desired order on the disk? Windows is preferrable, but I use Linux at the library sometimes. The Evil Spartan (talk) 18:41, 19 July 2008 (UTC)

ls -f shows filenames in the order they're found in the directory, which is probably what you want. --tcsetattr (talk / contribs) 20:12, 19 July 2008 (UTC)

Being able to see the order is well and good, but it doesn't help me organize them (especially on a Linux system without a usable terminal). The Evil Spartan (talk) 23:37, 19 July 2008 (UTC)

If the directory order (ls -f) is the correct one, I have some bad news for you: there's no mechanism in unix to explicitly request a specific ordering of the entries in a directory. The filesystem lays out the directory however it sees fit. A tool that presents a list of filenames in a directory is supposed to sort them, like ls without -f does. Simple filesystems just keep the filenames in the order they were created, until one is deleted, which leaves a hole, then the next one goes in the hole (if the hole is big enough to contain the new name). Then there are more complicated layouts in which the order might be based on a hash of the filename making it completely impossible to get the order you want.

So the only suggestion I can make is to make a new directory, and move the files into it in the order you want them listed, then remove the old empty directory and rename the new one. That might work, although it'll be annoying if you change the contents frequently and have to move all the files in and out every time. Or maybe there is a reorganization tool that I don't know about, and some other refdesk person will tell us about it... --tcsetattr (talk / contribs) 00:36, 20 July 2008 (UTC)

Does your MP3 player support sub-folders? That might worth a try. ---J.S (T/C/WRE) 02:53, 20 July 2008 (UTC)

It does support subfolders, though I'm completely lost at what tcs is suggesting above. The Evil Spartan (talk) 18:14, 20 July 2008 (UTC)

Your MP3 player uses some variant of the FAT file system. In FAT, a (sub)directory is just a list of files in arbitrary order. Some software will show the file names in alphabetical order, but that's because it sorted the file names itself after reading them in. Most if not all FAT implementations will just add new files to the end of the list. So if you create a new directory, move the files into it in the order you want, then remove the old (now empty) directory and rename the new one to the old one's name, you'll end up with a directory with the same files as before, but listed in a new order. This is probably the order used by your MP3 player. -- BenRG (talk) 13:04, 21 July 2008 (UTC)

I've heard of, but never used, FATSort, which is designed for this type of problem. --Bavi H (talk) 20:27, 20 July 2008 (UTC)

Internet use in non-English

Do people who speak and use on their computer languages other than english type web addresses with the english alphabet or with their own alphabet/character set? For example, do spanish web addresses make use of the characters á, é, í, ó, ú, ü, and ñ? Do arab web addresses use the arabic alphabet? Do chinese web addresses use chinese characters? Or does everbody only use abcdefghijklmnopqrstuvwxyz and whatever symbols are normally allowed?--71.175.123.61 (talk) 19:07, 19 July 2008 (UTC)

I think everyone uses the same characters as us for now, but I also think that's set to change very soon - as in it's already been decided, just not implemented. I'd apreciate someone who actually knew backing me up or teling me I'm wrong, though. Blood Red Sandman ^{(Talk) (Contribs)} 19:19, 19 July 2008 (UTC)

article:: Internationalized domain name --tcsetattr (talk / contribs) 20:05, 19 July 2008 (UTC)

credit card security

Hi folks, we all know not to give our credit card numbers out etc. etc., but some time ago I heard of something more sinister, where people like computer hackers could search for numbers using some computer algorithm, without having to read the number where the user had entered it. The algorithm, as I understood it, simply detected or generated valid numbers, and that was enough. Is this true, and how does it work? If it is true, does it generate numbers that belong to actual cardholders, so the transaction would appear on their bill, or does it generate numbers that belong to no one, but simply fool the receiving system into thinking they are genuine? thanks, 203.221.127.240 (talk) 19:56, 19 July 2008 (UTC)

Generating bank card numbers is trivial - the prefix identifies the issuer, most of the remaining digits are just a serial number, and the last digit is a simple checksum using the Luhn algorithm. Most numbers won't correspond with actual accounts, but bruteforce will find those that do. But a number by itself isn't useful - you need to know the expiry date (that's trivial to bruteforce, if you can do so without setting off the accepting bank's security system - for current cards there are only a few years worth of monthly digits). So it's not very hard for brute force to find a valid credit card number and the associated expiry date. But that's not enough to commit fraud. For a card-present transaction (where you give the card to the merchant) the terminal sends a bunch of other info (that's encoded on the mag stripe); there's no practical way to brute force or guess the name etc. encoded there. For a card-absent transaction (that's where the merchant doesn't touch the card, e.g. over the internet or phone) then the transaction has to be accompanied by other information (name, address, security code) which again is impractical to guess.

Excepting very small amounts in out-of-the-way places that have offline stored authentication terminals ("zipzap machines", which don't validate small transaction online but instead store a recent blacklist of stolen cards) all transactions are verified live online, and Visa (etc.) check through to the issuing bank and make sure the account really does belong to a specific person and they're in good credit. And they verify all that additional info I mention above. So generating random numbers is futile.

Rather than brute force, the way most credit card fraud is done is to steal the numbers, and the associated authentication details, pertaining to a live account and a live person. Most common is places like hotels and restaurants, where crooked workers swipe a victim's card through the criminal's card reader as well as the hotel's own, so they get a copy of all the mag stripe track contents - the crooked worker can also take note of additional info (hotels are best for this) like customer's address. Then the criminals make a new card with those same details and buy resettable goods with it. The industry's main defense against this is sophisticated spending-pattern heuristics (that block cards that do uncharacteristic things, such as buying a speedboat in Macao when you normally buy gas in Tulsa); increasingly Chip and PIN and its ilk are being deployed to back this up. -- Finlay McWalter | Talk 20:53, 19 July 2008 (UTC)

I should correct myself - "serial number" should be "number based on account number"; numbers are generally not issued in plain dumb series, although there is often a serial element. -- Finlay McWalter | Talk 21:18, 19 July 2008 (UTC)

Thanks, that was a great answer. Does anyone know if there are any vendors who nevertheless don't bother with all the authentication (Finlay mentioned above that some don't for very small amounts, but I mean for larger ones), or is this simply banned by the credit card companies (meaning the merchant would automatically bear the cost of fraud)? 203.221.126.187 (talk) 15:41, 20 July 2008 (UTC)

Visa (etc.) sets all the terms of the deal - if a merchant doesn't follow the procedure for authentication then the merchant doesn't get paid; worse, if a merchant gets too much fraud or too many charge-backs the credit card company will drop them (technically the accepting bank from which the merchant leases their terminals will drop them). For a credit-card based business like a hotel or gas station not being able to accept credit cards is a death sentence. The one exception I can think of is pre-auth (where a hotel reserves an amount on a guest's credit card, in anticipation of a future bill); hotels that have an existing relationship with a regular guest (or a company that regularly has its people stay at the hotel) might well not bother with the preauth.; In that circumstance the guest could theoretically leave without paying and the hotel couldn't get the money back from their credit card. But hoteliers are rarely silly enough to extend such unsecured credit to anyone they don't already trust. -- Finlay McWalter | Talk 15:57, 20 July 2008 (UTC)

In the case of hotels, the guest may be required to sign a lodging agreement prior to each stay. This agreement would state that the guest must pay all fees incurred during the stay, whether they are planned or incidental (such as pay-per-view movie rentals, phone charges, or even damage to the room). If the guest's credit information is on file, the hotel can then charge the card for the total amount when the guest leaves. The lodging agreement allows them to make this charge without further authorization. Similarly, I have heard of restaurants occasionally doing a "pre-auth" of the expected tip on a bill paid by credit card, in case a guest doesn't leave a tip. This practice, if it exists, does seem to fall on the more illegal side of things. --Several Times (talk) 20:48, 21 July 2008 (UTC)

cd

can anyone please tell me what a number followed by the word DRIVE means when displayed on a portable cd player thanks Ken Robinson —Preceding unsigned comment added by 86.13.215.119 (talk) 21:08, 19 July 2008 (UTC)

what is the make/model? 71.164.103.123 (talk) 21:40, 19 July 2008 (UTC)

PDF from Scanner

I need software that will take pages scanned from my all in one machine and output them to a PDF. My all in one does scanning with an automatic document feeder, however, it only does one sided. All of my originals are two sided. I would like to scan the document, flip the pile over, and scan the other sides, and end up with a PDF with all of the pages in order. The document does not need to be scanned OCR, just needs to be a PDF with all of the pages in order. Is there anything out there, preferably open source or freeware, that will do this--for people with one sided machines with 2 sided originals? Thank you so much

John 71.164.103.123 (talk) 21:38, 19 July 2008 (UTC)

pdftk + this -- Finlay McWalter | Talk 22:09, 19 July 2008 (UTC)

Thank you so much. I've installed the windows binary. You wouldn't happen to know the command for inserting a page from pdf2.pdf after every page from page1.pdf, would you? —Preceding unsigned comment added by 71.164.103.123 (talk) 22:20, 19 July 2008 (UTC)

Date format in PHP: WHAT WAS WRONG?

echo date($A_Date) . ' ---- ' . $A_Date;

Output: 2001-12-11 00:00:00 ---- 2001-12-11 00:00:00

echo date(ymd, $A_Date) . ' ---- ' . $A_Date;

Output: 691231 ---- 2001-12-11 00:00:00

echo date(Ymd, $A_Date) . ' ---- ' . $A_Date;

Output: 19691231 ---- 2001-12-11 00:00:00

$A_Date is retrieved from MySQL. The original data set is compiled on Microsoft Access and imported into MySQL. What was wrong with this? Does the PHP parser reads 00:00:00 as the date? -- Toytoy (talk) 22:07, 19 July 2008 (UTC)

date() works on Unix timestamps, not date/time strings. You need to run $A_Date through strtotime() before you pass it to date(). — Matt Eason ^{(Talk • Contribs)} 01:22, 20 July 2008 (UTC)

It is working correctly. In the first one, you asked it to print "2001-12-11 00:00:00" using the current timestamp. Notice that you didn't ask for any year, month, or day. You told it to literally print "2001-12-11 00:00:00" - which it did. In the next two, you gave it a completely invalid timestamp. So, the timestamp is 0, which is the stroke of midnight between Dec 31, 1969 and Jan 1, 1970. Therefore, it is printing the year, followed immediately by the month, followed immediately by the day. All in all, it appears that you haven't taken the time to read the documentation here. -- kainaw™ 18:39, 20 July 2008 (UTC)

Also, beware: for mysql allows day and month to be zero, but functions that parse date-strings tend to give (sometimes very cryptic) error messages if day or month is zero. I once spent days trying to find the cause of a "strptime parsing error" reported by some Python library. 195.35.160.133 (talk) 15:00, 22 July 2008 (UTC) Martin.

What's the easiest way for me to vertical sync in Windows (2D)?

Say I want to do something very simple but in a way that is synced vertically, it has to be fast enough to work on an OLD computer (pentium III, < 500 mhz) and with a very very old graphics card. But it is only 2D. I'm willing to consider: C; C++; Python; Perl; Java; anything else. THe point is for me to be able to get to there as FAST as possible, I've listed the languages I know... Thanks. —Preceding unsigned comment added by 83.199.57.103 (talk) 22:59, 19 July 2008 (UTC)

What is ADFU?

I know it is related to MP3 players. But what this acronym stands for? 93.80.171.61 (talk) 23:33, 19 July 2008 (UTC)

I've been able to find out that DFU stands for "Device Firmware Update" [1], at least in the context of IPhones. I suspect that ADFU is something similar. My Google searching hasn't revealed much about it, except that it's a mode that some people seem to get stuck in. ---J.S (T/C/WRE)

Automatic Device Firmware Update, perhaps? - SigmaEpsilon → ΣΕ 02:48, 20 July 2008 (UTC)

Maybe, but no google results, so I didn't have any confirmation. ---J.S (T/C/WRE) 02:51, 20 July 2008 (UTC)