Wikipedia:Reference desk/Archives/Computing/2012 June 7

Computing desk
< June 6	<< May | June | Jul >>	June 8 >

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.

June 7

Graphics card causing reboots?

Hello! I bought an nVidia (evga) GTX 550 Ti card close to the start of May, installed it, and it was working fine. Games played well at 1080p at full frame rate with no problems. Then, about a week and a half ago (so three weeks after installing the card), my computer would start randomly rebooting...at first just one freak occurrence, then once every few days, up to the point it couldn't be up and running for more than 5 minutes. I was using a 500 W PSU, so I replaced it with a 700 W model...same problem. I've now removed the graphics card and have been running smoothly for 3-4 hours so far. The card specs recommend a minimum PSU of 400 W, so the 500 W and current 700 W should be more than enough to support it.

I dualboot to Ubuntu 12.04 and Windows 8 CP...same problem in both. In fact, a reboot was triggered even just being booted in memtest86+, so I doubt this is a driver issue. I was, however, able to run a pass in memtest86+ failsafe mode, but I don't know if that's significant, or just dumb luck that it stayed on for that after rebooting twice in non-failsafe mode of memtest86+ (no memory errors). Besides rebooting, it would occasionally just freeze the display in Ubuntu and Windows, and not reboot by itself. There was no useful information in the Ubuntu logs, which acknowledged the kernel was going through the boot process again but did not log any shutdown. These reboot problems happen even if I'm not doing any CPU- or GPU-taxing operation; simply browsing in firefox or even being away from the keys has triggered a reboot.

The CPU is always in a suitable temp range, and there are no thermal problems logged in Ubuntu. I strongly suspect the graphics card and am about to RMA it, but I figured I'd check here to see if anyone else might offer some other diagnostic advice. Unfortunately, I don't have access to another computer where I can pop this gfx card in and see if the same problem occur, nor do I have an extra graphics card to check if the motherboard pci-express slot is bad (there's only one pcie slot on my board). Looking online, I find most people with defective gfx cards get them dead on arrival or get some obnoxious BSOD that points to the gfx card (I've never gotten a BSOD), and I was surprised that this card worked so well for a few weeks and then suddenly brought on these problems, so I wanted to check here and see if that sounded plausible. Thank you very much for any advice you can give on this matter. I will update you if I have any problems with the card now removed, but so far, this is very much longer than I've been able to go without a reboot.--el Aprel (^facta-_facienda) 00:46, 7 June 2012 (UTC)

Have some paragraph breaks, on me: ¶ ¶ ¶. :-) StuRat (talk) 01:32, 7 June 2012 (UTC)

I agree with the possibilities of a bad slot or bad graphics card. Another possibility is that the card just wasn't seated properly in the slot. Did you put it back in all the way to check this ? StuRat (talk) 01:21, 7 June 2012 (UTC)

Following your advice, I reinstalled the gfx card inside the machine, and made sure it's secure (screwed into the case and mobo pcie latch flipped to hold it in), but still have the same issue. I was able to get a good 2 hours out of it, but it rebooted again. However, as I was messing around with it after its self-provoked reboot, I found I could trigger hangs and reboots manually by applying a bit of force onto the top of the card with the tip of my index finger. To further explain: If I pushed down on the graphics card with my index finger (pushing it in the direction of the mobo), sometimes giving it a gentle shake with just the index finger (as if you're making a smudge on a piece of paper with your finger), this action will often trigger the reboot process, leading me to believe it's a gfx card or pcie slot issue.

Sometimes this action doesn't work, but it seems if the computer is up for about 2-3 minutes and I try this, that little bit of agitation usually acts as a way to trigger this problem. Of course, not doing this, the system eventually reboots on its own anyway. I've checked and checked again, and the gfx card is certainly secure, firmly held in the pcie slot...there's definitely no feeling of looseness when I use this finger tactic. Maybe this can be diagnostically useful. Also, what's the best way to test for a faulty PCIe slot? FWIW none of the components in this build are older than January 1st of this year.--el Aprel (^facta-_facienda) 04:23, 7 June 2012 (UTC)

Well, we seem to have narrowed it down to an intermittant connection, but it's still not clear whether it's on the card, in the slot, or between the two. The best way to test the slot is to try a new card in there. At this point, assuming you can't borrow another card to try, or borrow another computer to test the card, I think you need to bite the bullet and buy a new graphics card. (Keep the receipt just in case that's not the problem, though, and you want to return it.) StuRat (talk) 04:41, 7 June 2012 (UTC)

I also added paragraph breaks to both my posts above to make them more readable.--el Aprel (^facta-_facienda) 04:28, 7 June 2012 (UTC)

Update: After removing the graphics card, I noticed that giving the whole computer case a gentle shake also triggers this problem, so it must be a loose connection somewhere, but I'll be damned if I can find it. The obvious point of interest was the PSU connection to the motherboard, but I've checked that and it's secure. Again, I've tried two different PSUs, so it's not a problem with the connecting wires themselves. I'll probably have to resort to taking it apart and reassembling it, since no loose connection is visually obvious. If that doesn't work, though, what are some diagnostics I can do to see if the problem is embedded in the motherboard itself? Thank you.--el Aprel (^facta-_facienda) 18:48, 7 June 2012 (UTC)

Have you got automatic restart disabled in windows? This will tell you if the computer is 'crashing'(blue screening) or actually rebooting. If it is actualy rebooting, my guess is it's most likely a RAM issu. How many ram chips are you using? If it is more then 1, simply try taking each out one by one and seeing if your problem stops.(source: ten years in IT support). Vespine (talk) 01:47, 8 June 2012 (UTC)

If you get the problem only when using the graphics card (it disappears completely using the built-in graphics), then I don't see how it could be the motherboard or ram chips. It's the graphics card, the slot, or the connection between the two. StuRat (talk) 04:49, 8 June 2012 (UTC)

The post from the OP directly above mine says they tried to remove the card but still got the problem.. Unless i'm misunderstanding something. Vespine (talk) 04:28, 12 June 2012 (UTC)

OK, I was reading "after removing the graphics card" to mean "after removing and then reseating it", which would be more consistent with what they reported earlier. But, you may well be right. StuRat (talk) 19:32, 12 June 2012 (UTC)

PHP Function.

I'm learning php, so I did a "proxy" to test my 'skills', I showed it to someone experienced in netwroking. He managed to acess my computer files (He even downloaded some files, and pasted them as a proof) and access my network.. I asked him several times how he did it but he didn't show me..

Anyway here Is the function that get's the files, as you can see, the function that gets everything is file_get_contents, how he managed to do it?

 

$url = $_POST['pageurl'];
$opts = array(
  'http'=>array(
    'method'=>"GET",
    'header'=>"Accept-language: en\r\n" .
              "User-Agent:php"
  )
);
$context = stream_context_create($opts);
$contents = file_get_contents($url,false,$context); 

Thanks for reading 190.158.212.204 (talk) 01:52, 7 June 2012 (UTC)

I discovered how the vulnerability works. how can I filter it? 190.158.212.204 (talk) 04:39, 7 June 2012 (UTC)

The vulnerability (as you have probably figured out) is that the URL doesn't have to be a real URL — it can also be a file path on the server, and you haven't filtered it at all, so unless you are really careful with how your PHP's permissions are set, you can put in all sorts of nefarious paths like "../../../Users/YourUserName/Desktop/SecretFiles.txt" and so on. The base lesson here is "be careful when you take user input and use that in a function like file_get_contents which can access local files." How to fix it? Sanitize the user input — if you only want it to refer to URLs, make sure that the string starts with "http://" or "https://", and if not, add it to the front of it (http://../../../etc would not resolve to a valid URL, but it won't access your server filesystem, either). If you need it to access local files (or relative paths), reject any paths that include double periods (..). --Mr.98 (talk) 00:36, 8 June 2012 (UTC)

What is the price to put video ad on youtube (the video ads that plays before the video start)?

What is the price to put video ad on youtube (the video ads that plays before the video start)? — Preceding unsigned comment added by 201.78.132.208 (talk) 05:35, 7 June 2012 (UTC)

http://google.com/search?q=youtube%20advertising ¦ Reisio (talk) 07:09, 7 June 2012 (UTC)

Memory Space

Whenever I but a pendrive or a portable hard drive - say 1000 GB portable hard drive, I dont get to use the entire space in it. In the case of the 1000 GB hard drive, I am able to use only 930 GB. So is the case with pen drive - around 95 % of the space is only usable. Why is it so ? — Preceding unsigned comment added by 117.193.139.99 (talk) 08:09, 7 June 2012 (UTC)

It's likely you are in fact using all (or at least most) of the space, but that the capacity as printed on the device and the capacity as given by your operating system are arrived at by using different measurements, specifically: the device manufacturer has measured in base 10, and the OS (for historic reasons) has measured in base 2. ¦ Reisio (talk) 09:09, 7 June 2012 (UTC)

In addition to base-10 vs base-2, a file system reserves space for metadata - file names, modification time, ownership, permissions, and such. This space will be unavailable for file contents. 88.114.124.228 (talk) 16:34, 7 June 2012 (UTC)

Wikipedia article Binary_prefix#Consumer_confusion... -- AnonMoos (talk) 02:01, 13 June 2012 (UTC)

gmail

Resolved

How do i delete a pre-existing email address which is remembered by gmail "compose"? Kittybrewster ☎ 10:06, 7 June 2012 (UTC)

I think you have to delete the email from your "contacts" list. Heck froze over (talk) 13:26, 7 June 2012 (UTC)

it isnt on it. Kittybrewster ☎ 13:37, 7 June 2012 (UTC)

You probably have multiple contacts lists. Click on the red "Gmail" top left of screen, and select Contacts. If you see multiple contacts lists, you will need to work through each of them to find the one you want to remove. Also turn off "Create contacts for auto-complete" in your general settings options.--Shantavira|^{feed me} 15:02, 7 June 2012 (UTC)

Many thanks Kittybrewster ☎ 15:16, 7 June 2012 (UTC)

Email SMS alerts.

Does anyone know of such a service reminding a user that he/she has email via a text message. I know there is one, but it's only available for subscribers in India, Pakistan and elsewhere, but not in the Philippines. Blake Gripling (talk) 10:26, 7 June 2012 (UTC)

All implementations of such service (which i have used), consisted of email to SMS gateway (email address which forwards received emails as text messages to specific phone) and ordinary email forwarding (which is set up at mail server). First part (SMS gateway) depends on cellphone operator (they owns it), second part (email forwarding), depends on email server. -Yyy (talk) 09:17, 8 June 2012 (UTC)

Scorewriter licensed under BSD License or similar licenses

Are there general-purpose scorewriters with extensive functionality and licensed under BSD License or other copyfree licenses? 117.5.13.6 (talk) 11:01, 7 June 2012 (UTC)

Wikipedia has a list of scorewriters that should answer your question.--Phil Holmes (talk) 08:25, 8 June 2012 (UTC)

If Comparison of scorewriters is to be believed, the answer is no, unless the one LGPLed program counts. I'm not sure I know what copyfree means. -- BenRG (talk) 06:27, 9 June 2012 (UTC)

Computer security

Hi all, I remember back in the 90s people used to tell me you couldn't catch a virus just from being connected to the internet, and you couldn't catch anything from a webpage unless the owner of the page had put one there (or, say, unless a virus on his computer had put one there). Nowadays all you hear is that a (Windows) computer only has to be connected to the Internet for a few minutes to catch its death of cold. What has changed in the meantime? Is it that I used to listen to people who didn't know what they were talking about? What is the situation for Mac and Linux? It all seems rather strange to me, and it sounds like a rather basic thing to be able to prevent a computer from catching a virus except in the case of downloading and running software. Note that by "downloading" I mean in the sense of being able to run downloaded software from the start button, rather than in the controlled environment of a webpage, as happens with javascript. IBE (talk) 13:29, 7 June 2012 (UTC)

Hi. Nowadays, hackers (I shall refer to them as crackers from now on, and crack instead of hack), and they also crack for very different reasons. It was technically possible for a cracker to spread viruses just because you were connected to the internet in the 90s. However, most crackers then cracked due to their own mischief, they do not expect to earn money out of it, they only want to see people's computer crash and the feeling of invincibility. Because of this, they mostly did not bother making their virus spread in such a way that even those connected to the internet without doing anything else would also get infected. Therefore, people just ASSUMED that virus cannot spread just because you are connected to the internet. Unfortunately, times have changed. Crackers now crack because of economic reasons, that is, they want to earn money. They code viruses that stays rather dormant in your computer, slowly logging your personal details such as credit card numbers and sending them back. Since it is credit card numbers and other personal details they want, it means that the more they have, the merrier. Therefore, more energy is spent on making sure that every, or at least, nearly every, channel for the virus to spread is exhausted. Therefore, we have viruses that can spread, for example, to all computers in a LAN from just one computer connected to it. Thus, we come to assume that viruses now can spread just by being connected to the internet. Cheers! ★ Oliverlyc ★ ✈✈✈ Pop me a message! 14:29, 7 June 2012 (UTC)

Well technically you can't get a computer virus just by being connected. You may get a computer worm. More seriously, what you're saying is only really true for a preSP2 Windows XP that isn't behing a NAT or firewall. PostSP2, the firewall does a decent job provided it's properly configured (and of course enabled). Also, I don't know what people you listined to, but the risk of services open to the internet was well recognised even before Windows 95 let alone Blaster (computer worm). (There were of course a number of Internet Information Services vunerabilities that predated blaster and were fairly widespread like Code Red (computer worm) although most Windows users didn't/don't run IIS.) Think of OpenBSD's mantra for example (which wasn't before Windows 95, but you get the idea). 2001:0:5EF5:79FD:20CB:1C04:833A:FA41 (talk) 15:29, 7 June 2012 (UTC)

You can't get a virus just by being connected, but a Windows computer that detects an internet connection will do some things automatically. For example, it will automatically download updates. Looie496 (talk) 03:09, 8 June 2012 (UTC)

The auto updates can be turned off (and updates done manually), but it can be quite tricky to ensure that they are all disabled. Dbfirs 21:04, 8 June 2012 (UTC)

Auto updates, at least those from Microsoft, are safe. They have a digital signature which is verified locally. Yes, there was just a story in the news about Microsoft revoking some bogus keys which could apparently be used to sign updates, but that's an extraordinary circumstance. Your average Joe villain doesn't have access to that kind of thing. Disabling auto updates is very unsafe since your system will accumulate vulnerabilities which are known to everybody (since the updates themselves represent a public disclosure). It's important to install security updates as soon as they are made public. Many, even most, real-world attacks exploit vulnerabilities that have already been patched and so affect only people who don't update (which is a lot of people, unfortunately). -- BenRG (talk) 22:08, 8 June 2012 (UTC)

Interesting answers - thankyou. So does that mean the only true vulnerability with Windows is the checking for updates? If you configured it so as not to check for updates, would it be safe, at least until you visited a dodgy porn website? It just strikes me that it should be simplicity itself to protect a computer from unsolicited code. Stuff you download and run can obviously be malicious, as with anything (even offline) that asks for your private information, but stuff just jumping onto your system looks preventable. Does anyone know the situation with Macs? IBE (talk) 05:42, 8 June 2012 (UTC)

The chief thing that has changed since the 90's is that programs are expected to interact with each other in to a much larger extent today than what was the case 15 years ago.

Is update checking the only vulnerability? No, definitly not, any program that downloads code or content that can contain code can be a vulnerability. So the updates are necesarry in order to plug those holes.

In principle it is possible to prevent attacks by blocking anything that could contain code. However the current trend is that content and presentation should be synchronized with the cloud and with a users several devices (Phones, pads, other PC's). So content will be downloaded from outside of the computer, and this will reach (or have reached) a volume that means that direct user inspection would be annoying. Also content is becoming more complex and thus the need for content that can contain code is increasing.

Mac is currently somewhat safer than Windows, arguably because Mac is a less atractive target (they still have a smaller userbase). Wether this will hold as mobile devices gets more important remains to be seen. Mac might have an advantage over Windows because of tighter control on third parties. OTOH Windows might have an advantage because there are several compaines developing antimalware for the platform. Taemyr (talk) 11:49, 8 June 2012 (UTC)

As I said above, disabling auto updates makes your computer much less secure. Please don't do it. -- BenRG (talk) 22:08, 8 June 2012 (UTC)

Yes, apologies for mentioning this. I agree with Ben that Windows updates are important. I have auto-updates disabled on my computers for local reasons, but I do update regularly.Dbfirs 07:17, 9 June 2012 (UTC)

Thanks again - so if the problem isn't auto updating, is it just what Taemyr says, that user expectations for interactivity are high? Because this still seems inherently fixable. Again, I still find it very strange that a remote blackhat can get a computer to download and run some software when all I was doing was registering a connection. Cloud computing is a very small thing to me, and seems a small thing to gain for an insecure computer. Why should the threats be so blatant? IBE (talk) 12:04, 10 June 2012 (UTC)

This is a very complicated problem, but I don't think what Taemyr said is accurate. Generally computers are much more secure now than they used to be. Consumer versions of Windows didn't even attempt to provide a real security model until XP, and Mac OS didn't even attempt it until X. Early Linux distributions, like the Unices before it, shipped with a lot of remotely-exploitable Internet services enabled by default. The Morris worm exploited bugs that had, I think, been known for years, yet remained unpatched on professionally maintained systems. That's unthinkable now. I don't think there's been an increase in "interactivity". People have always run software on their computers. Microsoft Office has had a full-fledged programming language built in for 20+ years. When you hear stories about malware that escapes the Javascript sandbox, just remember that the very existence of the sandbox is a step up from the traditional Unix/Windows model of always giving all of your privileges to all of the software you run.

To the extent that the situation is worse than it used to be, it's because malware is now mostly written for profit instead of fun and because most computers are connected to the Internet a lot of the time now, which makes it possible (or at least much easier) for malware to phone home. But fearmongering around this issue is nothing new. Don't be afraid just because someone says you should be. They're probably trying to sell you a magazine, an antivirus product, or a government agency. -- BenRG (talk) 16:47, 11 June 2012 (UTC)

Yes, MS office has had an inbuilt programming language for a long time. But when email started becomming common that fact was a mark against the suite, in that it was an uncessary risk that you could avoid by using other software. Today it's far less of a mark against MS, mostly because safeguards are far better, but also because it's getting much harder to avoid. Taemyr (talk) 09:26, 12 June 2012 (UTC)

Also to OP, no it's not just cloud computing. It's not just any single thing. It's all things at that same time, most serious threats today (Ie. threats that have consequences that the user will notice) will use several weaknesses in order to have an effect. Opening a connection means that you expect to get data in return. That means that your firewall will need to allow data from what it thinks is the source. And so creating a connection will open an avenue of attack. Also see BenRG's comment re security of computers. Computers today are, provided the software is kept up to date, reasonably secure.Taemyr (talk) 09:33, 12 June 2012 (UTC)

iPhone charger

I think that I have asked this before but mislaid the answer. Can I charge my iPhone with my iPad charger please?85.211.222.224 (talk) 14:07, 7 June 2012 (UTC)

I suggest you try searching the archives (there's a box at the top) but AFAIK you shouldn't have a problem. Charging your iPad with an older iPhone charger may not work however. 2001:0:5EF5:79FD:20CB:1C04:833A:FA41 (talk) 15:15, 7 June 2012 (UTC)

Thanks for this, found my earlier post and links, and the answers seems to be 'Yes'. Thanks again.--85.211.222.224 (talk) 15:49, 7 June 2012 (UTC)

wow, that's the first ipv6 I've seen here 190.158.212.204 (talk) 22:17, 7 June 2012 (UTC)

What is an ipv6? — Preceding unsigned comment added by 85.211.222.224 (talk) 06:08, 9 June 2012 (UTC)

We have an article about it. →Στc. 06:12, 9 June 2012 (UTC)

Java compiler

Can you suggest me some good compilers to make Java software? Cambalachero (talk) 15:37, 7 June 2012 (UTC)

The Java Development Kit, which you get from Oracle, comes with javac, the standard Java compiler. For a general developer, this is the first and most obvious choice. -- Finlay McWalterჷTalk 15:54, 7 June 2012 (UTC)

And if you are more into a GUI environment, Eclipse (software) is a popular free choice, though not the only one; there are links to others in the Eclipse article. A common answer to this is to play with javac first, then graduate to a GUI, in the spirit of "learn to walk before you run". 88.114.124.228 (talk) 16:18, 7 June 2012 (UTC)

BlueJ is also a very good developing environment for beginners. It's good because it automates a few tasks (like the compiling process) and has basic syntax highlighting, but doesn't have any advanced feautures like auto-completion, thus forcing beginners to get very comfortable with Java and not rely on features that do too much of the work for them. Oracle's javac seems like the way to go.--el Aprel (^facta-_facienda) 18:54, 7 June 2012 (UTC)

We have an article, java compiler. Another well-known one is GNU Compiler for Java. 81.98.43.107 (talk) 23:37, 13 June 2012 (UTC)