Wikipedia:Reference desk/Archives/Computing/2012 December 20

= December 20 =

Browseprotect.exe
I downloaded a screensaver (yesterday), since then a file "Browseprotect.exe" (in all cap) is asking permission to access internet. AVG pops up with options allow, block etc. Can you tell me if it is a legitimate system file and if it is a virus then how I can remove it (full scan or search and delete (though I can't find the file in search)! --Tito Dutta (talk) 05:45, 20 December 2012 (UTC)


 * There are plenty of "free screensaver" sites that bundle in malware, so I certainly would not trust this new program. I can't find much about "browseprotect.exe", but is some information online on "browserprotect.exe" that seems to come along with some free downloads and is hard to get rid of. Unfortunately, all of the sites that look promising are blocked on our company's network. (Although the obviously scammy "free pc fixer" sites are allowed...) I did a search for the filenames in ICE that confirms that there is no component of Windows 7 that contains "browseprotect.exe" or "browserprotect.exe". 209.131.76.183 (talk) 14:53, 20 December 2012 (UTC)
 * Yep, it's malware. These "free screensavers" almost exclusively make their money on selling advertisement and personal information.  i kan reed (talk) 17:45, 20 December 2012 (UTC)

How big can an internal HD be?
Low large of a capacity can an internal drive on serial ATA on a Windows computer be these days? Is the maximum capacity limited by the motherboard, SATA, Windows, NTFS, or what? Bubba73 You talkin' to me? 05:45, 20 December 2012 (UTC)


 * The largest hard drives manufactured today for PCs are 4 TB in capacity:.


 * The theoretical limit for the NTFS file system used in the newest releases of Windows is 256 TB. The ext4 file system used in Linux can support volume sizes up to 1 EB. Older motherboards without UEFI can only boot off of drives that are 2 TB or less in size. UEFI supports a maximum volume size of 8 ZB.


 * So, to answer your question, the capacity is limited by the tolerances within the hard drive itself, assuming the rest of the software and hardware in your computer is relatively new. Hard drives are increasing in capacity every year. The 4 TB drives hit the market in 2012 and 3 TB units were introduced in 2011, and so on. They've been increasing in capacity steadily since they were introduced in the late 1950s.&mdash;Best Dog Ever (talk) 06:41, 20 December 2012 (UTC)


 * It's usually possible to trick the computer to boot from the first < 2 TB (or 2 GB etc) and then load drivers needed to make use of the whole space. Common trick with BIOS limitations. Electron9 (talk) 17:27, 20 December 2012 (UTC)


 * The theoretical limit of NTFS is one yobibyte (264·64K). The implementation in Windows 8 is limited to 256TB, according to the article. -- BenRG (talk) 00:25, 23 December 2012 (UTC)


 * Thanks, and there is nothing on the motherboard, SATA, Windows, etc, that prevents a Windows system from using the whole 4TB? Bubba73 You talkin' to me? 07:19, 20 December 2012 (UTC)
 * It's basically down to which SATA standard version your controller complies with minus any chip bugs. As for Microsoft it's depend on driver (supplied by manufacturer) and filesystem (by MS). Beware the Windoze bugz ;) Electron9 (talk) 17:27, 20 December 2012 (UTC)


 * When I got this computer last year, 2TB internal drives were the largest you could get with the computer. It has eSATA and 8 or 9 months ago I got a 6TB external HD (two 3TB drives, RAID0), with eSATA and USB2.  I tried to use eSATA, of course.  But I couldn't write more than about 2TB to it.  At about the 2TB mark, it would drop down to less than 5MB/sec, and a while after that it would stop.  I tried it many times.  I got an eSATA to USB3 adapter, put it on my USB3 port, and it works OK.  I think something in my system doesn't support more than 2TB on SATA/eSATA.  Bubba73 You talkin' to me? 00:45, 21 December 2012 (UTC)


 * Which operating system are you running? Assuming you're running Windows, you must initialize any disks above 2 TB with a GPT &mdash; not an MBR: . Windows XP is not supported. It must be running Windows Vista or later.&mdash;Best Dog Ever (talk) 01:05, 21 December 2012 (UTC)


 * Windows 7 then, 8 now (both 64-bit). I think I did use GUID.  Bubba73 You talkin' to me? 01:23, 21 December 2012 (UTC)


 * So, you checked the GPT radio button shown in this picture: ?&mdash;Best Dog Ever (talk) 02:47, 21 December 2012 (UTC)

It seems to me that a few years ago, HDs were rapidly getting larger and cheaper/GB. They don't seem to be doing that now. Is that right? Bubba73 You talkin' to me? 02:40, 21 December 2012 (UTC)


 * It hasn't been consistent from year to year:


 * I think I rounded one of the figures above because I remember there was a year when they had 1.5 TB drives. But still, I don't notice a trend myself. In the 80s and 90s there were some years where it jumped like 300% and other years where it increased 50%, just like recent years.&mdash;Best Dog Ever (talk) 02:55, 21 December 2012 (UTC)
 * In terms of price/GB, price/GB has been continuing to drop and I don't believe there has been a major change in the rate. However there was a massive price/GB increase after the Thailand floods and prices are only now beginning to reach preflood levels. Nil Einne (talk) 04:36, 21 December 2012 (UTC)
 * I concur, the Thailand floods messed the supply and price of disc drives seriously. Electron9 (talk) 19:24, 21 December 2012 (UTC)


 * That probably explains my casual observation that it hasn't dropped compared to a year or two ago. Bubba73 You talkin' to me? 19:29, 21 December 2012 (UTC)


 * I don't think the original question specified if capacity was in bytes or cubic meters ;-) ANyway if you want to see how big a HD can be just look at the early ones in History of IBM magnetic disk drives which provides an extension of the figures above for data capacity too earlier dates. Dmcq (talk) 12:26, 24 December 2012 (UTC)

Parse two-dimensional array from string in C#
I have run into the following problem at work. I am given a string such as this: { {a, b}, {c, d} } I am supposed to parse this into a two-dimensional string array, such as, so that: s[0][0] == "a" s[0][1] == "b" s[1][0] == "c" s[1][1] == "d" The number of items in both dimensions can vary. I can get to know these dimensions beforehand. The format is otherwise always the same: the whole array is surrounded by  -   brackets, and so are the rows within the array. The rows, and the individual items inside the rows, are separated by commas. But I don't know at all how to write the parsing code. Is there a pre-existing library that can do this? Or if I have to write the parsing code myself, how can I do it? J I P &#124; Talk 07:21, 20 December 2012 (UTC)
 * I would tackle this by deleting the outer {} elements, using a regex to get all the text inside the inner {} pairs, and then string.Split to get the individual elements. Don't think it would be too hard to code up in 2 loops. --Phil Holmes (talk) 14:26, 21 December 2012 (UTC)
 * You don't really need str.Split when you use regex. Use one or the other. See here for regex, and this question for str.Split. --Wirbelwind( ヴィルヴェルヴィント ) 22:14, 21 December 2012 (UTC)

Serving scaled images and relative widths
Pagespeed is telling me to serve scaled images. However the design uses relative widths. Is there any way to combine the two? bamse (talk) 15:14, 20 December 2012 (UTC)


 * Firstly, of course, Pagespeed is just advice, and if you have good cause to ignore it, there's no shame in that. But do you really want the browser scaling images for you? The browser's image scaler tends to be rather poor. For background images in an adaptive design, I often find myself putting a fade-gradient on the edge of a background image, so that it fades to the background colour. Then I have the background non-repeating of a fixed size with a specified background colour too. That way it fills the background of the element acceptably, regardless of the size to which that cell is resized, and without scaling in the browser. -- Finlay McWalterჷTalk 15:42, 20 December 2012 (UTC)


 * Example; the original image (without the fade on the right) is on Commons -- Finlay McWalterჷTalk 16:11, 20 December 2012 (UTC)


 * Thanks a lot for sharing the idea and code. Unfortunately fading is not an option in my case. The images appear together with text in a table-like listing: think a forum with avatar images on the left and some text around it. Let's say the container of the image is 10% of the full width. As far as I understand, there is no way to prevent image scaling by the browser if the image is supposed to fill all of the container (correct?). I read somewhere that one could use javascript to provide differently sized images (to be further resized by the browser) based on the display size. Is this really worth the effort, particularly for a site which would typically not be accessed from small mobile phone screens? bamse (talk) 22:24, 20 December 2012 (UTC)
 * As a side question, is there some online-tool that allows to calculate recommended image resolution (of the stored image) based on the eventual display size (in % of screen size)? bamse (talk) 22:26, 20 December 2012 (UTC)

The connection to the server was reset while the page was loading.
The above message has been appearing in my browser with apparently increasing frequency - this applies across at least 7 machines using Palemoon and Firefox. Similar errors appear in IE.

Clicking on "try again" is usually successful, however.

I have attempted to discuss the matter with Talk Talk my ISP, but since they tell me that 0870 numbers are free to call, don't know the difference between dropped packets and a dropped line, don't return calls and will on no account pass me to a supervisor, I have little faith other than they may have performed a whoosh test, and it may have shown everything to be fine.

In an attempt to diagnose the problem I have re-pointed my DNS to Google DNS, reset the router many times over the last few weeks and preformed a few network diagnostics viz:
 * 1) emulated an http connection by using telnet
 * 2) checked traceroutes and pings for basic connectivity and possible route-flap (seems ok, but hard to be sure)
 * 3) used ping -t to check for packets being dropped - 0% to Google, .1 % to Wikipedia

My options seem to be based on getting a new router or switching ISPs. I am tempted to switch ISPs anyway since this bunch seem such a shower, but I'm not convinced that it will solve this problem. Any ideas? Rich Farmbrough, 16:00, 20 December 2012 (UTC).


 * Your symptom sounds like you may have a software proxy server or web cache - either on your machine, or somewhere upstream, provided by your ISP - that is misbehaving. Unfortunately, diagnosis of an amorphous symptom like this requires a lot of effort and often results in no progress fixing the problem; but from your described troubleshooting steps, it sounds like you are on the right track and have the technical proficiency to proceed.  So, here's my recommendations:
 * First, isolate whether the cache or proxy is local on your machine, or upstream by your ISP, by connecting a different machine to your cable or DSL modem. If the symptom reproduces on other computers on the same connection, the problem is not your fault!  An amorphous cache server configuration bug exists, it's your ISP's fault, and there is essentially nothing you can do to fix it until a skilled-enough technician happens to repair the error (and that sounds unfortunately unlikely).  Try to work around the bug; you can try using HTTPS/SSL/TLS in your browser, or try to disable caching, or so forth.
 * If the problem only occurs on your machine, and not on others, check for common cache and proxy and firewall software. It sounds like the problem occurs in all browsers, so it is not likely due to a browser setting.  Other places lower down the system stack to check include your system network settings and your antivirus programs.  Antivirus programs sometimes set up such network trickery for your "protection," and it has been my experience that their implementations are often imperfect.  Again, try to workaround the problem using non-caching or secure HTTPS pages.  Finally, check for malware "web accelerators" and similar ill-intentioned software that lives in your network stack; sometimes this can be quite well-hidden.  Good luck with the troubleshooting.  Nimur (talk) 18:38, 20 December 2012 (UTC)
 * Thanks, yes I thought some kind of caching, even worried that someone was intercepting my traffic clumsily. As I don't use an explicit local cache or proxy, and I'm getting this on 7 machines I shall provisionally blame the ISP, and probably move, hoping the new bunch aren't even more hopeless. Rich Farmbrough, 23:07, 20 December 2012 (UTC).


 * I've had the same thing, too.
 * Basically, screens start loading and even displaying, and when I cut the connection (I'm on 56k), the page vanishes and displays that silly error message. It's been increasing slowly and steadily from 2008 on, even after reinstalling, and jumped up recently when the "Edit saved" pop-up came up on wikipedia. (Is that malware? it looks damn genuine to me.)
 * To make a long story short, it's like some demon in my machine that decides, "If you don't stay connected 24/7, you don't deserve that WP content," which is silly. After all, I received most of the info, and all I want is a tall ship and a browser which shows me that info rather than a full-screen error message.
 * Cancel seems to avoid the error message, but it's a pain and a half to push cancel on each page before cutting the connection. - ¡Ouch! (hurt me / more pain) 06:51, 21 December 2012 (UTC)
 * Yes, I get the same error message sometimes, especially when it is raining heavily, but in my case the problem is almost certainly an intermittent internet connection (delivered via a microwave chain).   D b f i r s   09:31, 21 December 2012 (UTC)

How is Google connecting merchandiser ID data to the person using Google?
From this article: " If you bought a T shirt at The Gap in the mall with your credit card, you could start seeing a lot more Gap ads online later, suggesting jeans that go with that shirt." If you read the article, you'll see that what it's supposing is that Gap uploaded its sales data using this Conversions API that Google is going to offer to advertisers. So that's how your shirt purchase got known in the Internet. But how is your identity getting connected to Google ID whateverUser395, or even if the name Joe Blow on your credit card were loaded up by Gap and your Google ID was Joe Blow, there could be twenty other Joe Blows in the world. I don't remember giving any hard ID like address, and definitely not SSN, when I created a Google account. I wonder how uploaded advertiser data can get pinned to given Google IDs especially when it says in the article "Neither the advertiser nor Google will be able to track you individually." 20.137.2.50 (talk) 18:34, 20 December 2012 (UTC)
 * Google is getting scarier each year. OsmanRF34 (talk) 18:41, 20 December 2012 (UTC)
 * You might read business intelligence and data mining for a start. Also, keep in mind that 100.0% accuracy is not necessary; if a few errors occur, a few targeted advertisements "convert" to untargeted advertisement, which is not a total loss to the advertiser.  So, the specific data-mining and identity aggregation methods in use only need to be accurate enough to provide competitive advantage over alternative advertisement strategies.  Nimur (talk) 21:09, 20 December 2012 (UTC)

Content supervision
When a software designed to monitor what webpages a person visits (like the ones installed on computers in work and academic environments), does it merely record the URLs of the webpages you visit, or does it record the content of those websites as well? Like for example, if you're looking up porn on Tumblr, does it only know that you've been on Tumblr or does it know specifically what you were looking for on that site? And does using the "site:example.com" function of Google also give away your activities? 74.15.143.46 (talk) 23:04, 20 December 2012 (UTC)
 * The firewalls I used to administer would record the URLS, not just of the "pages" you visit though. If you visit a page the images are fetched by a separate request, so their urls will be recorded too.  it is also possible to record information over and above the url, but often search terms are in the URL.  I suggest for a fuller answer you look at the manuals of the software in question, familiarise yourself with HTTP and SSL, and get in the habit of looking at the URL in you address bar as you browse to understand more of what is going on.  Rich Farmbrough, 00:27, 21 December 2012 (UTC).


 * But the thing is, I'm not installing or running the software. I'm merely asking on the off chance that I'm having my personal web browsing activities monitored (I honestly don't know if that is indeed the case, and can't tell if it is). 74.15.143.46 (talk) 02:42, 21 December 2012 (UTC)
 * The thing is, it should be fairly obvious software could record everything if it wanted to, so it's a pretty meaningless question. Most set-ups don't record everything simply because there's little point (recording the entire URL would usually be enough, they could always visit the URL later to check what the content was if it hasn't disappeared and when it isn't you normally just have to record the request rather then the response to get an idea of what the person is doing such as by revisiting the page) and it would take a lot of storage space if you have a large number of users and they use a fair amount of bandwidth but you can't rule it our in some arbitary set-up. And even if the software doesn't normally record everything, it may still analyse pages and record stuff that may be of concern or the IT admins could choose to record stuff if a user is of concern. If you are concerned, your best bet is to read the appropriate policies or employment contract or whatever is relevant to your situation, seek help for any areas that confuse you and then don't do anything that violates it. For your personal internet access, choose a service provider that provides sufficient guarantees to allay your concerns. Note that presuming someone else has administrative control of the computer you're using then there's nothing stopping them recording everything done on the computer including stuff which you would not expect to pass through the network, as well as either recording a decrypted copy of stuff which uses end to end encryption lime HTTPS or otherwise enabling themselves to decrypt encrypted traffic. If you also have administrative control in addition, you could try to stop this but it may itself be a violation. Nil Einne (talk)
 * I work in IT for a fairly large corporation. I can tell you first hand that some of the systems in place these days used for large volume content filtering are very sophisticated indeed. We use a system that doesn't just "track" what you look at, but can heuristically analyse images in websites and email in real time and automatically flags it if meets the rules for inappropriate content, (the volume of flesh tone being the most obvious, but not only, example). It then goes on to be assessed by a human, but the system does 99% of the work.Vespine (talk) 05:39, 21 December 2012 (UTC)
 * Every color from black to brown to yellow to sallow white to white to pinkinsh white? Do you get lots of false positives like beach volleyball or lion films or teleconferencing with unusually skin-toned clothes or something? Porn is overrated. After seeing every kind of woman's body naked (about 3 months) I lost the phallomanipulative compulsion. (one of the last ones was Elaine Benes. Oh baby, you have no idea how hot she is until you've seen her (photoshopped) naked (and gotten a bit closer in age) Sagittarian Milky Way (talk) 21:22, 21 December 2012 (UTC)
 * I used to hope I'd get somewhat closer to you in age, too. But achieved fuck all. ;)
 * To OP. There could be some process capturing your screen in 30-sec intervals. That would defeat end-to-end encryption and many other tricks. The server could challenge the process using some kind of question only that process would answer, sl killing the process would draw some attention, too. Or maybe it's bayesically the kind of thing Vespine mentioned. - ¡Ouch! (hurt me / more pain) 18:31, 22 December 2012 (UTC)