Wikipedia:Reference desk/Archives/Computing/2013 January 29

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.

January 29

Running mobile apps on a Windows PC

To make discussion easier, I am dividing this question into four separate questions. Please answer in the sections below, and thanks for all answers. 180.245.210.151 (talk) 04:30, 29 January 2013 (UTC)

With great difficulty.

There are emulators that come with the developer kits of those platforms. (However, The iOS dev-kit is Mac-Only.) In theory you could use the emulators to run the apps on your desktop. In reality, emulators often don't work well and have many limitations. (For instance, as far as I know, the Android emulator has only the most rudimentary OpenGL support. It won't run games, or almost any other app that uses OpenGL for display.) Another stumbling block is actually getting the apps. You can't connect to app stores with an emulated device, only real ones. So you'd need to get a copy of the app from the developer (or pirate it, I suppose.)

In most cases it would be more trouble than it's worth. APL (talk) 09:14, 29 January 2013 (UTC)

How do I run Android apps on a Windows PC?

Use BlueStacks.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); January 29, 2013; 15:42 (UTC)

How do I run iOS apps on a Windows PC?

Run Mac OS in a VM. ¦ Reisio (talk) 16:00, 29 January 2013 (UTC)

How do I run Windows Phone apps on a Windows PC?

You can download the development tools for Visual Studio and this will allow you to run and debug 'your' applications either through an emulator or with an attached windows phone (you need to developer unlock it first). [1] Surely however, if you are serious about development, the small bit of googling needed to find this wasn't beyond you? ---- _nonsense ferret 16:58, 29 January 2013 (UTC)

could a kinect read depth information from something as small as the travel of a computer keyboard?

hi,

could a microsoft kinect read depth information from smething as small as the travel of the keys on a computer keyboard? 178.48.114.143 (talk) 13:21, 29 January 2013 (UTC)

I seriously doubt it. The resolution seems much lower than that. StuRat (talk) 14:37, 29 January 2013 (UTC)

Could you tell me your reasoning? I just google "kinect accuracy mm", https://www.google.com/search?q=kinect+resolution+mm Now, the first result (for me) is the stackoverflow question, where somebody says "Horizontal Resolution: 640 x 480 and 45 degrees vertical FOV and 58 degrees horizontal FOV. Simple geometry shows is about ~ 0.75 mm per pixel x by y at 50 cm, and ~ 3 mm per pixel x by y at 2 m." and then it says "Depth resolution: ~ 1.5 mm at 50 cm. About 5 cm at 5 m." 1.5 mm per pixel seems EXTREMELY precise, especially given that a SINGLE pixel's color would give away which key was pressed out of 104... and we probably have a lot more than one pixel. what I mean is, if the kinect were looking directly down, even if all the keys were "basically" the same color, wouldn't the ones that wasn't in the plane with the rest "obviously" be of a slightly different color? (I'm talking 'colors' as in the depth map that you can see, for example, in the pdf result on the google search, which is the second link). Thanks Stu for any reasoning or insight you have here. 178.48.114.143 (talk) 15:00, 29 January 2013 (UTC)

Having recently purchased a digital camera, I can tell you that the useful resolution is far lower than the nominal resolution. That is, reflections, dust on the lens, bad pixels, etc., make the result a lot worse than you might think (incidentally, the same is true of our own eyes, with our brains hiding all the defects in our vision, like the blind spot). To account for this, they generally don't have the software react to a change in a single pixel, but rather a large group of pixels. See computer vision for some of the issues encountered.

As for your idea of detecting the color change when a key is pressed, do you mean because there is a finger on it ? In this case, having a finger on a key doesn't necessarily mean it was pressed. Or do you mean that it will be in shadow ? This would only apply with little ambient light and most light coming from the side. Still, how could you see this with a finger in the way ? I imagine if you had a keyboard which would hold down the keys after you press them, and you remove your hand after pressing a key, and you set up the light conditions very carefully, and made the keyboard take up the entire image, then you could get it to work. Mapping the keyboard would be another issue, if you want to be able to use any keyboard, and they can have keys in slightly different places, with slightly different mappings. And, of course, there are far easier ways to determine which keys were hit on a keyboard. StuRat (talk) 15:18, 29 January 2013 (UTC)

Thanks. Actually by "color" i meant the depth map that is colored that you can see. I've given one example (in the top pdf result of my google search). so, in that sense do you think there would be a discernible difference in keys, if the kinect was as near as its range allows? A key seems like it's more like 1.5 cm across, giving you 10x pixels in each direction. That is, nominally, 100 pixels that would consistently a the same dropped plane versus all the other pixels. Wouldn't such a consistent change show up in the depth imaget? Look at our Wikipedia Kinect article: http://en.wikipedia.org/wiki/File:Kinect2-deepmap.png this picture is taken from quite far away, and not on the plane. If instead the kinect were looking more straight down and at the near end of its range, wouldn't it be able to pick up keys based on depth change? (By "color" I just meant the "colorized" visualization of the depth map, as you can see in the linked picture). Thanks. 178.48.114.143 (talk) 15:34, 29 January 2013 (UTC)

(Saying "the top Google result" is ambiguous, since different people get different Google returns from the same search.) That last image is rather poor quality. What are all the black areas ? Areas where the depth couldn't be determined at all ? StuRat (talk) 15:41, 29 January 2013 (UTC)

replied below, incl. exact title of pdf. 178.48.114.143 (talk) 02:17, 30 January 2013 (UTC)

Now for a practical suggestion. You haven't said quite what you are trying to do, but I imagine you want to get a Kinect to allow input of text information. In this case, a large printed grid with a subset of the keys would be easier for Kinect to read. You could make each key the size of a hand, and consider a key to be pressed when a hand covers it (although even this would be tricky, as your hand will also cover other keys on the way to and from the desired key). Perhaps you could have a box you place on a key to indicate that the key is pressed. A delay could be applied, and it would only register as pressed if the box remains there for a second. Once you get this to work, you could try smaller and smaller "keyboards" and boxes, until you find the practical lower limits. StuRat (talk) 15:50, 29 January 2013 (UTC)

Stu, the pdf is the one entitled "Accuracy and Resolution of Kinect Depth Data for Indoor Mapping Applications" ( https://www.google.com/search?q=kinect+resolution+mm ). see third page, color image on the right. Yes, black means that no depth information was available. See this faq: http://openkinect.org/wiki/FAQ . As for application, there is none: I would just like to know whether a kinect close to it can tell which key(s) are pressed on a keyboard and which are in the normal plane, using only the depth information. What do you think? 178.48.114.143 (talk) 16:20, 29 January 2013 (UTC)

The answer is probably "about as well as computers can do voice recognition". (While computers can recognize a small set of easily distinguishable words, getting them to recognize any word is another matter.) So, if the keyboard fills the entire field of view, and is aimed directly at the lens, and you find a way to keep the keys pressed after the hand is removed, then, some of the time, it could probably work, yes. However, as those large areas of black indicate, the depth perception is rather iffy in the Kinect. StuRat (talk) 02:29, 30 January 2013 (UTC)

Avoiding spam filters

For a new internet business we would like to send newsletters to a large number or recipients. How can we avoid being blacklisted for spam? Is it enough to use proper (html) style and to avoid certain words/phrases? Does the number of emails sent matter, e.g. if we send 1 million emails to gmail accounts, would we be considered spam? How important is sign-in/opt-out. Is it possible to send non-spam email without "opt-out" text or to send unsolicited email at all? bamse (talk) 14:01, 29 January 2013 (UTC)

What makes you think your newsletter isn't spam? 86.167.125.59 (talk) 14:23, 29 January 2013 (UTC)

If you're going to send a million unsolicited HTML emails with no opt-out, then lots of people will think it's spam. In general, it's very hard for commercial emails, even when sent with restraint and the best of intentions, not to be treated as spam. At one company I worked with, we sent 650 emails to current or past customers. Each email was customised with the person's real name and a snippet about how we'd already done business together (e.g. "since we worked together on the new business park at Fooville"). The emails were HTML with ASCII fallback, with only embedded images (no webside images). They were from the email address of a real person, who signed the message, and featured his full name, his direct phone number, and an opt-out link. We used none of the spammy tricks like misspelled words, foreign characters, or invisible text designed to skew statistical results. The email itself wasn't directly selling anything (just a company newsletter) and contained articles of genuine interest to the kind of person who'd be receiving it (it was a consulting company, so the articles were there to show how clever the consultants were). It got a very low SpamAssassin score. I checked with our ISP beforehand, and sent the emails over two days with a script that sent them one at a time (no CC, no BCC) with minutes between each send. I think we did everything we reasonably could to avoid tripping automatic spam filters or, much more importantly, pissing off people with whom we really wanted to do business. Still, out of that 650, 6 complained that it was spam (one, rather vociferously, by phone). When the aforementioned marketing guy (the emails' nominal sender) phoned around all 644 others to follow up (that was a fun week for him) about 10% denied they'd ever seen it, and only about 10% had paid more than glancing attention to its contents. -- Finlay McWalterჷTalk 14:48, 29 January 2013 (UTC)

Thanks a lot for this reply and interesting story. This will give us an idea of things to avoid and things to do. What do you think about services like mailchimp? Can this help to address some of the issues? bamse (talk) 15:57, 29 January 2013 (UTC)

So, 10% of the people who read the email complained about it? APL (talk) 23:34, 29 January 2013 (UTC)

One thing which there seems to be a lot of lack of understanding of in various companies is the Sender Policy Framework (SPF) - if you are sending email from a particular IP address, purporting to be from sendingaddress@mycompany.com - then you should make certain that you have the appropriate SPF record on the domain, as if this is wrong, it is pretty sure to get your email caught in people's spam filters. A lot of companies go for years blaming spam filters for losing emails, when in fact they have it in their own power to resolve the situation quite easily. Otherwise it can be quite difficult to predict the sort of heuristic algorithms employed by the larger email providers, but if you are effectively sending people email unsolicited and they don't like it they will flag those individual messages in their inbox as spam, and after this happens a few times, I think that provider is potentially going to start to flag ALL emails from that address. ---- _nonsense ferret 16:53, 29 January 2013 (UTC)

If you send out unsolicited emails without an "opt-out" option, and preferably an honest and reasonable explanation of why the user is on the list, then many of the recipients will report you for spam. Even if your email has the best of intentions. You probably won't get a chance to send out a second set of emails. APL (talk) 23:25, 29 January 2013 (UTC)

Website generating "screenshot"

The "Give us feedback" link at the bottom of the google.co.uk search results screen goes through a process whereby it generates a "screenshot" of the Google webpage and previews it prior to submission to Google. How does this work? I assume (certainly hope) that websites cannot truly take screenshots since that would seem to be a serious security hole. 86.167.125.59 (talk) 14:22, 29 January 2013 (UTC)

They just duplicate the process that renders a web page in your browser. They load the page's html content from the remote server and put it through a layout engine, except instead of showing the results on screen, they're stored in an image. That image is what you see. i kan reed (talk) 14:42, 29 January 2013 (UTC)

However, it seems to replicate local browser settings such as colours and fonts, as well as zoom level, exact scroll position and clipping, and so on. How would it know all that? 81.159.111.140 (talk) 18:07, 29 January 2013 (UTC)

Oh, that is odd. Are you using chrome? Maybe they had some hidden features they whitelisted google.com for? i kan reed (talk) 18:27, 29 January 2013 (UTC)

No, I'm using IE 9. 81.159.111.140 (talk) 18:35, 29 January 2013 (UTC)

Here's a presentation by a Googler about it. -- BenRG (talk) 18:53, 29 January 2013 (UTC)

Thanks. Most of that is a bit over my head, but is it saying that the page is completely re-rendered using JavaScript? So the Google web page has access somehow to all those local broswer settings like font, colour, zoom and stuff? Another thing that I don't understand is that both in that presentation and in the actual web page widget there is some stuff about "privacy" and blacking out parts of the page you don't want Google to see. But surely Google already has access to everything that it's allowing you to black out, so what is the point? I suppose unless it just means it doesn't "save" blacked out parts to its comment log, or make them available to the Google employees who deal with that aspect? 81.159.111.140 (talk) 20:38, 29 January 2013 (UTC)

It's presumably the case that they expect you to give more personally identifiable information (name, contact details) when submitting feedback than you give anyway when you do a search. Thus, it is important that you don't link the two. - Jarry1250 ^{[Deliberation needed]} 00:05, 30 January 2013 (UTC)

Sorry, I don't understand how that explains why you are prompted to black out parts of something that must already be available to Google so that it won't be available to Google. 81.159.111.140 (talk) 01:39, 30 January 2013 (UTC)

Assuming good faith, which is not something I commonly do about Google, we could speculate that not everyone inside Google has access to all information. So, if you file a feedback submission, the person who reviews that submission might not have internal permission to browse everything that the company-as-an-ensemble knows about you. In fact, this is their public version of their employee-code-of-conduct, which discusses internal disclosure of user-data on a need-to-know basis. Nimur (talk) 05:52, 30 January 2013 (UTC)

Hello I Need Your Help (Google search result special pages)

Hello.i need Your help about article of Leo Jee i want to now: when i search leo jee on Google.com i find there leo jee wikipedia but not like this(u can see photo for example) i mean discussion on right side of page Google.com.

plz help me what i must to do that Leo Jee search like this ... thank U so much!!!--AlexSpancer (talk) 14:24, 29 January 2013 (UTC)

Google seems to be setting up special pages for certain topics, like the Eric Saade one you show here. While they use some content from Wikipedia, we can't control which topics get these special pages and which don't. I'm not sure if they've published their criteria for setting up a special page, but I imagine the number of searches for that topic must figure in. StuRat (talk) 14:34, 29 January 2013 (UTC)

Thank U so much !!!--AlexSpancer (talk) 14:37, 29 January 2013 (UTC)

You're welcome. Also, I added to your title to include some indication of the actual question. Please try to do this in the future, so we know what it's about without reading through the entire Q. StuRat (talk) 14:55, 29 January 2013 (UTC)

Win32s

Win32s allowed some 32-bit programs to work on a 16-bit OS, but not all because it was only a "partial implementation of the Win32 Windows API". Technically, could the whole Win32 API have been implemented, allowing it to run all 32-bit programs? Hypothetically, could a 64-bit environment be created on the 16-bit OS using the same thunks technique? 92.233.64.26 (talk) 15:51, 29 January 2013 (UTC)

Hypothetically, one could emulate all of the necessary hardware and software, and if we take the mental exercise to its natural limit, one could implement a Universal Turing machine. In practice, this might require writing a lot of software that doesn't yet exist for the 16-bit platform in question: things like a truly infinitely-scalable file-system (or at least, a really really large scalable file system). And, whenever software emulation must use many layers of indirection to perform the same task that other hardware does very quickly, the execution times become prohibitively long, often requiring orders of magnitude longer to complete. Nimur (talk) 16:11, 29 January 2013 (UTC)

Windows 95 was basically Windows 3.1 with a more complete version of Win32s built in, so yes. Of course, 64-bit x86-compatible processors didn't exist then, but I see no reason you couldn't write a 64-bit DOS extender. Multitasking DOS programs, as Windows 3.1/9x did, would be more of a hassle because there's no support for V86 mode in 64-bit long mode. However, the 64-bit versions of VirtualBox and friends do it, so it is possible. -- BenRG (talk) 17:39, 29 January 2013 (UTC)

*Manually* factory-restoring Windows 7 from recovery discs

Hello! I'm restoring my HP laptop from Windows 8 Preview to its original Windows 7 OS. I put in the first of the recovery discs I had made when I first bought the laptop, booted into it, and selected the factory-default option, while opting to do the recovery from the media, not the hard drive partition (not thinking it would make any difference).

Now the problem: the HP recovery software successfully started off reformating my drive, and then as soon as it started writing the old image, it immediately failed with 0xe0ef000e error, which is a generic error that it couldn't read properly from the dvd, or write properly to the drive. So I booted into the latest distro of Ubuntu with toram so I could use the cd drive, put in the recovery disc, and can read the entire disc, no problem. gparted can play around with the hard drive fine, but of course all the partition data has been lost from the recovery utility. Looks like windows includes md5 checksums with images parts on the dvd, and I verified a few arbitrary ones without any problems. It really seems to me it's a problem with the recovery software.

And now the proposed solution, with the full arsenal of Linux and GNU software utilities, how can I manually restore the image of the original windows split across the recovery dvds? There has to be a way with gparted and dd. Perhaps the recovery images are encrypted, though; then again, since Windows is so obsessive with checking that the hardware it's running on is all OEM, maybe not. Thank you for any help or useful links; everything I've been able to find via search engines assumes competent system-recovery software from the factory.--el Aprel (^facta-_facienda) 23:44, 29 January 2013 (UTC)

Resolved

Never mind. Since the recovery program failed before it could do any overwriting of disk data, I was able to recover the recovery partition with the TestDisk program. Word to the wise: backup the recovery partition itself with, e.g., dd and do your own verification. Don't rely on factory software to create functional optical recovery media.--el Aprel (^facta-_facienda) 04:15, 31 January 2013 (UTC)