Computing desk
< March 13	<< Feb \| March \| Apr >>	March 15 >

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.

March 14[edit]

Sign language annotation for sound files[edit]

For File:Sextortionagent_interview.ogg I was wondering if the software supports a sign language annotation for the sound files, indicating what signing would be done if one used American Sign Language. I think that having sign language for audio files could be another way to add support for deaf users (some instructional videos and news broadcasts use sign language) WhisperToMe (talk) 08:01, 14 March 2013 (UTC)[reply]

That seems unnecessarily complicated. I guess there could be better encouragement and support for linking to a transcript if it exists when a sound file is referred to, in this case there is a transcript at File:Sextortionagent_interview.ogg#Introduction and transcript (original English). The place to put an idea like that would be I'd guess WP:VPR. Dmcq (talk) 09:46, 14 March 2013 (UTC)[reply]

Alright, I'll use the village pump. Yeah, there are some videos that use sign language instead of subtitles even though subtitles seem simpler... WhisperToMe (talk) 14:16, 14 March 2013 (UTC)[reply]

motion detection for computers[edit]

how would I go about setting up a program to detect motion through a web-cam or similar and use that to control the movement of my computer's mouse pointer? and would I need any more advanced hardware or would my current cheap, simple camera be good enough to interpret 2D gestures across the screen?

Kitutal (talk) 13:01, 14 March 2013 (UTC)[reply]

Computer vision is not an easy problem, and there are no "cookbook" solutions. There aren't a lot of great software programs that work "out of the box," either.; though there are plenty of tools. You can make this problem a little bit less insurmountable by:

using an existing high-level language like MATLAB, or a powerful API, like OpenCV, to abstract implementation details so you can focus on algorithms
simplifying your scene - by using controlled scenes, good lighting, and high-quality cameras, or using standard tricks like brightly- and uniquely-colored targets with simple geometries.

Here's a link to lecture notes from Introduction to Computer Vision at Stanford. A few years ago, some teams made a "Kinect"-style motion tracker for the class project. I'll see if I can dig up a video or paper. Nimur (talk) 14:58, 14 March 2013 (UTC)[reply]

Are you asking about finding an existing program, or writing your own ? In either case, I'd be sure to have a sensitivity setting. For some applications, like turning on lights for 5 minutes when motion is detected, a low threshold is fine, while for other applications, like sounding an alarm, that would be bad. StuRat (talk) 15:38, 14 March 2013 (UTC)[reply]

Another link, "Introduction to Computer Vision", which has more of its lecture-notes available to the general public. Nimur (talk) 17:29, 14 March 2013 (UTC)[reply]

I was hoping for some existing software and/or hardware, otherwise I would need to look up some idea of how to go about creating the stuff myself, or so I can show friends what to do to help. I thought it best to rig up something simple that didn't work wonderfully and attempt to refine that through various methods, including better programming, better camera and the tricks mentioned above, rather than trying to do the whole thing perfectly first time around. Kitutal (talk) 12:06, 15 March 2013 (UTC)[reply]

Here's another slide-deck, Introduction to OpenCV. And here's a blog-post including a video of one of our classmates' final project, two-camera motion-capture system used to control a PUMA robotic arm. As you can see, they used bright targets (they taped orange and red color swatches to their hands) to make their program much simpler. For perspective, this required three graduate students and several weeks to design, implement, and tune. As above, the tools exist, but few are "ready to go" out of the box. Nimur (talk) 23:21, 16 March 2013 (UTC)[reply]

OK then, where can I go to learn everything I would need to know to do something like that? Kitutal (talk) 23:17, 17 March 2013 (UTC)[reply]

Stanford Engineering Everywhere has several classes that are directly relevant, including the Introduction to Robotics course. MIT's OpenCourseWare has dozens of classes that are relevant, including Machine Vision. Nimur (talk) 17:38, 18 March 2013 (UTC)[reply]

casses[edit]

is there any computing classes on the shanill road belast 13 — Preceding unsigned comment added by 62.253.58.255 (talk) 15:40, 14 March 2013 (UTC)[reply]

Belfast Metropolitan College has ICT classes which are held at various campuses around the city, including Chichester Avenue and Whiterock Road. -- Finlay McWalterჷTalk 15:49, 14 March 2013 (UTC)[reply]

Why can't you root iOS?[edit]

I have an Android and I have come to appreciate the great community involvement in the project in creating custom ROMs that are often much faster and feature packed than both Stock Android and various skins such as Touchwiz. If I understand correctly, the reason that there are no custom ROMs for iOS is because you can't root the devices? Is this purely a technical reason in the sense that developers have been unable to root the phone?

Would it be safe to say that if Apple opens up the phone and allows third party ROMs that there can be a lot of speed and performance improvements from third party devs like Android? Acceptable (talk) 16:15, 14 March 2013 (UTC)[reply]

A fully jailbroken iOS device allows the user the same totality of control that a rooted Android phone does, so the premise of your question is false. As a practical matter, developing an alternate firmware to run on an iOS device is much harder than for Android, because much of the software on a typical Android phone is open source and so available for re-compilation and alteration, whereas essentially none of iOS is. Perverse as it may seem, it's possible to run Android on jailbroken iOS hardware: see e.g. http://www.idroidproject.org, although that port isn't complete yet. -- Finlay McWalterჷTalk 16:45, 14 March 2013 (UTC)[reply]

Why can't developers just "hack" the non-open source software of iOS? My understanding of operating systems is limited, but is it not just a bunch of files? If it's because the OS prevents users from editing these files, can developers not just brute force or break the encryption or something? Or is this too technically challenging to do? I guess, in other words, if I bought an iPhone, isn't the code and all the files for the operating system already on my device? Acceptable (talk) 22:42, 14 March 2013 (UTC)[reply]

Jailbreaking breaks the encryption. But iOS, and its associated apps, is chiefly written in Objective-C (with some C), a compiled language. So what's on the phone is a bunch of Mach-O format binary executables, where the readable Objective C code has been compiled into thoroughly unreadable machine code. Patching that is a hideously laborious process, and decompilers produce fairly horrible results that aren't much better. Beyond removing some of iOS's security features, it's not worth the bother exploring all its depths. -- Finlay McWalterჷTalk 22:54, 14 March 2013 (UTC)[reply]

Oh I see, is this similar to Adobe Flash whereby the files that you edit on is the .FLA while the compiled version is the .SWF? Theoretically, would it be possible for an Apple engineer (who isn't afraid to be sued) to just load all the master files onto a USB and upload it on the Internet? Acceptable (talk) 22:59, 14 March 2013 (UTC)[reply]

Yes, and yes. -- Finlay McWalterჷTalk 23:06, 14 March 2013 (UTC)[reply]

I have a sneaking suspicion that even if the source code and tools and instructions necessary to engineer an operating system were all made available at no charge and with a fairly liberal license, very few people would actually use them. It probably wouldn't even light up the front pages of the rumor-mill websites if such material was surreptitiously leaked to the internet in a place nobody would look. This is probably because, while "jailbreaking" is frightfully exciting, actually programming at the lower levels of an operating systems is very boring. But, I don't know that much about programming iPhones; perhaps somebody might come along and correct me. Nimur (talk) 03:45, 15 March 2013 (UTC)[reply]

Uh, okay. So the only specific page you linked describes how to build clang, which is a compiler. Which is a useful thing to have, I suppose, if you're going to try to recreate an entire operating system from scratch, but... -- BenRG (talk) 04:23, 15 March 2013 (UTC)[reply]

I think Nimur is trying to assert that because Apple has released the Darwin kernel source code, the fact there is only limited interest in this demonstrates most people have little interest in the source code. However many would suggest the more likely interpretation is that a kernel does not make an operating system and with the large number of *nix kernels out there (including the ones Darwin was partially based on), it's not particularly surprising that few people were interested in remaking an operating system from the what Apple released (when they felt like it) as OpenDarwin and PureDarwin have shown. In any case, I don't get the relevance of this the OP's question. The OP's question was not why doesn't someone make a new OS for iOS devices (which a operating system based solely on the Darwin kernel would be) but why don't people customise iOS (the existing OS) as much as Android. The fact they can't easily do so because Apple has not released the source code for the operating system remains true whatever bits Apple have released. In fact, AFAIK, the amount of source code available for the iOS builds are minimal (our own article says 'except for the ARM variant, which has not been released in any form separately from iOS'), not even close to the whole kernel, hardware support etc. The fact it's a fork of something that has been partially released isn't likely to help much. I strongly suspect the Android option FM mentioned is still potentially far easier then trying something with what Apple released. And there's no particular reason why they'd want to, few people think there's something particularly revolutionary about the Darwin kernel, it's all the interesting bits of iOS that Apple did not release that people care much more about. (Of course one of Nimur's assertations is probably correct, there is little interest in doing much low level programming of the operating system but as the OP's original question has shown. There is a fair amount of interest in customising it in various ways as the very large number of highly custom Android firmware mods have shown. In fact many of those cases where e.g. sometimes the developers have had to at least partially reimplement hardware support due to the company not releasing it has shown people are even willing to go down to the nitty gritty when necessary. But as I said earlier, people need a lot more then what Apple released to be able to do this.) Of course when Apple does release something that's useful (whether entirely their own work or not), people may use it as WebKit showed. Nil Einne (talk) 07:22, 15 March 2013 (UTC)[reply]

Nil, have a paragraph break, on me: ¶. :-) StuRat (talk) 14:55, 15 March 2013 (UTC) [reply]

Windows 7 Shutdown Strategy[edit]

I'm looking for a way to use the shutting down of my laptop to enforce a solid sleep schedule. For a while, I was using a windows scheduled task to start up shutdown.exe /s on a five minute repeat for an hour. This was fine until I realized I the task could be disabled. I am currently trying out the 30 minute random delay, however that degree of variation is not ideal, and it would be better if it were narrower.

Ideally, I'd like to be able to have a scheduled log off time at night that can only be edited during the morning/day, though I don't know if windows or any other program would have that functionality. I also briefly considered setting up a "on task creation/modification" trigger to the task, which would likely prevent any changes to the task without getting the intended result, but I am wary of setting up scheduled tasks that I cannot change (such as a shutdown at log on task, I don't know how that would be bypassed). If anyone has experience with this sort of thing or any suggestions, I would greatly appreciate it. Thanks. Sazea (talk) 16:23, 14 March 2013 (UTC)[reply]

I'm not entirely sure I understand what you're trying to achieve. It sounds like you're trying to set things up so that the computer will shut off at night (so you're not awake until 4am playing Skyrim or something), and you're worried that your 4am self will just disable the software your daytime self has installed? The trouble is that it's the same administrative permissions you need to install software like this as you need to uninstall or disable it. The best you can do is set yourself up as a non-admin account, setup the schedule as admin, and then have some other person change the admin password so that they know it and you don't. You're probably better off with some sleep-cycle management software, if this is a real problem for you (I know people who use SleepCycle on iOS, but I've no personal experience of it). -- Finlay McWalterჷTalk 16:54, 14 March 2013 (UTC)[reply]

Computer solutions may not get at the root of the problem, which is why you aren't sleepy enough to want to go to sleep when you should. I suggest eliminating all caffeine after noon (or ideally eliminate it completely) and doing some heavy exercise about an hour before bedtime. Also, avoid computer games where you are penalized for signing off (say if competitors who stay online rank up and you don't). StuRat (talk) 17:16, 14 March 2013 (UTC)[reply]

why should we pass object of the class implemented from runnable to constructor of thread class?[edit]

Hi my doubt is in java. My doubt is on creating thread by implementing runnable interface. Please see the below programme.

In below programme I have implemented the class countdowneven from Runnable.
In main method I am passing the object of countdown even to constructor of thread class.
Why should we pass the object of countdowneven to constructor of thread class?
What operations will be performed by Thread class constructor on object of countdowneven?

. class thread
{
public static void main(String args[ ])
{
Thread ce=new Thread (new countdowneven());
ce.start();
}
}
class countdowneven implements Runnable
{
public void run()
{
for(int i=10;i>1;i=i-2)
{
System.out.println(i);
}
}
}

Output:
10
8
6
4
2
— Preceding unsigned comment added by Phanihup (talk • contribs) 18:38, 14 March 2013 (UTC)[reply]

We don't do homework for you, sorry. Those are clearly homework questions. Would you like to know more about Threads? If you're specifically wanting to figure out threads in java, I recomend reading the documentation in the java SDK i kan reed (talk) 19:12, 14 March 2013 (UTC)[reply]

Launching the wrong browser[edit]

Evert time I have tried to launch Google Chrome in about the last week, my laptop launches IE instead. This applies to all short-cuts, icons, files and lauching it directly from the C:// drive. I eventually got it to launch by running Chrome as an administrator. How can I fix this?--Gilderien Chat|List of good deeds 19:02, 14 March 2013 (UTC)[reply]

How are you starting Chrome? If it's by clicking on a .html file, or entering a URL in the start box, then Windows will run the default browser. There should be an option in Chrome to make it check if it is the default. and if not, make it, it. Unfortunately, I don't use Chrome, so I don't know where the option is. CS Miller (talk) 21:10, 14 March 2013 (UTC)[reply]

I'm clicking on the Chrome icon, and it is my default browser.--Gilderien Chat|List of good deeds 22:35, 14 March 2013 (UTC)[reply]

There has been some suggestion on other sites that it is likely malware. This link recommends running malware bytes (www.malwarebytes.org/). Sazea (talk) 23:34, 15 March 2013 (UTC)[reply]

Ok thank you. McAffee seems to have removed it when I ran a full and intensive scan, reporting 5 trojans as part of it. Not really sure how the writer would gain from writing that but anyway ... --Gilderien Chat|List of good deeds 23:49, 15 March 2013 (UTC)[reply]