Wikipedia:Reference desk/Archives/Computing/2011 November 19

Computing desk
< November 18 << Oct | November | Dec >> November 20 >
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.


November 19

edit

Could a PC be used as a radio receiver/transmitter?

edit
  Resolved

Software such as NI Multisim can be used to emulate electronic components. Could a standard laptop with its built-in Wi-Fi antenna be programmed to function as a radio receiver or transmitter? --Codell (talk) 02:06, 19 November 2011 (UTC)[reply]

You might be interested in Software-defined radio. -- Finlay McWalterTalk 02:21, 19 November 2011 (UTC)[reply]
I managed to find GNU Radio through that article which seems to be what I was looking for, although I wonder if software for Windows exists that can receive radio without needing external hardware. --Codell (talk) 02:51, 19 November 2011 (UTC)[reply]
This page references a number of projects, some free and some commercial, one or two for Windows but mostly Linux or BSD, which unlock the capability of wireless cards to some degree. If manufacturers distributed more broadly-capable drivers, they would have to endure greater FCC scrutiny, and any added sales just wouldn't pay for the resulting costs. I think Cisco has a Wifi interference detector with a near-standard chipset and some funky drivers that does something like this, but it is expensive, even by comparison to the more capable hardware GNU Radio works with. Besides, there's not that much interesting stuff going on in the bands WIifi cards operate on anyway- you might be able to snoop on your neighbour's cordless phones, but without extra hardware, you're not going to have access to much. Nevard (talk) 08:54, 19 November 2011 (UTC)[reply]

Where can I get affordable Maytag Gas Wall Oven Parts?

edit

Does anybody know of a place where I can get affordable parts for my wall oven? My less than 6 year old oven had several "faulty codes" on. I replaced the ignitor and the temperature sensor that cost me over $100.00. Now there is another faulty code on and the technician said is the hardware and watchdog circuits and that I need to replace the clock assembly (also called the ERC) (Honestly, I don't have a clue and nothing about this makes any sense to me)

I already went online and the cheapest cost almost $250.00. Spending so much money on a not so new appliance sounds absurd and it would probably make more sense to just buy a new one. So, I am asking if anybody has some advise or know of a place when this part can be found used or new before my oven ends up in a landfield. Thanks! 66.108.223.179 (talk) 02:50, 19 November 2011 (UTC)[reply]

Iphone's bluetooth

edit

Hi, I would like to know why an Android cannot transfar files to Iphone, and which devices can. Exx8 (talk) 10:51, 19 November 2011 (UTC)[reply]

They can. Some apps might make it easier. [1] ¦ Reisio (talk) 17:18, 19 November 2011 (UTC)[reply]


iPhones cannot send or receive files over Bluetooth, except to other iOS devices, AFAIK. Apple, like everything else, crippled Bluetooth. Until iOS4, iPhones and iPod touches didn't even have AVRCP, when older devices had it out of the box. User:Bodman456 | Come talk to me or ask me a question! (I don't bite ;D) 23:10, 19 November 2011 (UTC)[reply]

Overclocking and stability

edit

Computers are physical systems subject to wear, thermal degradation, and intervening factors like cosmic rays and electromagnetic interference. Nonetheless they perform remarkably well. Systems remain stable for very long periods of time. We can check system stability by running computational exercises where the expected answer is known already. Presumably if one runs any computationally intensive task for long enough, eventually an error must occur due ultimately to the physical limitations of the machine. Existing systems are often so reliable that the mean time to such an error may be months or years. I'm curious if there is any hard information of the typical reliability of recent CPUs in this way? For example, something like if you have 1000 CPUs computing MD5 hashes continually for 2 months you would expect 1% of them to make at least one error in one hash over that time.

Assuming such quantifiable data does exist, I'd also like to know how CPU reliability is impacted by overclocking. Obviously overclocking too much can lead to scenarios where a CPU fails almost immediately and no useful work can be done, but I'm wondering about the rare events where everything appears fine for hours or days before a small glitch. Can we measurably demonstrate that overclocking increases the chance of small glitches on systems that initially appear stable? Dragons flight (talk) 11:21, 19 November 2011 (UTC)[reply]

Soft error appears to have some very good external links measuring actual reliability vs. cosmic rays and radiation effects and such. Nothing about overclocking, though. Comet Tuttle (talk) 18:17, 19 November 2011 (UTC)[reply]
Most people who know what they are doing when overclocking the CPU will test for 'instability'. For those not experienced, they may think instability refers solely to the computer crashing or freezing but realisticly you should only get that when you are beginning to overclock and trying to work out the limits, at least if you aren't some sort of 'extreme edge' overclocker. (There are those who try to get records and only hope is the computer lasts long enough to get a verified screenshot of it running at the speed.) It's of course far easy to push the CPU until it's clearly unstable then ramp down then it is to test everytime for long term stability and slowly ramp up even if the later is technically safer.
Once you've gotten past that point, it gets down to the nitty gritty of stress testing#hardware. For the CPU, this will often mean running something like Prime95 or OCCT or IntelBurnTest or something of that sort for hours to see if any errors (either by comparing the results against known, pre-computed results or doing the same thing a few times and seeing if the results are the same) are detected. (Nowadays memtest86 should probably also be part of the mix since even if you aren't overclocking the RAM, the memory controller is built in to all modern x86 CPUs.)
While it depends on the user, on a personal note I've generally used 24-48 hours error free (often multiple sessions with at least all 3 types of Prime95 test) as a minimum before I'm willing to say the computer is stable. Although I admit, past 24 hours often means I use the computer during testing for general stuff like browsing, which isn't as bad as it sounds, at least for Prime95 if you don't chose 'Blend' or some other test that uses all the RAM the fact the test is usually on very low priority by default means it's not that noticeable for such low level usage, however it must reduce the chance of an error being detected.
Anyway to your point, most people reduce overclocking or bump up the voltage after any error but I believe in the past I've repeated testing more then once after getting an error in a few hours and got another error at some stage every time (the precious time tends to vary quite considerably of course) but received no errors after having reduced the overclocking. This is even easy to show further past the borderline, sometimes you get an error within 10 mins - 1 hour or so and you will find this happens fairly consistently, but if you reduce the overclock or the CPU is not overclocked this doesn't happen.
I'm sure you must be able to find some more systematic analysis of this in some overclocking forum. Although as said it depends on the invidual, I have heard of some people who ignore a single error after a long time and I think quite a lot test less then I do. Of course if you only ever got an error once and can't get it to happen again, it's difficult to be sure it had anything to do with overclocking, one of the reasons I generally hope if I do get an error, I can have it happen again but not happen when I've either reduced overclocking or stopped overclocking (statistically I doubt I test enough that it's really that meaningful and there's also the risk of confirmation bias but on a human level it provides at least some reassurance).
Of course this doesn't mean such tests detect all errors, a good case in point, I believe when IntelBurnTest (which I've never really used as it came out after my last major overclocking mission and also I haven't had an Intel CPU In a while although I believe by now it's supposed to be quite good for AMD CPUs as well) came out, some people found their previously seemingly 'stable' CPUs after hours of Prime95 would start to fail in some way after a short time with IntelBurnTest.
Nil Einne (talk) 03:20, 24 November 2011 (UTC)[reply]

Windows colour editing

edit

What I used to really like about Windows98 was the way you could bring up the customise window and change the colour or size of anything to make the computer look just the way you wanted. Having proven unable to get 98 on my new laptop, I am wondering if there is any way of getting a similar function on Windows 7. I have managed to find the option to return it to the original blue and grey, but not any other colour options.

148.197.81.179 (talk) 11:58, 19 November 2011 (UTC)[reply]

Right click on the desktop, select "Personalize" from the menu that comes up, click on the icon that says "Window color," and you get a screen which is more or less the same set of options as was in Windows 98. --Mr.98 (talk) 15:47, 19 November 2011 (UTC)[reply]

Ah, I've found it now, I have to set it to the classic version first, then after that go to the window colour options. Thanks. 148.197.80.214 (talk) 20:35, 19 November 2011 (UTC)[reply]

Word 2007 - Two problems

edit
  Resolved

I have two problems with Word 2007, which I would like to sort out.

  • When I start a document, the default spacing of lines is set to 'Normal', which leaves a blank line between lines. I want it to be set to 'No spacing' as default. How can I do this?
  • Whenever I enter a date, I am prompted to press Enter to change it to today's date (in a different format). I don't want today's date, and I want to press Enter anyway so I can start a new line, so this is troublesome. How can I turn this feature off?

Cheers KägeTorä - (影虎) (TALK) 16:39, 19 November 2011 (UTC)[reply]

1. Whenever you start a new normal document, it's based on a template called "Normal.dotm". Choose File -> Open, click "Templates" on the left, and choose "Normal.dotm". Select All, change the paragraph attributes and line spacing attributes and default font however you want, and save over Normal.dotm. And let me know if this works! I'm too chicken to do it on my machine. You should save a copy of Normal.dotm before you begin, of course, in case something gets messed up.
2. Haven't found an answer yet, but Office button -> Word Options has an Autocorrect Options button in the "Proofing" screen ... which to my surprise doesn't let you turn off this date feature. For what it's worth, my Word doesn't seem to do this. What exactly are you typing (in what date format)? Comet Tuttle (talk) 17:57, 19 November 2011 (UTC)[reply]
Thanks for that. Please bear in mind, I am using Word 2007, and your first solution is not possible. There is no 'File' and when I click open (from the Office button), I just get a list of files in my Documents folder and no option to open Templates. Clicking 'New', however, gives me the 'Templates' section on the left, but Normal.dotm is not there (I know where it is in my AppData folder, as I have had to locate it in the past). As for the date format, I type 1st November 2011, for example, and it will offer to change to 2011-11-19 (today), which, bizarrely, is Japanese date format. The language is currently set to English (United Kingdom). KägeTorä - (影虎) (TALK) 18:15, 19 November 2011 (UTC)[reply]
I'm using Word 2007, too — sorry for saying "File"; what I actually did when testing it was to hit alt-F, which in my mind will always mean "File". So what I did is: Alt-F, Alt-O (for Open), click Templates. This brings me to username->AppData->Roaming->Microsoft->Templates and it does contain Normal.dotm. Did this work for you?
I just verified the same behavior you described on the date! I normally never type dates like that, because I am an American, so never saw this. By the time I get to "November 1," then that comma cuts off the annoying date autocorrect. Anyway, you could have found the fix by doing what I did, which is to google "Word 2007 turn off date autocorrect" where the second entry was this not-super-intuitive solution which did work for me:
To turn off date autocorrection in Word 2007, do the following: Press Alt+F11 to display the Visual Basic Editor. Press Ctrl+G to activate the Immediate window. Type
Application.DisplayAutoCompleteTips = False
and press Enter. Close the Visual Basic Editor.
Cheers — Comet Tuttle (talk) 18:24, 19 November 2011 (UTC)[reply]

Brilliant! Both work perfectly! Excellent, thanks! KägeTorä - (影虎) (TALK) 22:08, 19 November 2011 (UTC)[reply]

Kompozer

edit

Kompozer doen't work and blames xpcom.dll saying entry point not found. Recently had to replace myy hard disk (C:) due to warnings and crashes. Both Kompozer & xpcom.dll are new. Kittybrewster 17:31, 19 November 2011 (UTC)[reply]

I have added a section header for this question. Comet Tuttle (talk) 17:50, 19 November 2011 (UTC)[reply]
I don't know anything about Kompozer, but when you wrote that you replaced your hard disk, what exactly did you do? I am betting you did not (a) cloned the hard disk with software like Norton Ghost or Clonezilla, and that you did not (b) format the new hard disk, install Windows, and then use the Kompozer installer to install the software afresh on the new hard disk. I'm betting you (c) formatted the new hard disk, installed Windows, and dragged your files from the old hard disk to the new hard disk? If so, that is the problem; almost any software you install on a Windows system sets some values in the Windows registry, which do not get copied over if you just drag the files from the old hard disk to the new hard disk. Do a reinstall of Kompozer and that should fix it. If the installer allows it, run it and choose a "Repair" install. Back up everything first of course in case a badly designed installer chooses to overwrite some of your work files. Comet Tuttle (talk) 20:19, 19 November 2011 (UTC)[reply]
I have a world book backup, backing up my stuff pretty often. It was installed at a distance using logmein. I installed the c drive by handing it to a local shop. Kittybrewster 21:31, 19 November 2011 (UTC)[reply]
Did you pay the local shop to install Kompozer? If so, take it back to them. Backups work for data files, but program files need to be installed, as explained above. Dbfirs 09:21, 20 November 2011 (UTC)[reply]
Not specifically. But thanks to the ref desk, Kompozer now works.Kittybrewster 10:37, 20 November 2011 (UTC)[reply]

ICT4D

edit

ICT4D- Who proposed it and why and when — Preceding unsigned comment added by 120.151.181.98 (talk) 22:30, 19 November 2011 (UTC)[reply]

Why don't computers have money inserters, so we order online with physical cash?

edit
 
Imagine this coonduit starting from behind a desktop computer (or a laptop if it's docked to the tube's house station first), in every home.

Look at this- these are conduits leading from the house, to many parts unknown.

The idea here would be to insert coins through coin slots, and dollar bills through bill inserters, and into a waiting capsule, in order to send the capsule through a conduit to the payment recipient. (If the online store doesn't take e-checks, I suppose the checks can be inserted into the capsules as well.)

If we can have many conduits/pipes/tubes/et al. originating from every home to parts unknown thousands of miles away, what's wrong with doing the same for the sake of home-shopping with physical cash?

Besides, a child can't use plastic until their teens, but even kindergartners can buy with physical cash, so online toy, novelty and game stores will see a spike in sales if these conduits come along.--70.179.174.101 (talk) 22:54, 19 November 2011 (UTC)[reply]

The tubes at the bank just connect to the inside of the bank. The distance is short, and no switching mechanisms are needed.
Building infrastructure to connect to every single house is really expensive. A typical building will hook into networks for water (and also sewage), power (electricity/maybe gas), and information (phone/cable TV/Internet service). Each one of those is something that people are happy to pay dozens of dollars per month for (possibly indirectly, though taxes). Importantly, pretty much everyone is willing to pay for this, so it's practical to dig up every single street in order to put the infrastructure in. A pneumatic tube network would be cool (I'd think more for delivering groceries than dealing with money, since electronic mechanisms for money are pretty good already), but it's probably not worth the immense cost of construction and maintenance, especially since moving macroscopic objects through tubes is going to be more failure-prone than, say, applying current to a wire. Paul (Stansifer) 23:14, 19 November 2011 (UTC)[reply]
Parents could give their children an account with a PayPal-like service (not one hooked up directly to their bank account), and allow them to deposit money into it and spend money from it. But online businesses in the US are likely to refuse service to anyone under 13 due to COPPA, so the whole thing may be moot. Paul (Stansifer) 23:14, 19 November 2011 (UTC)[reply]
That would be an enormously expensive undertaking to supply every house with a pipe like that. And you'd most likely have people digging up the ground where the pipes are trying to intercept the money. It's simply not economical enough to warrant doing 82.43.90.142 (talk) 23:18, 19 November 2011 (UTC)[reply]

Operated two ways it would eliminate the need to step away from the computer at mealtime, as here.

Yet, we do NOT have an article on the Alameda-Weehawken Burrito Tunnel??? It's such a marvel of engineering, we need to put one up! — Preceding unsigned comment added by 70.179.174.101 (talk) 23:49, 19 November 2011 (UTC)[reply]
Let's imagine how this would work.
1. I put in my money.
2. The money thus travels to... somewhere? A bank? A central repository? The business in question? Somewhere. Not where I am.
3. This takes.... awhile. Because paper in tubes doesn't move as fast as electrons do. Ho hum. I wait, oh, I don't know, awhile. Now let's pretend, just for fun, that this money tube is set up so that it can't get clogged (not easy for such distances) and somehow we've managed to get over the fact that we're not likely the only one using the system so unless we have dedicated tubes from each house to the place (a lot of tubes), we're talking about a scheme where multiple dollars have to somehow share the tube at once, not get caught, and not get confused.
4. Eventually the bank or whatever gets the money, scans it in, determines it is not counterfeit, credits the account. Let's imagine this can happen very fast unlike in real life, where banks have actual humans verifying cash deposits before they are credited to your account.
All of this is... well, just very silly. What problem does it solve, again? Oh, children being able to buy stuff on the internet. Parents are going to be clamoring for that — they'd love for their children to just be able to send money out of the house at rapid speed. That's definitely worth the cost of this silly money-in-tubes scheme. --Mr.98 (talk) 23:35, 19 November 2011 (UTC)[reply]
Also, let's say I (in the UK) wanted to buy something from Japan. Even if my money travelled down this ridiculously long tube from my house to the store at the speed of an average Boeing 757, it would still take around 14 hours for it to get there - and then another 14 hours for me to receive any sort of acknowledgement of payment. Besides, kindergarteners don't tend to buy toys or games. They are bought for them, by their parents, whether online or physically at the store. Why? Because kindergarteners don't earn. KägeTorä - (影虎) (TALK) 01:04, 20 November 2011 (UTC)[reply]
I would have to assume that the allowance is a foreign concept in your culture. Parents, to varying degrees, pay their kids to do chores and their best on their academics. Children do earn after all. It's the parents' decision on whether they earn money or something else. --70.179.174.101 (talk) 01:34, 20 November 2011 (UTC)[reply]
We call it pocket money, but when I was a kindergartener (which the OP specifies), it was just enough money to keep me going over the week with sweets, etc., and certainly not enough to buy myself a game or toy from a specialist game/toy store. Without supervision. At the age of 5. :) KägeTorä - (影虎) (TALK) 02:14, 20 November 2011 (UTC)[reply]
My son wants me to get paid in Legos to cut down the time between me giving him allowance and then waiting to go to the store to buy Legos. -- kainaw 01:36, 20 November 2011 (UTC)[reply]
In one of many possible post-apocayptic futures, Legos will be the only viable worldwide currency. --Mr.98 (talk) 12:27, 20 November 2011 (UTC)[reply]

Why would someone put his money in this thing? It's true that cities, or at least boroughs could have some sort of pneumatic tubes, but the idea somehow didn't pan out...At least not at a a level higher than a building. I also ask me why, since we invest in many other large expensive infrastructure ... Quest09 (talk) 22:14, 20 November 2011 (UTC)[reply]

The router logistics are just not physically workable on the scale of a city or country with each house connected, though with modern robots and postcodes/zipcodes, I suppose it would be technically possible using sorting offices similar to the mail system. I wonder if the system was considered when designing a new city such as Dubai Meydan City Dbfirs 23:22, 20 November 2011 (UTC)[reply]