Wikipedia:Reference desk/Archives/Computing/2010 June 26

Computing desk
< June 25 << May | June | Jul >> June 27 >
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 26

edit

Microsoft Office 2010 Student Edition

edit

Does the Microsoft Office 2010 Student Edition contain/display the "non-comercial use only" stamp like the Microsoft Office 2007 Student Edition? Thanks! --Tyw7  (☎ Contact me! • Contributions)

Yes, probably. Let me find a link.... Chevymontecarlo 14:44, 26 June 2010 (UTC)[reply]
Argh, can't seem to find one. I think there is a separate version for commercial use. Chevymontecarlo 14:47, 26 June 2010 (UTC)[reply]
Anybody has Office 2010 Student Edition? --Tyw7  (☎ Contact me! • Contributions)   Changing the world one edit at a time! 19:28, 26 June 2010 (UTC)[reply]
Yes, it does. 124.214.131.55 (talk) 02:22, 28 June 2010 (UTC)[reply]

New Features

edit

When I log on to Wiki I get New Features activated by default. I have to log in to choose to no-New Features mode. Is there any way I don't get New Feautres by default ?  Jon Ascton  (talk) 04:35, 26 June 2010 (UTC)[reply]

Unfortunately no, the developers of Vector apparently didn't give a shit about non-logged-in users who might not want the horrible new skin. However, there are two ways that I know of to revert Wikipedia back to the old style. Firstly, you can add "?useskin=monobook" (without quotes) to the end of any url, reload, and it should now display the page with the monobook skin. The second way is this userscript for greasemonkey, which adds "?useskin=monobook" to urls automatically while your browse. 82.43.90.93 (talk) 09:52, 26 June 2010 (UTC)[reply]
Actually the developers did look into methods to allow non logged in users to choose but were informed the servers couldn't handle it Nil Einne (talk) 10:31, 28 June 2010 (UTC)[reply]

antenna connectors on cell phones

edit

Old phones had antenna connectors but new ones don't. Is that just a cost saving thing, or is there a better reason? I'd like to add an antenna since my reception is lousy at home, but it no longer seems possible. Thanks. —Preceding unsigned comment added by 75.57.243.88 (talk) 08:06, 26 June 2010 (UTC)[reply]

Speaking as a former cell-phone technician, exterior antennas are superior to internal antennas when it comes to line noise and dropped calls. Even better would be an antenna mounted outside of your cell phone altogether! Some people mount antennas on the dashboards of their cars and even on their roofs. They then hook up those antennas to their cell phones. If you wanted the best possible reception, you would mount a tall antenna on the roof of your car. Someone at Radio Shack can set you up with one.--Best Dog Ever (talk) 08:42, 26 June 2010 (UTC)[reply]
Thanks, but 1) I rarely use my phone in a car and when I do, it almost always works fine. The trouble I'm having is indoors in a residence. 2) I don't see how to use an external antenna with any recently made phone since they don't have antenna connectors any more. That's what I was asking about. 75.57.243.88 (talk) 08:48, 26 June 2010 (UTC)[reply]
This blog post from an antenna manufacturer, about the recent iPhone 4 antenna problem, touches on the fact that the antennas have disappeared due to consumer preference. Comet Tuttle (talk) 15:07, 26 June 2010 (UTC)[reply]
Maybe you can attach an antenna through the charging point. Kittybrewster 15:33, 26 June 2010 (UTC)[reply]
Still unlikely, with any modern phone (more and more are switching to USB connectors). What might work is an external antenna (as in outside the building, not simply outside the phone's casing) connected to a piece of leaky coax inside the building. -- 78.43.71.155 (talk) 22:40, 26 June 2010 (UTC)[reply]
If you have broadband at home, you can of course use a Femtocell if you network operator supports that. Seems a bit overkill in this case however and I'm not sure if the OP is interested in solutions but more the why Nil Einne (talk) 10:28, 28 June 2010 (UTC)[reply]
You can open up your phone and tape the end of an external antenna to the internal antenna. People say it helps. What make and model of phone is it, by the way?--Best Dog Ever (talk) 22:50, 26 June 2010 (UTC)[reply]

Information theoretic loss of JPEGs through displaying

edit

I'm asking this out of curiosity, not for any practical application reason. Suppose one had a JPEG file, displayed on the screen at 1x zoom, no special effects applied or anything, and one took a screencap, and saved the screencap in a lossless format (PNG or BMP or whatever). From an information theoretic point of view, would the resulting copy have lost image information? Would it be in any information theoretic sense an inferior version of the original image? Note that I'm not interested in metadata or anything like that, I'm only talking about the actual image itself. Any thoughts? Maelin (Talk | Contribs) 08:37, 26 June 2010 (UTC)[reply]

No.--Best Dog Ever (talk) 08:44, 26 June 2010 (UTC)[reply]
Unless the computer being used to display and take the screencap only had, say a 16 color display. Then the screencap would be limited to 16 colors, regardless of whatever the original jpg was 82.43.90.93 (talk) 09:46, 26 June 2010 (UTC)[reply]
I think that in some situations, the raw JPEG might have a higher color depth than what can be displayed on a screen, no? --Mr.98 (talk) 12:02, 26 June 2010 (UTC)[reply]
So there's no kind of interpretation or processing algorithm that is used that is not universal to all JPEG displayers? Every possible program you could use to open a JPEG will send exactly the same pixel pattern to the monitor, and hence, one would always get the exact same BMP no matter what app it was screencapped from? Also: assume the system has maximal available colour depth. Maelin (Talk | Contribs) 19:04, 26 June 2010 (UTC)[reply]
I think the answer to the spirit of the question is "yes". The image that gets displayed is the product of the JPEG algorithm, which is applied slightly differently from vendor to vendor, and from compression session to compression session. So, if you took a screen capture of the JPEG image output (or just saved the image into another lossless format like a simple BMP) you would get the derived output of the original file. However, you could *not* ever get that original file back unless you had a tool to carefully reverse the JPEG decompression process under the exact conditions in which it was compressed. So, the details of the compression activity are indeed lost. The image itself (what it looks like) won't be changed but from an information theory standpoint you are losing the information tied up in what made it a JPEG in the first place. This is the exact problem related to "generational losses" in lossy image compression; without careful measures taken each time the picture is edited, the new compression round will not take in any information from the previous compression round and instead generate a new, even loss-ier image. --144.191.148.3 (talk) 15:03, 28 June 2010 (UTC)[reply]
It is unclear to me what you mean by "the original image". By this, do you mean the first JPEG file that you mention? If you have a JPEG file on your system, it has already gone through a step of lossy compression, hasn't it? Comet Tuttle (talk) 15:04, 26 June 2010 (UTC)[reply]
Yes, nomenclature I intended was "original" -> jpeg, "copy" -> bitmap from screencap. Maelin (Talk | Contribs) 19:04, 26 June 2010 (UTC)[reply]
Mr.98 is right; see the "Color profiles" section of the JPEG article. It's not certain that the JPEG will render to your display totally accurately, which would mean that some loss of information occurs. That said, I do this occasionally and it's certainly "close enough" for my work. Comet Tuttle (talk) 21:00, 26 June 2010 (UTC)[reply]
Sure you can lose information this way. Suppose you have your computer set up to produce 16 bits per pixel video (5 bits red, 6 bits green and 5 bits blue - typically). In that case, the color depth has been cut down in the display's frame buffer - and when you do your screenshot into PNG, the low order 2 or 3 bits of each color will be zero. The quality of the resulting PNG image would be worse than the original JPEG. Even if you use 24 bits per pixel, some applications pay attention to the JPEG 'gamma' setting and display gamma-corrected pixels. The PNG image would then have the gamma settings 'built in' and if you then displayed the image on a screen with a different gamma (an older CRT versus an LCD - for example) then the quality of the PNG would be worse than the JPEG would be on a correctly adjusted monitor. I'm sure there are other ways to lose quality too...but that's enough to prove a point. SteveBaker (talk) 02:47, 27 June 2010 (UTC)[reply]

Userscript

edit

Would it be possible to write a userscript that can scan a wikipedia article page a randomly select a link, then navigate to that article? 82.43.90.93 (talk) 10:17, 26 June 2010 (UTC)[reply]

Well, you could use your browser's find feature. In Safari it's Command-F and in IE on Windows it's Ctrl-F. It won't just search for links or automatically navigate to the articles though, sadly. Chevymontecarlo 14:41, 26 June 2010 (UTC)[reply]
Picking a random link and then automatically navigating to it is the whole point of my question. 82.43.90.93 (talk) 14:53, 26 June 2010 (UTC)[reply]
What operating system are you on? this could (obviously) be done in javascript, which you could add to your user javascript page, or if you are on a Mac you could write something in applescript to do it. The first is not an entirely trivial problem, since you'd need to do some dom manipulation to set up a button or link to trigger it, and integrate that with the rest of the page.
(e/c) It's tough to do, since all links aren't just floating around in the <body> element, they're buried in several layers of divs and <p>s. So if you just randomly pick a link from the whole page, you're quite likely to get one of the sidebar links, or an external link. But if you go into the bodyContent div (the actual article text), you can't easily get a list of all the <a> elements. This might be avoidable by using the WP API, but since you're not a bot or bot owner, you don't have access to it (I think). [flaminglawyer] 17:53, 26 June 2010 (UTC)[reply]
API access is not limited to bots or bot owners. Although as a user's limit on queries is 500, and flagged bots is 5000. Avicennasis @ 18:30, 26 June 2010 (UTC)[reply]
I just found the proper API call for just links. I'll try to write up a short script. [flaminglawyer] 19:08, 26 June 2010 (UTC)[reply]

Free wi-fi at a cafe

edit

Why would they use WPA2 security when they freely give out the network key? It seems to defeat the purpose of encryption. 67.243.7.245 (talk) 14:55, 26 June 2010 (UTC)[reply]

Presumably so that they can limit it to customers rather than anyone within a few dozen yards of the building? ╟─TreasuryTagassemblyman─╢ 14:57, 26 June 2010 (UTC)[reply]
Since the name of the cafe is the SSID of the network, anyone nearby could just walk in and ask for the key. 67.243.7.245 (talk) 17:35, 26 June 2010 (UTC)[reply]
They could ask, but I doubt they would be given it. Usually you only get the key when you buy something. --Tango (talk) 22:17, 27 June 2010 (UTC)[reply]

Most likely whoever installed the wi-fi set it up with security as that's the done thing these days. 82.43.90.93 (talk) 15:00, 26 June 2010 (UTC)[reply]

Prevents Reduces the likelihood of casual snooping of the wireless traffic. Comet Tuttle (talk) 15:02, 26 June 2010 (UTC)[reply]
the straight forward answer is "Because they don't know any better". The cafe contracts with someone who comes in and set up the system (who doesn't usually know anything about the cafe business or put much thought into why they might or might not want security). The system is operated by the store manager (or more likely, whatever employee happens to be handy), where 'operates' means 'turns the computer on and off at need'. No one cares beyond that. --Ludwigs2 17:50, 26 June 2010 (UTC)[reply]
The encryption 1) decreases freeloading by non-customers as Treasury Tag mentions; and 2) if I understand how WPA2 works, it prevents eavesdropping of active connections even if the network key is known. That is, if someone walks into the cafe with sniffing tools and sees you working at your laptop, they can't read your traffic. Your session uses a temporary key that was created when you first logged on, and if the attacker didn't intercept that initial handshake, it's too late after that. Without encryption, they could start listening in the middle of your session even though they missed the beginning. That said, anyone doing anything even slightly sensitive at a wifi cafe should do it through a VPN, whether the cafe uses WPA2 properly or not. 75.57.243.88 (talk) 19:35, 26 June 2010 (UTC)[reply]
You misunderstand what that "network key" they're giving out is, and how it's used. It's not an encryption key, it's an authentication token - it's used as part of the (rather complex) 802.11i authentication scheme. Here (rather schematically) is how things work. When you connect to a WPA2 access point (AP) you are mandated to authenticate - at this point you are unauthenticated, and the AP will only let you pass IEEE 802.1X packets until the authentication is complete. To do this you enter into a 802.11i authentication scheme sending EAP packets encapsulated (in EAPOL) in which you pass that authentication token and you're given a Pairwise Transient Key (PTK) using the scheme described at IEEE 802.11i-2004#The Four-Way Handshake. That key only lasts for the session, and is unique for you. The PTK is used as the key for one-to-one communications between you and the AP using an AES based cryptosystem called CCMP. The strength of CCMP (which relies on that of AES) and the uniqueness of the PTK means that no-one can read your conversation, even other people in the cafe who know the same authentication token. Now you might have spotted one fly in this ointment - how did you pass the authenticator to the AP, and receive the PTK, without others intercepting that and being able to read your subsequent CCMP conversation. That's where the extensibility of EAP comes in - EAP can host a number of plugins that allow the EAP transaction to be encrypted, so that information can be passed securely. In the cafe scenario (as for basic home and SOHO routers) that's either LEAP (boo hiss) or a TLS-based public key exchange (TLS is the fancy-pants new name for SSL), generally PEAP or EAP-TLS. So, tl;dr: that number the cafe guy gives you is sent by ssl to the router, which sends you a temporary AES key, and you use this for the rest of your session - that way other patrons can't read your stuff (but the cafe guy can). Some comments: in this scenario, the authentication part is pretty worthless (all you need to do is visit the cafe once, or know anyone who ever has) and you know the authenticator. There are modes for EAPOL which don't need an authenticator - which just give you (still securely) a PTK to use for your visit. These seem not to be terribly widely supported, which is why your cafe is just using the way they are doing. The second thing to think about is how to actually make that authentication useful to the cafe (as some cafes have a problem with people who buy one small coffee and sit there all day taking up space). There are router systems intended for cafes which generate temporary authentication tokens (which live in an internal RADIUS database for a few hours, or for the rest of the day). These can be generated at a web app (so the barista prints one off for you when you ask) or can be integrated with the cafe's till (POS/EFTPOS), so when you buy a coffee there's a fresh authenticator on it, which gives you a couple of hours of access. As with most products targeted at rather tight verticals like this, such systems tend to be rather pricey. -- Finlay McWalterTalk 22:28, 26 June 2010 (UTC)[reply]
I do not have the impression that random wifi routers do TLS, and anyway, as long as there are no secret keys at the beginning, then an attacker can potentially spoof the access point. 75.57.243.88 (talk) 04:26, 27 June 2010 (UTC)[reply]
That long explanation above is right. Only a few tweaks. The EAP system might not use TLS, it could use something else. And it could use RC4 in what is essentially per-packet-WEP rather than AES. And, the system is vulnerable to snooping, but much less since an attacker has to see the initial handshake. Otherwise they're flying blind against 128-bit AES. Shadowjams (talk) 21:27, 27 June 2010 (UTC)[reply]
Correct me if I am wrong, but the attacker needs to be *part* of the initial handshake in order to get in on the information. The TLS handshake allows for secure handshaking by way of asymmetry, so unless the attacker decides what is passed (by orchestrating a man in the middle attack) he has no way of knowing what took place even with the entire conversation visible to him. This is how web sites manage to create 100% secure sessions, with the addition of a certificate authority to prevent a MITM attack. --144.191.148.3 (talk) 14:56, 28 June 2010 (UTC)[reply]
All of the handshakes are [necessarily] in the open. There's no way to avoid that (you could create a public-key mediated connection but that would be vulnerable to MiTM too). But that said, the WPA handshake relies on secure secrets, not a MitM attack. The enterprise editions are relatively immune because they do this with public key exchanges (and even if they are vulnerable against a MitM attack, they might too have cert protections). The issue is if it's a pre-shared key. The pre-shared-key version of WPA2 does the handshake in the open, and so if you know the pre-shared-key you can decrypt that conversation and recover the PTK and the GTK, in other words both of the temporal keys. Even still, WPA is miles ahead of WEP because it uses a good hashing function that requires a lot of processing power. That's why the security recommendations really only warn against dictionary passphrases. If you use a 12 character ascii-random pass-phrase, it's a very good security. Shadowjams (talk) 10:04, 29 June 2010 (UTC)[reply]
Thank you 75.57.243.88. Wish you would make an account so I could thank you proper-like. --mboverload@ 23:14, 27 June 2010 (UTC)[reply]
The IP appears to be resonably static so there's no reason you can't leave any messages at their talk page Nil Einne (talk) 10:35, 28 June 2010 (UTC)[reply]
But notice that the big reply came from User:Finlay McWalter not the IP. APL (talk) 03:59, 29 June 2010 (UTC) [reply]

"non-comercial use only" stamp in MS Office?

edit

An earlier question prompts me to ask: What "non-comercial use only" stamp is Tyw7 asking about? I was going to recommend this version to my sister, but I'm worried this might mean a visible watermark on every document, which is clearly not suitable for letters to the bank and such like. Astronaut (talk) 16:11, 26 June 2010 (UTC)[reply]

The text in question appears on the top bar of the window of the program (see pic) (nb my latest version is Microsoft Office 2007 Home and Student Edition) I haven't seen any visible mark in documents, but it is possible you can tell what version was used to write a particular document somewhere in the individual document properties. File:Office07homestu.png
--220.101 (talk) \Contribs 17:08, 26 June 2010 (UTC)[reply]
No there is no watermark. And there is no way to know which version of office you are using. Plus, you can "scrub" these data from your documents. Just go to Office button-->Prepare-->Inspect documents-->Scan-->Click remove on everything you see. --Tyw7  (☎ Contact me! • Contributions)   Changing the world one edit at a time! 00:46, 27 June 2010 (UTC)[reply]
Thanks Tyw7. I am more familiar with Office '97! If you don't scrub the document as detailed, I take it that some version data might remain? Or if you examined the code in a 'hex' code editor or similar?--220.101 (talk) \Contribs 01:25, 27 June 2010 (UTC)[reply]
I created a blank file and check the propertitiees. Nothing indicates Home and Student Edition. The only thing the other person know is that its created by an office of 2007 or later. Cause Office 2007 & 2010 the default file type is .docx. --Tyw7  (☎ Contact me! • Contributions)   Changing the world one edit at a time! 09:20, 27 June 2010 (UTC)[reply]
Ah, that's what you mean. Thanks. Astronaut (talk) 12:42, 27 June 2010 (UTC)[reply]

Modem - router - computers

edit

After reading Internet Problem above, is it correct to think that if I plug a (small old) router into my broadband modem, and then plug this XP computer and another Linux computer into the router, then it should all start working by itself without me having to do anything further except wait? Thanks 92.15.5.103 (talk) 19:20, 26 June 2010 (UTC)[reply]

You may be able to just wait, or you may have to restart the computers. (Or manually renew their IP connections from the command line.) Comet Tuttle (talk) 19:46, 26 June 2010 (UTC)[reply]
(e/c) As long as it's had its initial setup (which it probably has, since it's "old", as you say), and assuming it hasn't stopped working altogether, then yes, that should work.. [flaminglawyer] 20:00, 26 June 2010 (UTC)[reply]

Thanks, if I try that, but then decide to go back to having my XP plugged into a modem as currently, would I be likely to get any problems? 92.15.5.103 (talk) 20:02, 26 June 2010 (UTC)[reply]

If your router is anything like mine, or my previous router, or the one before that, then it should be fine. [flaminglawyer] 20:38, 26 June 2010 (UTC)[reply]
Worst case scenario, you have to reset your router. [flaminglawyer] 20:47, 26 June 2010 (UTC)[reply]

Would this be the ideal order to turn them on for the firast time: computers, router, modem; or would the other direction be best? Thanks 92.15.1.65 (talk) 13:40, 28 June 2010 (UTC)[reply]

userscript help

edit

I made this userscript that picks a random wikilink on a article and follows it. I got it to display using addPortletLink, but it doesn't work. Any help on what I'm doing wrong? [flaminglawyer] 21:16, 26 June 2010 (UTC)[reply]

There's no need to re-invent the wheel. You can accomplish the task without using any third-party APIs:
function getLink()
{
	var content = document.getElementById('content');
	var links = content.getElementsByTagName('a');
	window.location = links[Math.round(Math.random()*links.length)].href;
}
		
window.onload = getLink;

--Best Dog Ever (talk) 22:22, 26 June 2010 (UTC)[reply]

well, hmmm... try:
function getLink()
{
	var content = document.getElementById('content');
	var links = content.getElementsByTagName('a');
	x = links[Math.floor(Math.random()*links.length)];
	window.location = x;
}
		
window.onload = getLink;

You'll probably want to add a loop to exclude undesired links (test to see if the link contains things like 'special' that you don'y want to follow)

You forgot the href.--Best Dog Ever (talk) 04:38, 28 June 2010 (UTC)[reply]

Problem Downloading RealPlayer

edit

Hi, my friend is having problems downloading "Realplayer" onto her computer. She is using Windows XP on an HP, when she tries downloading RP, it tells her that she has limited access and is restricted. She tried changing her Account Type, but there is only one User Account on the computer, so this option is not available. So, how can she download RealPlayer? Moptopstyle1 (talk) 22:29, 26 June 2010 (UTC)[reply]

It sounds like someone else set up her account as a non-admin account using a hidden admin account. Is it XP Home or XP Pro? --Phil Holmes (talk) 11:10, 27 June 2010 (UTC)[reply]

It's XP Home. Is there any way to change that setting? Moptopstyle1 (talk) 04:25, 28 June 2010 (UTC)[reply]

Sorry if this seems off topic, but see RealPlayer#Criticism. Then, if you still really want to put that useless, spyware- and adware-infested program on her computer, you can try logging into the account named Administrator without a password. Press F8 while the computer is booting up and select safe mode with networking, or press CTRL + ALT + DELETE at the regular login screen. If that doesn't work, then try a portable version. And if those steps don't work, then your friend is a very lucky lady.--Best Dog Ever (talk) 04:37, 28 June 2010 (UTC)[reply]

Both I and my friend use RealPlayer SP, and neither of us has received any spyware or adware, we don't use the message center or anything, we only use it for the converter for AAC to MP3 files so we can add iTunes music into Windows Media Player for home movies. Moptopstyle1 (talk) 04:53, 28 June 2010 (UTC)[reply]

(iTunes will convert AACs to Mp3s, as well, if you already have it installed.) Kingsfold (talk) 15:36, 28 June 2010 (UTC)[reply]

Oh my crap, you can do that? How, my fellow Wikipedian? Moptopstyle1 (talk) 21:46, 29 June 2010 (UTC)[reply]

Google Adwords, Dynamic Keyword Insertion not working as expected

edit

Hi, this page [1] explains that to use Dynamic Keyword Insertion, all you have to do is type {Keyword: Alternate_String}, and it wll replace the {...} part of your ad with the keyword typed in by the user, or replace it with "Alternate_String" if the keyword/list of keywords was too long to fit into the line.

However, when I try this with my ads, it always pulls the alternate string, even though the keyword in question is way shorter than the alternate string. Say my keywords were

wprdc
Ref Desk
wikipedia reference desk computing

and the ad was set up to say

Check out {Keyword: Reference Desk}

then I'm always getting

Check out Reference Desk

even though, from my understanding, I should be getting

Check out wprdc

and

Check out Ref Desk

for the first two keywords.

(Let's not get into the details of using {Keyword} vs. {KeyWord} vs. {keyword} here, I'm aware of that distinction, it doesn't work with any of them.)

I've tried googling the problem but never got any further than the description of how it *should* work, like shown in the Google help page.

What am I doing wrong? Should I have asked bing instead? ;-) -- 78.43.71.155 (talk) 23:09, 26 June 2010 (UTC)[reply]