Wikipedia:Reference desk/Archives/Computing/2010 November 23

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 23

Web hosting services' choice of their (virtual) dedicated servers' Linux distribution

I wonder why certain web hosting services still offer less stable, outdated, and short-lived Linux distributions (like an old version of Fedora) as an operating system choice for their (virtual) dedicated servers. Aren't such configuration particularly risky and inefficient (especially for public, Internet-facing servers like web servers), because either you get a less stable system, don't receive the latest security updates, or have to upgrade (and reconfigure) the operating system every so often in order to keep receiving the updates? 118.96.162.121 (talk) 01:00, 23 November 2010 (UTC)

Domain hacks and political instability

Two hypothetical questions that just floated in my head:

• If I visited, say, a website with URL of http://example.ex/, what does "country .ex" (i.e., its government, military, and intelligence agencies) know about my visit?

• Suppose the "country .ex"'s government is overtaken by an "evil regime", what the worst can the said "evil regime" do?

118.96.162.121 (talk) 02:00, 23 November 2010 (UTC)

• Country .ex know nothing about your visit. Your internet service provider keeps a log of all sites you access and the site you've visited may keep a log but the country whose country code the site uses are never involved. bit.ly is a service for shortening URLs. .ly is the country code of Libya, however the site is based entirely in the United States of America (country code .us), in New York City. It would violate many international and domestic laws if Libya were able to see who was accessing the site bit.ly worldwide. The Data Protection Act of the United Kingdom springs to mind.
• If an "evil regime" were to take over the country I live in, the UK, then they'd be able to gain access to everything I had visited simply by talking to my ISP, under the Regulation of Investigatory Powers Act. I'm not sure what they'd do with that information though. (I'm 17 - they could probably have guessed my internet history would be revision materials and porn). I think they're much more likely simply to cut all access to the internet (or excessively censor it) in order to prevent you from expressing your opinion about their evil ways.

Hope this has answered your question. --tb240904 ^{Talk Contribs} 04:07, 23 November 2010 (UTC)

Trustees of Country code top-level domain are generally given wide latitude in what requirements they may set for registering subdomains. Some do limit registrations to businesses with a local presence or citizens. Theoretically I can't see that there would be anything to stop a trustee requiring the site be hosted locally although I don't think anyone does it, it's a bad idea for numerous reasons.

In any case if domains are limited to companies with a local presence and citizens, the companies or citizens could be required to follow whatever laws the government may have set for their citizens and companies and I can't see any reason these laws couldn't keeping logs of all site visitors and handing these over on request.

The requirements for registering a .cn domain name may be of interest [1] (chapter III) which says for example "The domain name holder shall comply with the laws, rules and administrative regulations of the state concerning the Internet" and "Domain Name Registry and Domain Name Registrars have the obligation of conducting website inspection in concert with the national governing departments, and request to suspend or cease the resolution service of the domain name concerned".

I'm not sure what 'international laws' you think are at play here. In terms of domestic laws, it seems unlikely the UK Data Protection Act would stop companies from other countries who are hosting their websites outside the UK from obeying whatever their local laws may require of them.

Note that I'm not suggesting this is happening anywhere. Simply that I don't see any thing which will really stop it.

Nil Einne (talk) 10:32, 23 November 2010 (UTC)

This is the second time in as many days that this particular poster has alluded to "international laws" that do not exist. --Mr.98 (talk) 12:47, 23 November 2010 (UTC)

Note my 'don't see anything which will really stop it' was only really considering the technical and legal issues. One reason why it doesn't tend to happen is because there's generally little point for a government to know what sites (and pages in the sites) IPs are visiting that happen to be hosted by local businesses (or whatever). They still don't know what the IPs do most of the rest of the time and it's also likely to discourage people from visiting such sites. Further the IP part is important because without cooperation from the ISP or perhaps a person voluntarily giving up the info to one of the sites, they still have little idea who's behind the IP. Nil Einne (talk) 14:38, 23 November 2010 (UTC)

Try a privacy test here: http://www.privacy-test.info/ I wish someone would come up with a Firefox add-on that stopped websites being able to see what other websites you had visited, without having to clear your History. I bet an unscrupulous website could see other things, such as the "cookies" or as I like to call them, the "rats" you have on your computer. 92.15.6.86 (talk) 11:10, 23 November 2010 (UTC)

It's not possible for any site to see a cookie unless the site is the origin of that cookie. Paul (Stansifer) 13:20, 23 November 2010 (UTC)

True but misleading when a web-page are accessed many sites are accessed to get images and so on, some are so common that their cookies can track a large part of a users surfing behaviour se Third-party cookie#Privacy and third-party cookies --Gr8xoz (talk) 15:07, 23 November 2010 (UTC)

Due to a bug in the way that URL colors are handled in CSS, it is possible for any webpage to check if you've visited any specific URLs from a list. However, only checks if the URLs are exactly the same as the website is targeting; the webpage needs a list of URLS of interest. Capturing that someone has visited en.wikipedia.org is easy; to check which articles have been visited the website needs to create a list of all articles of interest. For an example, check which social websites you've visited. The bug is being fixed for most of the major browsers. CS Miller (talk) 19:58, 24 November 2010 (UTC)

Cripes! Does that mean they can a) only check if you have visited a website on their pre-existing list of websites, or b) save something from your visit and at their leisure figure out all the websites you've visited? 92.28.251.194 (talk) 19:50, 25 November 2010 (UTC)

Boleto bancário

Can anyone explain boleto bancário, which is getting popular in Brazil? [2] It seems to be a generic thing, like a cheque, as opposed to a specific company or brand, like PayPal? The relationship with e-payment is a bit unclear. It seems like we should have an article on it, or at least a mention in some related financial article. There's a Portuguese Wikipedia article [3] which at least talks about the general notion. -- Beland (talk) 04:19, 23 November 2010 (UTC)

what's the Features of virtualization software, which is available in the market today ?

i would like someone to help me to answer out this question —Preceding unsigned comment added by Alraeeq (talk • contribs) 09:05, 23 November 2010 (UTC)

See Comparison of platform virtual machines and Comparison_of_application_virtual_machines.Smallman12q (talk) 20:20, 23 November 2010 (UTC)

command prompt

On a Windows computer with restricted privileges and the command prompt and .bat files disabled, would it be possible to still run wget somehow? I only know how to run it from the command prompt and .bat files, but I assume since it's a program which can run without a cmd.exe window there might be some way to start it? 82.44.55.25 (talk) 15:13, 23 November 2010 (UTC)

Run it from the Run box? Hint: To open the Run box, press ⊞ Win+R. 118.96.155.216 (talk) 15:38, 23 November 2010 (UTC)

Thanks, but run is also disabled 82.44.55.25 (talk) 16:25, 23 November 2010 (UTC)

Create a shortcut to wget.exe, then add wget's parameters to the shortcut's properties? 118.96.155.216 (talk) 17:03, 23 November 2010 (UTC)

That might work, but right click is disabled so there's no way to create or edit a shortcut on the restricted computer. Preparing a shortcut on another computer and then copying it over wouldn't really be practical, but it's worth a try if there's no better alternatives 82.44.55.25 (talk) 17:13, 23 November 2010 (UTC)

Either:

• use a portable "wget GUI" kind of program, or
• use a portable Perl distribution as alternative command prompt (use the system function to launch wget).

118.96.155.216 (talk) 17:34, 23 November 2010 (UTC)

By the way, here is an even stranger way of launching a command line without any external "helper" program, by misusing the Object Packager. 118.96.159.246 (talk) 01:42, 24 November 2010 (UTC)

Thanks but I think Object Packager has been removed from recent versions of Windows (Vista/7). I probably should have specified which OS, sorry for the confusion 82.44.55.25 (talk) 12:18, 24 November 2010 (UTC)

Is .vbs disabled as well? I presume .cmd is. Nil Einne (talk) 18:46, 23 November 2010 (UTC)

What is .vbs? I think it's probably going to be disabled too, basically everything is disabled on the computer except usb drives and .exe 82.44.55.25 (talk) 12:18, 24 November 2010 (UTC)

VBScript, on a default Windows installation either cscript.exe [4] or wscript.exe [5] can run VBScript files, I think wscript is the default association for .vbs files if run from the GUI. See also Windows Script Host. If this is Windows 7, Windows PowerShell would also be available on a default install. Is cmd.exe removed or simply not accessible? I seem to recall doing something odd once where I used either Windows Paint or Notepad or Office or something to browse to the cmd.exe location and run it even though I couldn't access it from the desktop although this was Windows XP so Microsoft may have fixed that bug by now. (Or perhaps it was simply copying files from the system directory.) Nil Einne (talk) 13:50, 24 November 2010 (UTC)

Go here and download the Open Command Window Here application. (Here's a direct link: [6].) When the download box appears, click on "Run." Then, right-click on a folder (assuming right-clicking is not disabled, too) and select the option to open a command prompt. If right-clicking as been disabled, you should still be able to right-click on a folder on the left-hand side of a Windows Explorer menu. For example, see the folders on the left under "Other Places" in this window? Try right-clicking on them and running a command prompt that way.--Best Dog Ever (talk) 08:59, 25 November 2010 (UTC)

It is worth pointing out that someone seems to have gone to a great deal of trouble to prevent you from messing with their computer. If it is your employer, you could get in trouble running an "unauthorized" program like wget, the kind of trouble that could lead to disiplinary action. 212.123.243.220 (talk) 17:04, 25 November 2010 (UTC)

AVL-tree: Deleting leaves an ambigous rotation situation

After deleting a node, I end up with a unbalanced tree that can be fixed with either a right-right followed by right-left rotation or only a right-left rotation. Which one should I choose? To get the situation: Draw the tree by inserting {1, 3, 2, 4} (without rotating for balance). Moberg (talk) 15:54, 23 November 2010 (UTC)

I don't see how either of the resulting trees would have an advantage over the other, thus I'd suggest the second method, as it takes less steps.—Emil J. 16:55, 23 November 2010 (UTC)

File indexing, backup and dedublication

I want to have a program for indexing all my files at different computers and storage media. The index should contain file-name and other meta-data and include a secure hash of the content in each file. Preferable the system should be open source but even free-ware with good reputation could be interesting. The system should work on Windows and Linux. The index should be portable as files on a usb-key or similar. I should be able to search for (As an example) important files that only exist on media in my apartment and thus are sensitive to fire and theft and make copies of these files. It should be possible to remove copies of files with more redundancy than necessary. It should also be easy to replicate any indexed directory tree including meta-data at any of the times it was indexed given that a copy of each file-content needed are found on any media.

I have searched for such a program without finding any that are even close. Does anybody know any program or group of program that can be used in this manner? If no are there any good reason it does not exist? Does the concept appear useful?

If I do not find such a program which I think would be useful to very many people in many situations then I will probably write it my self in Python. The system will probably be better if I find other people interested in the system to discus the design and review code and so on and maybe also write code. How do I start such a open source project? Where do I host the project? (Google code, Github, Sourceforge and so on.) Are there any good discussion forums or mailing lists to find interested people? I have experience from writing code for my own research project and similar but not in writing code that could be of wider interest. I have only limited experience with writing code in cooperation with others. --Gr8xoz (talk) 17:21, 23 November 2010 (UTC)

Why are the URLs different?

Several months or years ago I asked a related question and was told this is done by Javascript. What I do is copy and paste newspaper articles I see online to print them out in a compact form rather than waste money and paper printing out a bunch of white space and ads. Certain papers include a URL with the text, which links to the article I just copied. But both this and this give you the same article.Vchimpanzee · talk · contributions · 17:25, 23 November 2010 (UTC)

This is also the same article. I clicked on the link just now.Vchimpanzee · talk · contributions · 17:35, 23 November 2010 (UTC)

If you look at the two URLs:

    http://www.thesunnews.com/2010/11/12/1808513/retail-news.html#ixzz155QMyMyg
    http://www.thesunnews.com/2010/11/12/1808513/retail-news.html#ixzz155QJgK46

you see they only differ with the text that's after the #. As the URL article notes, this is the "fragment". For HTML documents, this mostly is intended to refer to an HTML A tag with an id="ixzz155QMyMyg", allowing a browser to skip to that section. You'll see that's how links in this page (from the table of contents) are done. But those pages don't have A tags with those IDs, so the browser just jumps to the top of that page. So why do they do this - it's a way of tracking how links are shared. So for example if someone posts a link to an article on a news site like Reddit, the newspaper will see all the links with the same fragment id, even if the links are emailed or cut-and-pasted first. So it is, in essence (but not in implementation) a cookie. And you can safely, in this case, omit the # and the fragment ID and get the same page anyway. Or you can put any old rubbish after that # - e.g. http://www.thesunnews.com/2010/11/12/1808513/retail-news.html#Ive_got_a_loverly_bunch_of_coconuts -- Finlay McWalter ☻ Talk 17:40, 23 November 2010 (UTC)

Note that this a rather home-brewed solution of theirs (and it would either not work or would break their page caching if they actually had A anchors they wanted to use). People who write their own webapps can do clever (or just "clever") stuff with their URLs if they want. Amazon allows (meaning ignores) descriptive text in its URLs: so the DVD for The Phantom Menace is canonically http://www.amazon.com/Star-Wars-Episode-Phantom-Widescreen/dp/B00003CX5P , but you can use any old text in there instead: http://www.amazon.com/the-annoying-one-where-you-wish-someone-would-just-shoot-jarjar/dp/B00003CX5P -- Finlay McWalter ☻ Talk 17:46, 23 November 2010 (UTC)

Thanks. I did notice the first part was the same. I thought it might have something to do with outdated URLs I have used in Wikipedia articles. The correct, full URL would go to the updated version, whereas the old URL I used here was linking to something that didn't support the statement for which I used it as a source.Vchimpanzee · talk · contributions · 17:58, 23 November 2010 (UTC)

MediaWiki encodes such information properly, using a well-formed query string. If the Sun News did it that way, their URLs would look something like http://www.thesunnews.com/2010/11/12/1808513/retail-news.html?token=ixzz155QJgK46 - and they would be able to use # properly. -- Finlay McWalter ☻ Talk

What's interesting about this technique is that it does depend on JavaScript; the fragment isn't sent to the server when requesting the page. --Tardis (talk) 20:28, 23 November 2010 (UTC)

Mounting in Ubuntu

When I try to mount a specific drive in Ubuntu (clicking on it in the file browser), I get an error message (Invalid mount options). In Puppy linux I don't have any problem mounting the same drive. And in Ubuntu, I do not get the message with: the option sudo mount /dev/sda1 /media/plus, which works fine.

How can I correct this?

--Mr.K. (talk) 18:18, 23 November 2010 (UTC)

I'm going to take a punt you need to provide a lot more info before anyone is likely to be able to provide any help, at the very least some sort of error log. Nil Einne (talk) 18:40, 23 November 2010 (UTC)

I suppose the information above is enough. There are some invalid mount options somewhere. I want to know where I can change these mounting options. The command mount (without parameters) works fine, so I suppose Ubuntu has some invalid mounting options somewhere. Mr.K. (talk) 18:44, 23 November 2010 (UTC)

You could find out what the options are by temporarily replacing /bin/mount with a stub script that logged its arguments (and exited 1 to indicate failure). You could even have it run the real mount and replicate its (error) output to a file for later examination, if the GUI is oversimplifying:

echo "$@" >/root/margs
realmount "$@" >/root/mout 2>/root/merr; x=$?
cat /root/mout; cat /root/merr 2>&1
exit $x

Obviously be careful to keep the real mount safe. --Tardis (talk) 20:36, 23 November 2010 (UTC)

computers

how do i tell if my computers audio/video jacks are imputs or outputs —Preceding unsigned comment added by 72.242.208.10 (talk) 19:00, 23 November 2010 (UTC)

They may use audio jack color codes. If there are no colors, you will have to check your computer's documentation. Nimur (talk) 21:10, 23 November 2010 (UTC)

If not, there may be arrows pointing in and out of a shape of some sort. Output points out, input points in. General Rommel (talk) 21:14, 23 November 2010 (UTC)

Use a headphone and where you get any sound. You can also try to record sound with the headphone, most work as a (bad) microphone. Do not try a microphone in a jack that can be output, I think the microphone can be damaged. --Gr8xoz (talk) 10:28, 24 November 2010 (UTC)

PDFCreator error

Just recently, whenever I try to create a multiple-page PDF using PDFCreator (eg. from a Word 2007 document), it does the first page, and then a blank second page which says:

ERROR: undefined

OFFENDING COMMAND: H’R×jF’oÆ

STACK:

I tried this using the latest version (1.1.0) and its immediate predecessor; both the same. I'm on a Windows 7 64-bit Dell laptop: any advice? This is driving me insane! ╟─TreasuryTag►estoppel─╢ 19:06, 23 November 2010 (UTC)

You could try using Microsoft's save as pdf addon for Office 2007.Smallman12q (talk) 20:19, 23 November 2010 (UTC)

1. That was not the question I asked.
2. That function only works with TrueType fonts.
3. ╟─TreasuryTag►person of reasonable firmness─╢ 08:52, 24 November 2010 (UTC)

No offense, but its difficult to provide help for an undefined error.Smallman12q (talk) 13:33, 24 November 2010 (UTC)

That is the error message I got. What else can I say? ╟─TreasuryTag►Speaker─╢ 17:23, 24 November 2010 (UTC)

The error seems to say that H’R×jF’oÆ is an undefined command. I guess the PostScript interpreter was expecting a command at this point in the PostScript file, but found data there instead. --Bavi H (talk) 02:33, 25 November 2010 (UTC)

Is it happening with all documents or just some? If just some, what's special about them? What font(s) are you using? What's the source - are you creating them yourself or does someone else on a different system? What happens when you try the "print as image" option (if it's available)? What happens when you print just the second page to the pdf driver? What about just the first page? What happens if you try Word's own pdf printer - error message, document not coming out the way you want, other? :) Have you tried using another pdf printer driver?

I played around with PDF Creator a little, but wasn't able to reproduce your error. Not surprising: I've always found PDF to be a somewhat temperamental format. However, the first step is probably to further narrow down the cause of the problem; I'm not sure how active the developers of PDFC are, but you may be forced to work around it rather than getting a fix, especially if it's an old GS bug. --Kateshortforbob _talk 23:01, 24 November 2010 (UTC)

All documents. I'm using Myriad in OTF format. There's no "print as image" option. Individual pages are fine (it's 38 in total, though!) ╟─TreasuryTag►Africa, Asia and the UN─╢ 09:11, 25 November 2010 (UTC)

I recognize that type of error is a PostScript error, but that's about it. You might try looking in the PDFCreator settings for any PostScript settings. You might try changing settings related to memory (use more or less memory) or compatibility (use more compatible mode or more advanced mode) and see if that has any effect on correcting the error. --Bavi H (talk) 02:05, 25 November 2010 (UTC)

I saw that you asked the question on the PDFCreator forum, and that's really the best place to get an answer. This is clearly a bug in PDFCreator; there's little hope that anybody except the developers can help. Looie496 (talk) 17:46, 25 November 2010 (UTC)