Wikipedia:Reference desk/Archives/Computing/2012 January 19

= January 19 =

Xbox 360 emulator
Has anyone succeeded in making a good Xbox 360 emulator? If they did, I suppose one could install such an emulator on a small form factor PC with a DVD drive and play Xbox 360 games on that. --Melab±1 &#9742; 00:41, 18 January 2012 (UTC)
 * I don't think so. Not for now, at least. Emulating sixth-generation systems has been quite difficult even for console manufacturers themselves, i.e. the Xbox compatibility mode on the 360 and PS2 software emulation on the PlayStation 3. Anything else that claims to emulate the 360 or any recent console is most likely to be a scam. Blake Gripling (talk) 01:14, 18 January 2012 (UTC)
 * Not really... I had the same thoughts until I downloaded and evaluated this PS2 emulator. It ran my copy of God of War 2 (a premium game for its time) amazingly well on my five-year old PC. Also, earlier editions of the PS3 did have some degree of PS2 compatibility, and I'm not sure why Sony removed that ability. Sandman30s (talk) 11:48, 20 January 2012 (UTC)
 * Each core of the 360's Xenon CPU is clocked at 3.2 GHz. It's never going to be possible to emulate that in real time on a ~3 GHz x86. It might be possible to emulate games that don't use the CPU to its full potential. -- BenRG (talk) 23:07, 19 January 2012 (UTC)

Tor Proxy
Just how anonymous is Tor? Is it really untraceable? Is it even anonymous to the courts if someone did something mildly bad from it, but not anything serious like fraud or threatening to blow up the White House? (Not asking for legal advice here, obviously since I'm not providing enough details for somebody to provide good legal advice, I just want to know exactly how anonymous Tor is) 31.204.130.206 (talk) 01:47, 18 January 2012 (UTC)


 * See Tor_(anonymity_network). It's only as anonymous as the servers you are routing your traffic through, and cannot hide from traffic analysis. But what's interesting here is that I'm not sure either of these weaknesses would be available to the justice system after the fact. It's not like a centralized proxy, where you could presumably require the proxy company to divulge its server logs. It's the sort of weakness that you can discover if you are actively trying to monitor traffic — the sort of thing an intelligence agency, not a court, would find useful. But keep in mind that your own computer probably contains lots of evidence of your doings on it, so if there was some plausible way to put you as a suspect, there could be a lot of ways to tie you to the deed, even if your IP wasn't directly traceable. --Mr.98 (talk) 02:25, 18 January 2012 (UTC)


 * You have to consider what you mean by anonymous. As emphasised by the source Mr.98 used, exit nodes can of course record any traffic going through them. So for example, if you write send an email over plaintext where you reveal your name, the exit node is going to have a copy of that email including your real name. Nil Einne (talk) 05:24, 19 January 2012 (UTC)

See also device fingerprint. ¦ Reisio (talk) 03:10, 18 January 2012 (UTC)

It would be interesting to set up a number of POP3/IMAP/FTP/et cetera accounts on a offshore server and log in via as many TOR exitnodes as possible (1 account per exitnode is ideal). If you closely monitor the activity on the server you should be able to spot which exitnodes are being used to steal passwords. Von Restorff (talk) 11:21, 19 January 2012 (UTC)
 * People have done that. Read the article I linked to. --Mr.98 (talk) 12:15, 19 January 2012 (UTC)
 * I just reread it, but I am unable to find it. Von Restorff (talk) 12:42, 19 January 2012 (UTC)
 * Oh, you mean people have set up exitnodes to steal passwords. I know that, but I was thinking about a way to detect which exitnodes are used for that purpose. Von Restorff (talk) 12:55, 19 January 2012 (UTC)

Anyone who has enough money to buy lots of (virtual) servers can create hundreds of TOR nodes. If you control all the nodes there is no security. More nodes under your control = less security for the users of those nodes. Von Restorff (talk) 11:29, 19 January 2012 (UTC)

Weird letter characters
I have on occasion stumbled upon very weird letter characters (not sure if they are ascii) that have stuff "attached" to them above and/or below. An example is m̘̟͖͔̮̗̗͖͚̺̩͉ (which in the edit window looks like an m horizontally followed by a string of symbols, but in the rendered page these symbols appear below the "m", overlapping the lines below). I've been googling off my fingertips trying to find a list or table or really anything about these characters or what they are used for. Could anyone please help me out? What are these, are they ascii and where can I find a complete list, ideally including HTML code or other character reference? Thanks in advance, --213.168.119.48 (talk) 11:38, 19 January 2012 (UTC)


 * It's an ordinary 'm' followed by the unicode character u0318, which is a combining character (meaning it modifies those around it, rather than appears as a glyph by itself). Looking it up it's a "combining left tack below". Some Googling suggests that this is used to indicate phonetic things in various languages, but you'd be better to ask at the language reference desk, giving them an idea of context (so they know which language is being used). -- Finlay McWalterჷTalk 12:01, 19 January 2012 (UTC)
 * Thanks! Combining characters, right? I had never heard of those before. (My question was rather of a purely technical nature and perfectly answered by you. I'm not sure RD/L could help me further, since my curosity was of a technical nature, not related to any specific application.) Thanks again, --213.168.119.48 (talk) 12:10, 19 January 2012 (UTC)
 * Aside from representing text written in actual languages, combining characters have been found useful in conveying a sense of cosmic horror. Paul (Stansifer) 13:51, 19 January 2012 (UTC)
 * Thanks Paul, I like that link a lot! Von Restorff (talk) 15:25, 19 January 2012 (UTC)


 * Actually looking at it a bit closer, it's a string of combining characters after the 'm': U+0318 U+0356 U+0354 U+032e U+0317 U+0317 U+0356 U+035a U+033a U+0329 U+0349 - he's just built a meaningless stack of diacritical marks just to make a splodge. -- Finlay McWalterჷTalk 15:43, 19 January 2012 (UTC)


 * If you want to see all the those combining characters un-overlapped, here's the string printed with a normal 'a' between each character: ama̘a̟a͖a͔a̮a̗a̗a͖a͚a̺a̩a͉a  -- Finlay McWalterჷTalk 15:47, 19 January 2012 (UTC)


 * These kinds characters are normally used to construct the letters of some south and south-east Asian alphabets. For example: the Thai alphabet, Gujarati alphabet, Devanagari (the script used to write Hindi), Lao alphabet, and a few others.  Astronaut (talk) 16:40, 19 January 2012 (UTC)


 * One thing I've not found, that would be very useful, is a resource that, given an arbitrary unicode code-point, would tell me more than just the name, but the circumstances (languages) in which it is used. -- Finlay McWalterჷTalk 17:14, 19 January 2012 (UTC)
 * One unicode point is too ambiguous. It might not even be unicode!  (For example, consider the character 'a' - or 0x61... is this byte-value a character or an unsigned integer 97, or ... what?  The byte, by itself, is not sufficiently distinct to provide that information).
 * Given a stream of bytes, we can use charset detection to at least determine which character encoding is being used.
 * To confound matters more - a valid Unicode stream need not use a single human language: it's perfectly valid to switch to French or Arabic or Chinese, in-line. And of course, many streams-of-valid-unicode are valid in multiple languages.  For this reason, Language identification, the first step in many machine translation algorithms, must be performed statistically: it is likely that a character sequence is German, or French, or Cherokee, or whatever.
 * If you're looking for APIs - there are loads of third-party libraries (I saw several linked on the article references I posted above). I've also been playing with NSLinguisticTagger, now available on Mac OS X Lion, making natural language processing a simple call to a system-library (instead of a long and horrible slog through academic research-papers).
 * As a last note, given a value that you know is a Unicode code-point, you can look up the code plane and code block. Some of these code blocks have plain-names that might clue you in: e.g., "Latin" or "Greek and Coptic", and so on.  As above, caveated by the ability to mix characters and languages - Cyrillic or even Greek characters can be used to write Russian or Serbian language words.  Nimur (talk) 19:45, 19 January 2012 (UTC)

Then what about glyphs that are at the top of the letter, such as seen in "H̴̨̨̜̯̝̙̯̠̯̝̭͉̠̖̯͉͚͉̳̊̆̽̇̔̀̄̚͜͞Ą̿̒̈ͯ̎͗ͮ̾̈́̔ͯͭ̈ͦ̓͝҉͔̬͕̬̪͉̮̗͇̪L͉̭̥̥̩̘̳͇̥̫̬͈̦̘̱͇͖͕̫͂̓͛̓ͬ͂̃͜P̢̺̘͎͖̠͔̥̳̙ͧ̅̉́͂́̾ͭͅ!̃͛ͪͫ̓͒̽̇̓̾͛̑ͤ̏͒͏̨͖̮̗̕͞ͅ 	Z̸̧̛̛̗̣̩̝̬̣̖̳͔͓͇̈̑ͭ̒͋̏̇̇ͪ̈̀̾̀ͧ͋̀Aͣͬ̉͊̉ͭ̿̈̍͐̈́ͦ̆̈́̀҉̶҉̼̥͉͢Ľ̷̶̛͙̦͇͔̦̭̄̆̍̽̎̎̉͊̄ͬ̏͜G̵̡̢̥͔͙̦̥ͣͬͣ̓̓̐͑ͫ̆̍̔͑͛ͯ̽̐̾ͦƠ̧̛̞͙̲̹̜͚̼͖͈͓̜͕ͧ̓̃ͧ̑͗ͮ͊͂̋̅ ̢͖̤̯͇̗̻̙̰͙̮̂̒ͣ͛ͨ̍ͣ̎̔͑̃ͯ͐ͥ͑ͩ̚̕͠H̷̛̩̜̯͌͐ͣ͐͊̔̂̐̕͟͡E͎̪̮͈͕̗͎̹̦͕̼̜̾̐̎̓͌̊ͪ̽̍ͨͤ̒͗ͨ͞͡ ̛̝̱͍͕̞̫͉̲̮̝̻̠̬͉̝ͯ́̿͘ͅC̝̯̳͉̠͙͇̞̞̰̜̉̓̇͂͛ͨͫͣ̾̎͛̌̊̀͝͡Ő̒͑ͪͪͬ̈̓̒͏̸̼̲͈̫̳̥̯̭͎̥̖̭̣͉̪M̢̹̬͕̬͈̪̬͈̝͖̣̭̓ͥ̌ͯ͌̎ͣ̽ͯͩ͛̇̿͆͢͟͝͡Ǐ̷̛̤̗̝͓̯̦̯̈́ͬ̎̃̋̾͌̑̌̎ͧͤͥ̀͛́́̚̚͝Ṋ̶̳͓̜̰̖̟̤̇ͭ̄͆̀ͨͪ̇̆ͨ̀ͨ͐̿̓̒̓̚̚͘͟ͅͅG̵̮͍̗̘̮̝̖̖̮͉͚̻͇̞̞͌̿ͤͯ͆͐ͤ̇ͅͅ ̜̹̪͛̊ͥ̀ͭ̂̚̕͢Iͫ̉̊ͪ͑̂̿̇̃̈̎̅̔͛ͥ̾ͤ͊҉͍̙̩̝͎͜ͅS̷̛͓̯̬̥̟̱̈́ͫͣ̒̾͗ͯͥ͌ͭͯͥ̋̒̋ͮ͜!̴̧̛̦̺͓͙͕̙̙̼̞̱͕͔̰̭̘̩̬͊͑͌͒ͭ̓́͠"? → Σ  τ  c. 08:23, 25 January 2012 (UTC)

One question about SOPA.
With sopa it would be illegal to put infringing links in your websites. But my question is: Why this part of sopa law exist? If a infringing website will be removed because this website would be against the law, wouldnt it be impossible to link to infringing websites, since they would not exist in the first case anyway? — Preceding unsigned comment added by 201.78.223.152 (talk • contribs)
 * Think about foreign websites, different laws in different countries. Also, the fact that there is a law does not automagically mean everyone obeys it. According to MLK one has a moral responsibility to disobey unjust laws. Von Restorff (talk) 15:15, 19 January 2012 (UTC)
 * To know a link is illegal and not allowed to be used on another website, they would to..... know this link is illegal, if they know some link is illegal they would try to remove the link. By removing the link the website would not link to a illegal website because this website doenst exist anyway. Also foreign websites would be blocked following the same paradox. 201.78.223.152 (talk) 15:26, 19 January 2012 (UTC)
 * That's it exactly; it's impossible. Bear in mind what big media companies really want - they want the Internet to be like cable TV: a walled garden consisting only of controlled, pre-approved content, running on a platform that (like the iOS market) also controls what can and can't be used. They try to sell this claiming this prevents copyright theft, child-porn, and hate speech. But it also constructs a cartel, where only members of the cartel can distribute content, and so they totally lock out new players and disruptive competition. 87.115.127.51 (talk) 15:36, 19 January 2012 (UTC)
 * By penalising people for posting links you're making it harder for people to find illegal content in the short period between the link being posted and the content being blocked or removed. Other measures in SOPA such as DNS filtering will prevent Americans from accessing infringing foreign content, but it may take a few days to get a filter in place.  The only way to fill in that gap is to force people to police themselves.  It's already illegal to run a website whose primary purpose is posting links to infringing content (see e.g. Legal issues with BitTorrent), but SOPA will clamp down on otherwise-legal websites who happen to post a link to infringing content.  The result, as already mentioned, be to will make websites think twice about carrying any links or user-generated content. --Colapeninsula (talk) 16:03, 19 January 2012 (UTC)


 * It needn't even be user-generated content. Imagine you're a modest sized media production company, producing a weekly hour-long programme featuring new bands. You work with small indie labels and feature their new signings playing in concert. You have legal agreements with the bands, their labels, and the venues involved - you're a totally legit outfit. But say occasionally one of the bands plays a cover version of some obscure Jacques Brel song, and they don't tell the label, and so you don't know, and put the video out anyway, entirely in good faith. A few months later someone notices and the Brel people complain.  If it goes out on TV, you're totally okay - it's covered under existing songwriter broadcast compensation schemes.  Right now if it goes out on the internet, they have to send you a DMCA takedown notice, and so the worst that happens is you have to take that video off the internet until you've got the permissions problem cleared up.  But under SOPA you can be declared a copyright violator, your internet domain taken away, and anyone who even links to you (not that video, to anything on your site) is breaking the law. So, although you acted in good faith and with reasonable professional diligence, it's a death-sentence for your business. The little indie labels and self-managed bands you deal with can't always be absolutely thorough about this kind of thing; only giant top-heavy lawyer-rich recording companies can - so if you want to avoid this, you take all your content not from the indies but from the big record companies (and thus shut the indies out of the exposure they need). And look who is pumping money at congress for PIPA and SOPA - it's not those indies (in music, in movies, in games) it's the big players seeking to protect themselves from competition. 87.115.127.51 (talk) 16:27, 19 January 2012 (UTC)
 * Or just imagine you link to a site and the site's content changes from when you initially set up the link. There are myriad problems for holding websites responsible for what they link to, if there isn't some kind of due process involved. --Mr.98 (talk) 19:28, 19 January 2012 (UTC)


 * Part of the intent of that part of SOPA is to get around the "we're not hosting it, we're just linking to it" argument that is common amongst aggregator and torrent websites. Torrent databases (like Pirate Bay), for example, don't actually host any infringing files — they just provide the connection data necessary to find other people who are sharing the file. It's a serious difficulty if you're in the business of trying to stop piracy, but the solution in SOPA is a poor one for the many reasons given. Criminalizing the use of links in a broad manner is essentially incompatible with the current usage of the Internet, even without all of the due process issues. In general the most problematic parts of SOPA come from the fact that its attempt to stop piracy is rooted in the idea that they can "hide" pirating websites outside of the US from people inside of the US. It is a misguided approach. --Mr.98 (talk) 19:04, 19 January 2012 (UTC)

Is there some way to look up the non-toll free number that a toll free number goes to?
Like there's 800-123-4567 and it goes to 202-123-1212 (both numbers are made up). That kind of thing. Now if there's just an 800 number given out, is there some way to look up the non-toll-free number that it goes to? Are you ready for IPv6? (talk) 20:10, 19 January 2012 (UTC)
 * Not in general. The redirect can be performed in a variety of ways.  If it's performed by the telephone network, then you (as a general user) aren't privileged enough to see the network route, unless you own the part of the routing network on the public switched telephone network that's responsible for the redirect.  Further, if the redirect is done at the endpoint, even the privileged operator telephone network cannot perform such a look-up.  Conceivably, a "highly capable adversary" could perform traffic analysis and monitor the endpoint, but it would not be easy.
 * Our article is telephone number, and the concepts that underpin this amazing system are much more complex than many people realize. Nimur (talk) 20:22, 19 January 2012 (UTC)
 * And yes, Wikipedia is ready for IPv6. Von Restorff (talk) 21:18, 19 January 2012 (UTC)