Computing desk
< November 5	<< Oct \| November \| Dec >>	November 7 >

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 6

causes of limited access on wifi hotspot

OP curiousMahfuzur rahman shourov (talk) 05:22, 6 November 2015 (UTC)[reply]

Not sure what you are asking. Some possibilities:

A) Why is range so limited on WIFI hotspots ?

B) Why is access restricted on some WIFI hotspots ?

Are you asking one of these Q's, or something else ? StuRat (talk) 05:36, 6 November 2015 (UTC)[reply]

Perhaps he's enquiring as to whether certain sites are inaccessible (most likely the hotspot owner choosing not to serve "adult" content or similar for personal or legal reasons) or why speed/downloads are capped/restricted (most likely to reduce costs to the hotspot provider - free wifi is a courtesy mainly and abusing it with high bandwidth activity can be costly for the provider I would surmise). Unless of course he's having issues with the "limited connection" you get when there is a problem with your wireless connection to the hotspot, in which case he would need to consult technical support for the hotspot, his device or both. Quintessential British Gentleman (talk) 01:50, 7 November 2015 (UTC)[reply]

Assuming both questions are being asked and certain sites are restricted. The British Gentleman has sufficiently answered the question if this were just one hotspot. Assuming there is a connection problem with all hotspots, it would be logical that the device is the issue. Hotspots are very backwards compatible so the age of the device should not be a factor. You should be looking at the hardware or software to solve this issue. Depending on your device, you can download apps that will test the hardware and software. If there were any recent updates prior to the problem, undo them. If you dropped the device, the app should recognize the damage to the hardware. Tools like Speccy from Piriform are free and do an excellent job. I found several answers on the LG help pages for troubleshooting my phone. Brnin60s (talk) 03:14, 10 November 2015 (UTC) — Preceding unsigned comment added by Brnin60s (talk • contribs) 01:36, 10 November 2015 (UTC)[reply]

Seamless, High resolution, dark space images

Hello,

I am looking for some very specific types of images, and i can't seem to find anything regardless of my search terms. I need a few samples of different images that satisfy the following criteria:

The content of the image must be of outer-space. This does help some in the "mostly dark" requirement below.
The image is seamless. This means that if you tile the image vertically or horizontally, the left/right sides match, and the top/bottom match. Usually some of the objects in this type of image are partly on the top, and wrap around on the bottom... same for left/right.
The image must be relatively high resolution. 1080p preferred.
The image must be mostly dark, but have some interesting brighter areas.

Some images which are similar to what i am looking for:

Nebula 1 Nebula 2

... I would almost use these images, but the glaring issue is that these are not seamless. If i tile them, its obvious where the image begins to repeat. If it helps, i am wanting this for a subtle dark scrolling background in a main screen menu for some research i am doing in Unity Game Engine.

Thanks to everyone in advance for the help. Believe it or not, googling "Seamless galaxy images" in Google Images brings up lots of bras and photos of Galaxy5 phones!

216.173.144.188 (talk) 05:43, 6 November 2015 (UTC)[reply]

There is a fundamental problem with what you ask. The sky is most easily thought of as a sphere around the earth. The photographs you find out there online are rectangular. If you think about pasting rectangles onto the insides of a sphere, you soon realise that it's not possible to do that without overlaps or gaps. In fact, it's a topological property of a sphere that you can't do that...so what you're asking for is impossible. That said, you can approximate what you're after for a small area of the sky.

For a game, you either need something called a "sky box" - which is typically just a cube surrounding the world with six textures pasted onto it - or you need to compress your map into a circle and use a shader to reconstruct it. If can afford a small licensing fee, I strongly recommend going to a company at http://www.1000skies.com who specialize in high resolution pictures of the sky for movies and video games. Their work is stunningly good - and they've taken care of all of the wrapping and morphing needed to pack a high res sky image into a rectangular texture map. SteveBaker (talk) 15:48, 7 November 2015 (UTC)[reply]

You can get an image (I suggest one with just some stars in it). Then, open it in Gimp and select Filters:Map:Make Seamless. It will make the image seamless for you. 209.149.114.132 (talk) 14:50, 6 November 2015 (UTC)[reply]

Thank you! This is a most valuable answer, because it does not rely on the availability of something; I can make the image i desire myself! The key thing i am missing from many many of the images i do like is this seamless property. Thanks once again! Problem solved!

216.173.144.188 (talk) 19:37, 6 November 2015 (UTC)[reply]

And if you need source images, nasa releases high resolution hubble images to the public domain. Literally hundreds of awesome space images. That's what I use for all my wallpapers. Vespine (talk) 23:57, 8 November 2015 (UTC)[reply]

Nitpick from someone who's spent a lot of time reading about copyright law: they're in the public domain to begin with. Any work of the United States government is, by U.S. law, unable to be copyrighted. They're never "released" into the public domain because they're never copyrighted to begin with. --71.119.131.184 (talk) 02:34, 9 November 2015 (UTC)[reply]

Why Depth First Search uses linear space time?

I would like to know why the space complexity for Depth-first search is Ο(bd)?I have searched the web but it all tells about equations like (m-1)*(b-1)+b that I can't understand.Could anyone give an intuitive explanation with an example.JUSTIN JOHNS (talk) 08:08, 6 November 2015 (UTC)[reply]

Assuming b is the base and d is depth, if you search every item in a depth-first search, you will search about b*d items. If it is a well-balanced tree, you will search 0.5*b*d items. If you don't understand that, go back and look up how to calculate the area of a rectangle and a triangle in geometry. It is normal in O complexity to drop constants. In 0.5*b*d, the 0.5 is a constant. So, we drop it and get O(bd). 209.149.114.132 (talk) 13:24, 6 November 2015 (UTC)[reply]

b is the branching factor. The number of nodes at depths up to d is O(b^d), not O(bd), and in any case the question was about space complexity, not time complexity.

In fact the space complexity of DFS is O(n) where n is the number of nodes, or O(d) if you don't prune multiple paths to the same node. It's not O(bd).

The article on Iterative deepening depth-first search said that its space complexity is O(bd) but also that it's O(d). O(d) is correct, so I removed the O(bd) claim, which dates all the way back to the article's first version from 2004.

Outside sources that claim O(bd) may have gotten it from the Wikipedia article. -- BenRG (talk) 00:09, 7 November 2015 (UTC)[reply]

Bitcoin ransoms

I was reading https://protonmaildotcom.wordpress.com/ in which Protonmail admitted to paying DDOSers a ransom. Needless to say, it didn't work, and they published the address of the DDOSers ( 1FxHcZzW3z9NRSUnQ9Pcp58ddYaSuN1T2y ). What I don't understand is... I didn't think Bitcoins were truly anonymous. I thought each one was a number, a solution to some kind of math equation ... isn't it? So why can't Protonmail publish the list of actual bitcoins they paid for the world to see? Then if those bitcoins ever turn up again, people know they were through the extortionists' hands, right? (OTOH maybe the chain of transactions written for Bitcoins already has the address they published coded into it? I don't know the details)

I still don't *really* understand how Bitcoins are protected against double-paying two different people, but if the extortionists try to spend a Bitcoin, the receiver will want to check it. Can't that check be spotted and flagged - whether by the receiver or the party doing the check - to set off an alarm?

I understand if there are a lot of people who say they're just not going to report any crime, whatever it is, then Bitcoins might be safe to use. But the same is true of cash. But even if the guy you pay cash to doesn't look up the serial numbers, eventually someone will, and then there will be a swarm of agents following the trail backward. Isn't that true for Bitcoins also?

What am I missing here? Wnt (talk) 16:44, 6 November 2015 (UTC)[reply]

Bitcoins don't have serial numbers. The Bitcoin network is just a transaction log. Each transaction has the form "take the money from these previously-unspent outputs of previous transactions, and distribute that money amongst these new transaction outputs." What makes a particular transaction output (TXO) yours is that only you know how to satisfy the cryptographic spending criterion that's attached to it (probably because you have a private key corresponding to the public key it uses). That spending criterion is only meaningful until you spend that TXO (and define new criteria for the new TXOs).

If you tried to taint the descendants of a particular TXO, the taint would spread because transactions can take money from multiple sources. For example, if a transaction takes 1 BTC that's tainted and 1 BTC that's untainted and redistributes it as 0.5 BTC each to four outputs, you have to decide whether the amount of tainted money just doubled, or the outputs are each half as tainted as before, or two unlucky recipients (which you'd have to choose at random) got tainted money and the others are okay.

Worse, if the tainted money was deposited in a Bitcoin bank, there's no longer anything in the public Bitcoin log that says what the offenders turned it into. Even if they withdrew it as BTC shortly afterwards, it probably came from a different, untainted TXO owned by the bank, and all future recipients of the tainted money are innocent victims.

If law enforcement had access to an up-to-date list of TXOs owned by banks (so they knew when the money was deposited in a bank) and could demand that the bank turn over its internal records of where the money went after that, they could possibly trace it. I don't know to what extent that's happening in reality. Obviously the international nature of Bitcoin makes it difficult. -- BenRG (talk) 18:46, 6 November 2015 (UTC)[reply]

@BenRG: I wasn't actually thinking about recovering funds, but finding the extortionist. From that point of view, either of the BTC's derived from a particular TXO would give authorities a chance to haul that recipient in and ask him where the BTC came from. (Or am I still confused?) And I assume a Bitcoin bank can't be entirely anonymous/unfindable or else ... why don't they just vanish one fine day and disperse their funds as their own money rather than returning it to depositors? Wnt (talk) 20:27, 6 November 2015 (UTC)[reply]

Yes, if they can identify the owner of any downstream TXO, "bank" or not, they might be able to trace that back to the original offender one transaction at a time. I agree that people shouldn't trust a Bitcoin "bank" without some kind of accountability, but it's their money and a lot of them trusted Mt. Gox, so who knows. One thing a bank could do to increase trust is publicly pay all its Bitcoin to itself periodically to prove that it still controls it. That wouldn't prevent the owners from skipping town with the money one day, but might allow law enforcement (or anyone else) to identify deposits to that bank in the block chain. -- BenRG (talk) 22:48, 6 November 2015 (UTC)[reply]

Also worth noting is that most of these ransom schemes, and online computer crime in general, are based in Eastern Europe or mainland Asia, where law enforcement doesn't care, and is often inept anyway. Bitcoin is being used here more for the quick and unregulated funds transfer than the anonymizing aspects. --71.119.131.184 (talk) 06:30, 7 November 2015 (UTC)[reply]

The more I think about it, the more this seems prone to regulatory action. If a Bitcoin necessarily contains a TXO chain that includes a traceable attestation that it was once used in a ransom, then it seems pretty obviously to be some kind of stolen/forfeitable property... no matter how many were mixed in what proportion to make it that way. At best the owner might hope to recover a portion by filing suit (and putting up a 10% bond first) under U.S. notions of civil forfeiture. A prosecutor using this sort of action could surely say that a Bitcoin owner "knew or should have known" that the provenance on his coin included markings showing it had been spent as ransom in the past - and the flip side of this is that I'd expect online websites to pop up that can "scan" a bitcoin's history against all known ransom transactions. Such a website would seem like a good business model because the moment it is available it becomes obligatory, according to the known-or-should-have-known model. Wnt (talk) 12:58, 7 November 2015 (UTC)[reply]

If everyone knows or should know which transactions are tainted, then the criminals know, and can just demand as part of the ransom condition that the payment not be tainted. If you wait and taint it later, whoever has the money at that point is screwed, because they couldn't reasonably have known not to take it, but now they can't get rid of it because no one else will take it. -- BenRG (talk) 23:08, 7 November 2015 (UTC)[reply]

Well, I still don't truly understand how Bitcoins work, but I assume that you can't just transmit one of these TXO records that says "I received 1000 bitcoins, honest"; there must be some kind of proof you received particular ones. Is this code (1FxHcZzW3z9NRSUnQ9Pcp58ddYaSuN1T2y) actually embedded as one of the 'previous transactions' in the blockchain, readable by anyone who cares to look? If not, are there other unique identifiers that could have been retained by Protonmail that could be looked for?

As for the moral argument, yeah, it applies, but .... police are under no obligation to keep Bitcoins valuable. Some guy invented Bitcoins in his head - nobody ever guaranteed they were worth something. Police are under an obligation to stop extortion. I think this means that yes, some people who receive Bitcoins are going to get screwed. However, I imagine that they would arrange something whereby they could cooperate with the extortion investigation, possibly paying a portion of the bitcoin in compensation, and then get their particular bitcoins 'whitelisted'. For example, if a bank did the thing suggested above where they reissued all the bitcoins with a commingled history, they'd have to pay back the whole ransom, but then all their bitcoins would be circulable again. OTOH, if no whitelisting is done, then taking Bitcoins out of circulation permanently might make the others more valuable, which would just contribute a bit more volatility/risk to an already volatile/risky investment. Wnt (talk) 13:04, 8 November 2015 (UTC)[reply]

Bitcoin transactions are valid if the total amount of the inputs equals the total amount of the outputs. A 2+8→1+9 transaction is accepted by the network because 2+8=1+9. Any TXO can be spent only once, in its entirety; if you want to send a fraction of that amount to someone, you send the rest to a new TXO owned by you (your "change"). The transactions are nodes and the TXOs are edges of a flow network (with fixed acyclic flows). If you introduce red dye into one of the pipes, eventually a large fraction of the downstream water will be reddish. As an actual example, most U.S. currency has traces of cocaine not because most of it has been used illicitly but because the contamination spreads from note to note when they're handled together.

You can absolutely identify Protonmail's ransom payment in the transaction log, and all downstream transactions. But imagine if the ransom was paid in $20 bills, the criminals deposited it in a bank, the bank put it in their ATMs, a random client of the bank withdrew $60 from an ATM, and then they couldn't spend it until they told the police everything they know about the ransom payment. That's how cash works. People ignore the serial numbers and BTC transactions histories and treat it as fungible. Making people liable for the transaction history of paper currency would fundamentally change the nature of cash. It probably makes sense for the police to deal privately with BTC banks and exchanges to track criminal activity, but the self-destruct system you're proposing is not a good idea. -- BenRG (talk) 20:52, 9 November 2015 (UTC)[reply]

@BenRG: I was pretty sure banks are under some obligations to look for certain serial numbers on currency they receive, though from [1] it seems more unclear than I'd have thought. Note that I'm not really "proposing" this in the sense of going out and trying to make people do it - rather, this is a situation like chemical contamination that is going to remain until it becomes an issue, and it is useful to think about it in advance. Note there is a difference between your example and the situation for counterfeit bills: a counterfeit would be seized from the holder no matter how he received it. Bitcoins are intermediate in status - they are not explicitly counterfeit, but the government has made no promise to uphold them as genuine as they do with legal tender. Given that even keeping encryption itself legal remains constantly under threat, I wouldn't want to rely on their ideological love of Bitcoin.

Also, there is no all-or-nothing decision to be made here. If a bitcoin is 1/100000 ransom money, then a public or private agency tasked with "whitelisting" bitcoins could skim that 1/100000 off as restitution to the person ransomed, and certify the rest as freely circulable. So there's no point at which the money is truly unrecoverable, nor is it especially draconian to track it down late in its distribution. Wnt (talk) 12:41, 10 November 2015 (UTC)[reply]

Perhaps a more apt comparison would be to diamonds, which are a commodity. According to our article the serial numbers on stolen diamonds can be removed rather easily, but if you go to sell one you haven't done this to, it won't matter if you're not the one who stole it - you'll be out of luck and probably looking to cooperate however you can simply to avoid being prosecuted. Wnt (talk) 16:25, 10 November 2015 (UTC)[reply]

.NET C# question about HTTP and HTTPS

I recently got a bug report from customer on-site testing at work. An ASP.NET web application I had worked on wasn't able to open one of its pages.

It turned out that the link to open it was using a plain HTTP URL scheme, whereas the customer on-site installation was deployed under HTTPS.

What puzzles me here is that the code should already have worked this out. It uses the Request.Url.AbsoluteUri property, which should have worked out the correct scheme. But it did not. I added a check that if Request.IsSecureConnection is true, then it uses the HTTPS scheme on the link, but that didn't work either. In the end I resorted to hard-coding the link scheme as HTTPS.

The testing engineer told me that if the site was accessed directly from its own web server, it worked OK. But if it was accessed from anywhere else, it kept writing the link as HTTP even though the site was deployed as HTTPS.

What is happening here? Is this supposed to work like this? Or is there something I don't understand here? Or is this a bug in the actual .NET libraries? JIP | Talk 19:59, 6 November 2015 (UTC)[reply]