Wikipedia:Reference desk/Archives/Computing/2013 February 17

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Computing desk
< February 16 << Jan | February | Mar >> February 18 >
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.

February 17[edit]

PC graphics cards with scrollable screen[edit]

I have a feature on my laptop I've never seen on a PC, but would like. It allows a larger display than the screen can show at once, and when you try to move the mouse off the edges, the display scrolls to show the hidden portion. What is the name for this feature, so I can search for it online ? StuRat (talk) 07:50, 17 February 2013 (UTC)

VNC people call it “bump scrolling” (as in when you bump into a window’s side, scrolling happens if it can). I couldn’t tell you how widespread or standardized this term is. No doubt many projects have independently implemented it just thinking it to be sensible; I know several Unix window managers manage it as well. ¦ Reisio (talk) 09:21, 17 February 2013 (UTC)
This is a feature often found in packages meant to help people with bad vision. That way they can keep the magnification maxed on everything they see and then the mouse moves the portion of the screen that is displayed. Sorry, I don't know the general name of it. On the Mac that I'm using right now, it's under the Universal Access control panel. Dismas|(talk) 14:46, 17 February 2013 (UTC)

Thanks so far. It might help if I explain why I want it. I find it to be very quick on my laptop, presumably since this operation only involves the graphics card, not the CPU. On the other hand, I find scrolling using a window scroll bar to lag badly on my PC (it apparently must reload every image each time I scroll). So, my hope is that I could have a much higher resolution, which could display the entire page at once, eliminating scroll bars, but I would still only look at, say, 1920×1280 of that image at a time (the native resolution of my current monitor). One downside might be that it would then take longer to load the page, initially, since all the pics must be loaded, not just a portion of them. However, I would find this preferable to a lag with every scroll. I could take a coffee break, then come back to a fast page. StuRat (talk) 18:12, 17 February 2013 (UTC)

Interface to wiki-style editing Offline[edit]

There is a recommendation for someone? (talk) 14:06, 17 February 2013 (UTC)

TiddlyWiki? -- BenRG (talk) 19:40, 17 February 2013 (UTC)

Need help fixing my computer[edit]

I'm not sure what's wrong, I was running my computer like usual, but the moment I opened the internet browser, it all just froze, and since then refused to load up fully. I managed to get it working in 'safe mode' and was able to back up all my files and set a system restore to an earlier point, which got it working ok, but with the same result, as soon as I went to open an internet browser, it stops working. Except now, it doesn't load in safe mode either, just comes up with a black screen with a mouse pointer and nothing else. So, is there any way of resetting the whole thing without having to go through anything on screen? And would that even fix the problem?

As well, I note that it crashed both times on trying to access the internet, and only worked on the version of safe mode with no network access, wondering if this is anything to do with the cause of the problem? (talk) 15:05, 17 February 2013 (UTC)

Some questions:
1) What operating system ? Windows 7 ?
2) Can you boot it now, in safe mode with the network disabled ?
3) Which internet browser ? Internet Explorer ? If so, I suggest reinstalling it or running another internet browser, like Firefox, Opera, Google Chrome, etc. Of course, if you don't already have it downloaded, getting it on your PC with no internet browser is a problem. Do you have access to another PC you could use to download it and burn it to a CD ?
4) Do you have a way to restore the system to it's state prior to the trouble ?
5) Might be a virus. Do you have an anti-virus program ? StuRat (talk) 15:17, 17 February 2013 (UTC)

1) Yep, windows 7

2) not any more, but it used to.

3) Originally chrome, then IE, both with the same effect.

4) I've already restored it to well before the trouble (hence why I needed to use IE), and it still crashed.

5) I have AVG, it hasn't found anything, if there is a virus, though, how would I remove it now? (talk) 15:22, 17 February 2013 (UTC)

6) Do you have your original disks, so you can reinstall the O/S ?
7) It might be a problem with the network, are you able to run anything else over the network ? (Another PC, etc.) StuRat (talk) 15:30, 17 February 2013 (UTC)

6) No, although my housemate says he can get some for me soon, and I have a friend dropping off a shiny new Mint 14 installation disk tomorrow, though chances are that's not going to help any more.

7) Everything else works fine, on wired or wireless internet, even using the same ethernet cable on my laptop. (talk) 15:36, 17 February 2013 (UTC)

8) How old is this PC ? If it's about due for replacement anyway, this might be a good time. I just did this when my old PC refused to boot. StuRat (talk) 15:58, 17 February 2013 (UTC)

8) about five or six years, I was thinking of getting a new one, but I would need to wait a couple of months before I can afford an actualy decent one. I guess I can talk to my housemate, in theory he can salvage those parts that still work and reuse them to make a new computer, but if we don't know what's wrong, how can we know what parts to recycle? (talk) 16:02, 17 February 2013 (UTC)

I suggest you buy a refurb PC. I did, and got quite a decent computer for $99 at Micro Center. You can then swap out each component from the old PC to the new, until it refuses to boot, too, then you know that's the problem, so toss out that component and keep the rest. Incidentally, the most expensive part of a computer these days is the monitor, and that can't possibly be the problem, so you know you can reuse that. (The graphics card could possibly be the problem, but that's in the PC, not in the monitor.)
If you want to get the old PC working, for now, you might try a bootable USB thumb drive (assuming your PC can boot from USB) or a CD/DVD with an O/S on it. A Linux variant is one option, and a Windows version is the other. Of course, if you don't already have a bootable USB or CD/DVD, making one without access to a working PC is problematic. StuRat (talk) 16:43, 17 February 2013 (UTC)
I noticed that when I tried to set up my raspberry pi as a mini computer, the cheapest monitor I could find cost twice as much as everything else together. And I can see if anyone wants the left over parts. And I don't know if it would load from the CD or not, but it might be worth a try, at least, when I get it tomorrow. (talk) 16:52, 17 February 2013 (UTC)
Those installation disks are worth a try, too, although installing on your hard disk requires that it be working properly. Incidentally, how are you editing here, without a working PC ? Using the Raspberry Pi ? BTW, if you do need a monitor on a tight budget, a used CRT can be had for free or very little, as people are dumping those to get flat screen monitors. StuRat (talk) 16:53, 17 February 2013 (UTC)
You could use the RasPi to create a boot USB drive. Dismas|(talk) 16:59, 17 February 2013 (UTC)
on my old laptop, which if I remember rightly did pretty much the same thing, one of the times it crashed. And it works sort of ok now. the raspberry pi doesn't work, because that cheap screen turned out to be so small I can't actually see anything on it, need some way of changing the resolution, but I can't do that unless I can see what I'm doing... (talk) 17:01, 17 February 2013 (UTC)
Can you hook the Raspberry Pi up to the PC monitor ? StuRat (talk) 17:04, 17 February 2013 (UTC)
I'm not sure if it has the right socket, I expect if it did we would have done that before, and the shops will be closed by now, so no chance of getting an adaptor. We tried it on the tv, but that couldn't pick up the signal for some reason. (talk) 17:20, 17 February 2013 (UTC)
OK, let us know how it works out. StuRat (talk) 19:54, 17 February 2013 (UTC)
Well, I got the new OS booting from the CD ok, after a couple of tries, then formatted the harddrive, installed it properly and it all seems to work well, no problems at all, no idea what was wrong before, but whatever it was seems to have gone now. Just need to get the computer set up ready for everyday use again now. Thank you all for your help :) (talk) 16:34, 18 February 2013 (UTC)
OK, glad to hear it. I'll mark this Q resolved (remove tag if you need more help, or add a new Q). StuRat (talk) 18:42, 18 February 2013 (UTC)

functional computing with only folds?[edit]

Consider a strict functional programming language (say Haskell), but with explicit recursion disallowed. If we do allow Folds and unfolds, is this enough to make the language Turing complete? I'm not an expert in Lambda calculus. I see at our fold article that a fixed point combinator can be implemented as a fold. Is that enough to implement any recursive function? Staecker (talk) 17:22, 17 February 2013 (UTC)

Yes, it's enough. In place of f args = ... f modifiedArgs ... you can write f = fix (\self args -> ... self modifiedArgs ...). -- BenRG (talk) 19:47, 17 February 2013 (UTC)

WiFi networks[edit]

How do wifi networks send the data each device on it wants especially in public wifi networks where in busy places, there could be up to 100 devices on it at once. Does it just send everything to every device and let the devices sort out which one it wants or does it get sorted by frequency? Clover345 (talk) 18:52, 17 February 2013 (UTC)

Wireless networks that are unencrypted, or using WEP are the functional equivalent of hubs. They send to everyone on their network. And they don't distinguish by frequency. The frequency for a given AP is always the same.
Individual clients on the network disregard packets that are not addressed to their mac address (although there are ways they can instead collect them, which is part of why wireless networks are less secure). In a WPA network the networks are separated because each client negotiates an individual key that's unique for that session. There's actually 2 keys negotiated... one for 1 on 1 communication, and another for broadcast packets, that is shared. However WPA networks that share a common password are vulnerable because everyone can watch that initial handshake that establishes the key, so if you see the beginning of the conversation you can decipher the rest. Shadowjams (talk) 02:43, 18 February 2013 (UTC)
As an aside, like all Ethernet networks (which 802.11 is), there's a jam signal which is registered when two clients try to transmit at the same time. There's a random exponential backoff algorithm that allows them, statistically speaking, wait until they both talk at the same time. Same thing used to happen on hub switched networks all the time. Now because switches are ubiquitous that's unique for wired networks. Shadowjams (talk) 02:46, 18 February 2013 (UTC)

So in WPA, are the different networks on different frequencies? If they're on the same frequency, surely everyone receives every packet. (talk) 11:21, 18 February 2013 (UTC)

No, they're all on the same frequency. The individual packets are encrypted though. Shadowjams (talk) 11:32, 18 February 2013 (UTC)
But isn't that insecure then? Surely people could hack this easily or release viruses on public wifi networks? Clover345 (talk) 18:27, 18 February 2013 (UTC)
Yes, and they do. The virus issue isn't a problem because of the architecture because computers ignore messages not intended for them. Any normal network has a flurry of broadcast packets that go to everyone anyway, there's nothing about that unique to wireless. It is insecure though in the sense that everyone can see everyone else's packets, which is why encryption is important for anything requiring any security. Shadowjams (talk) 20:23, 18 February 2013 (UTC)
It's also worth pointing out that all cellular/mobile 'phone networks broadcast all their communications, so everyone with a suitable wireless receiver can listen to them. However, they, like WiFi, are encrypted communications, so listening to them is not the same as understanding them. So far, the encryption has proved robust, and so mobile phone networks remain secure, despite their broadcast nature.--Phil Holmes (talk) 09:25, 19 February 2013 (UTC)
The A5/1 encryption used on most GSM networks is not at all robust by any technical sense of the phrase. It's fairly obvious that GSM traffic is not secure against any serious attack. However, most cellphone traffic is secure because it's harder to intercept cellphone traffic practically, and then breaking the encryption is another impediment. Phil's also confounding a little bit of the issue. Yes, anything broadcast is "broadcast to all", but the OP was asking about logical (i.e. ethernet level 2) broadcasts... which 802.11 is... but GSM is not. Shadowjams (talk) 11:13, 19 February 2013 (UTC)
I did not say the encryption is robust - I don't know the answer to that. I said "the encryption has proved robust" which is true - AFAIK no-one has demonstrated cracking encrypted cellular traffic. And the OP did not say anything whatever about logical broadcasts, or ethernet level 2. He said "Does it just send everything to every device and let the devices sort out which one it wants or does it get sorted by frequency?". In the sense of answering that question the mechanism used in cellular technology is relevant - since they send everything to every device and let the device sort out which one it wants - just like WiFi. How it sorts that out is different, but the fact that the device picks which signal to attempt to decrypt is the same.--Phil Holmes (talk) 16:54, 20 February 2013 (UTC)
It will help to answer the question if we properly use terminology, e.g., from the OSI model. At the physical layer, most WiFi traffic is broadcasted to everyone; this is because most WiFi antennas are omnidirectional antennas for convenience. So, every device receives every radio-wave. At the next level of abstraction, a "link" is established between a router and a client; that link contains physical information (like channel frequency and modulation methods). As before, any client can receive this data, so an eavesdropper could maliciously intercept this data. But as we go farther up the network stack, we reach the first protocols that govern security and integrity: WEP and WPA. These protocols affect the network and the link layer (though dfferent explanations sometimes disagree on the detail, the canonical answer is "read the OSI standard and the IEEE spec"; because these publications do not use the same terminology, there's room for debate. Assuming the WPA-like protocols have no flaw, any data at a higher level of abstraction than this link-layer is "scrambled" for everybody except the intended recipient. A secure application will also encrypt at the transport- and application-layers. (For example, using SSL and additionally encrypting sensitive user-data). And a truly secure communication platform will exercise the maximum caution at all layers. In general, if a wireless network is set up properly, and uses modern technology that is not known to have defects, it is effectively impossible to intercept useful (application-layer) data. It is trivial to intercept useless (link-layer) data. This same rule of thumb also applies to cellular phones. In the United States, cellular networks are additionally protected by government control - it is not practically possible to purchase radios on the consumer market that are capable of operating at mobile telephone frequencies. Acquiring such a radio that could intercept mobile telephony data, or acquiring the parts to build such devices in any volume, attracts regulatory oversight from the FCC. So, from that perspective, mobile telephony has all the security protection of proper cryptographic algorithm design, plus the extra imposed hassle of making the interceptor rely on difficult-to-acquire, difficult-to-operate equipment. Even HAM equipment is sold with blackout frequencies at mobile telephone bands; much skill and effort is required to circumvent the protection (and violating regulations may expose the operator to license violations and legal liability). This closes the loop; the network integrity is protected by both technical and by political methods. Nimur (talk) 16:13, 19 February 2013 (UTC)
Regarding "it is not practically possible to purchase radios on the consumer market that are capable of operating at mobile telephone frequencies": I can't find any evidence that a Universal Software Radio Peripheral, with the relevant daughter cards, is either restricted for sale or has cellular bands blacked out. As OpenBTS notes, it has been used to implement live GSM base stations in the usual US GSM bands. A NI/Ettus USRP certainly isn't cheap (they $1300 and up), but I don't see anywhere that they're restricted in sale, or only sold to people with relevant radio licences. Their FAQ says here puts all the licensing obligations on the user. Still, as you've said, being able to receive the raw signal is a long long way from being able to intercept the data it carries. -- Finlay McWalterTalk 17:08, 19 February 2013 (UTC)
Of course, if you're going to design and build your own radio... anything is possible; transistors aren't illegal, and a radio is, at its core, just an oscillator and an amplifier; but a useful radio requires a lot of specialized electronics. Hence my comment about "significant skill" and "hassle." Now, if you try to sell it... it's only a matter of time before even your Wikipedia article will soon be referenced by naught but dead links! If you look very carefully at the complete list of products from Ettus, one of the software radio vendors you linked to, you will see a few holes in the frequency spectrum - particularly in the transmitters. If you cannot transmit, you cannot interfere with a high-level protocol; this eliminates an entire class of problems. Nimur (talk) 05:06, 20 February 2013 (UTC)
But I don't see any evidence the dead links are anything to do with censorship, it's easy to find the documents from a quick search. More to do with the fact the article is partially outdated. In fact the article itself mentions that the USRP2 was discontinued and that was one of the deadlinks. I don't see any evidence it was discontinued because of legality concerns or that what replace it is any less capable. And in fact the deadlink was just to a brochure not some super secret document. Another of the deadlinks was just because changed to which I fixed. A third deadlink was another easy to find brochure this time on daughterboards, in fact you yourself provided a link to what's perhaps the best replacement for that deadlink. Meanwhile, I'm not sure what holes you're referring to, but they seem to be well outside the range normally used by mobile phones. In fact one of the daughterboards WBX 50-2200 MHz Rx/Tx is specifically advertised as suitable for 'cell phone base stations, PCS and GSM multi-band radios'. All in all, I'm not seeing any evidence Ettus has been forced to discontinue their products or hide stuff even in the US, let alone other countries. Nil Einne (talk) 06:10, 20 February 2013 (UTC)
Another deadlink is [1]. If you look at the history (2011, 2012), it looks like the most likely reason is because whoever owned the domain couldn't be bothered keeping it up. The key stuff i.e. the software is still [2] although hasn't been updated for a while. I suspect a big problem is no one is particularly interested in using a $1300+ USRP for a GPS SDR as nowadays (at least what my quick search when attempting to find a replacement for the deadlink) everyone seems to be concentrating on using a $10-$20 DVB-T USB stick with the right chipset (Realtek I believe) a well known source of a cheap SDR (I've heard of it despite having no particular interest in SDRs) albeit one that is only capable of receiving of course. Having said that the software is open source and uses GNU radio [3] so I doubt there's much stopping someone using a USRP if they wanted to instead of the very cheap and abundant DVB-T receivers from eBay/AliExpress/wherever. BTW, that means of the four identified dead links, there's no evidence any of them were removed because of censorship or anything of that sort. Nil Einne (talk) 15:38, 20 February 2013 (UTC) (belated signature, somehow I didn't sign)

How to go to a bookmarked page in Google Chrome ?[edit]

I know this is a rather basic Q, but my Google Chrome install lacks the command line on the top I'm used to from other browsers with options like File, Tools, Bookmarks, and Help. Does Google Chrome not have this at all, or is it just disabled in my case ? If so, how do I enable it ? If not, how do I go to my bookmarks without it ? StuRat (talk) 21:06, 17 February 2013 (UTC)

I believe by default they’d prefer you to use the location bar, which also matches against bookmarks. There is no doubt an extension to get the more traditional approach you’re after, however. ¦ Reisio (talk) 21:36, 17 February 2013 (UTC)
Click on the Tools icon (the three horizontal bars in the top right corner). The fourth item in the menu is a subsidiary Bookmarks menu. Top of that is "Show bookmarks bar". Shortcut is Ctrl+Shift+B. Rojomoke (talk) 22:12, 17 February 2013 (UTC)

Thanks !