Wikipedia:Reference desk/Archives/Computing/2009 March 21

From Wikipedia, the free encyclopedia
Computing desk
< March 20 << Feb | March | Apr >> March 22 >
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.


March 21[edit]

net[edit]

i'm buying a usb modem to have net at home. Can interconnect 2 PCs using one usb modem. one is downstairs and one in my room. what do i need. And is there a wireless way i can do that? —Preceding unsigned comment added by 193.220.225.251 (talk) 00:06, 21 March 2009 (UTC)[reply]

The easiest way to connect multiple computers to have access to one internet connection is a Router. You can get Wireless routers pretty much everywhere these days - even the supermarket. To be able to be wireless you will need a wireless enabled computer/laptop - though most wireless routesr have at least 4 wired-connection slots too (at least from what i've seen) so you could have a (wired-connection) desktop PC and a (wireless connection) laptop. ny156uk (talk) 00:23, 21 March 2009 (UTC)[reply]

To connect your modem to a "router" (confusingly known as either a residential gateway or a wireless router here on Wikipedia) you'll need to have a modem with a network connection. If you have a pure USB modem, you'll need to connect it to one of your computers and have that computer "share" the internet connection across a home network (where a "router" would also prove useful). The latter layout has some limitations (all traffic travels through the "host" computer, so it has to remain on; setting up the connection is somewhat more complicated; and you don't get the security benefit of the router's firewall), so it would be best to purchase a networkable modem instead (many support both USB and network connections). – 74  03:28, 21 March 2009 (UTC)[reply]

Slow Computer[edit]

My computer's gotten slow, both at bootup and while running applications. I have Norton and AdAware running, so I doubt it's a virus. Any idea what could be causing this or how it can be fixed? Black Carrot (talk) 07:00, 21 March 2009 (UTC)[reply]

What OS? Ζρς ιβ' ¡hábleme! 07:37, 21 March 2009 (UTC)[reply]
You have Norton running. That's one. Also, try doing Start->Run (or Win + R), and type in "msconfig". Go to the startup tab and see what's starting. May also want to check services.msc to see if you've got any strange services, but don't mess with any entries in either unless you know what they are. -- Consumed Crustacean (talk) 08:09, 21 March 2009 (UTC)[reply]
Once you've cleaned out the start-up as Consumed Crustacean suggests, do a defrag ... Start, Programs, Accessories, System Tools. — Ched ~ (yes?)/© 08:52, 21 March 2009 (UTC)[reply]
I think the most likely reason is because you need more system memory (not be confused with hard drive space). What version of Windows are you using and how much memory do you have? These days if you're running XP with Service Pack 2 or newer and antivirus etc like you are then you'll need at least 1Gb+ and if you're running Vista you'll need 2Gb to run things comfortably. In both XP and Vista you can go to "Control Panel" then "System" and after a short time the first page that shows will display the system memory. ZX81 talk 13:25, 21 March 2009 (UTC)[reply]

The startup tab of msconfig doesn't show a very long list, and I don't understand any of the names well enough to know what to change. The defrag system says I don't need to defragment. The General tab of the System Properties folder says, Microsoft Windows XP > Professional > Version 2002 > Service Pack 3, and AMD Athlon(tm) 64 Processor > 3200+ > 2.01 GHz, 1.00 GB of RAM > Physical Address Extension. Does that narrow anything down? Black Carrot (talk) 19:01, 21 March 2009 (UTC)[reply]

When the computer starts going slow does the hard drive light flash like crazy/constantly on? Sadly 1Gb in XP is pretty much bare minimum needed these days (with SP2 onwards/antivirus/spyware/firewall running on the computer) and quite likely it's ran out of real memory so it's using the hard disk which is much slower. If you load up Task Manager (Press Ctrl-Alt-Delete) under the Performance tab you'll be able to see memory usage. Although the whole page is quite helpful, if you can tell us what the available/system cache numbers are next to "Physical Memory" that'll help to confirm if you're running low or not (do it when you have your "normal" programs running, having everything closed will give a false reading) ZX81 talk 19:24, 21 March 2009 (UTC)[reply]
  • You can Google most of the executable names; anything exceptionally odd that doesn't come up on Google is suspicious. And like I said, Norton isn't helping. -- Consumed Crustacean (talk) 05:19, 22 March 2009 (UTC)[reply]

A friend's laptop had this problem and it turned out his hard drive was running in PIO mode for some reason. In XP you can view and change this setting in Device Manager, IDE ATA/ATAPI controllers, Primary IDE Channel, Advanced Settings. Just another thing to try. -- BenRG (talk) 13:21, 22 March 2009 (UTC)[reply]

I suggest you download Antivir's antivirus immediately, then uninstall Norton Anti-Virus with Revo Uninstaller (with an advanced uninstall). Procede to install Antivir. You will notice a difference immediately. Plus, CCleaner can help a little, and it's always useful in freeing up space. --71.98.14.96 (talk) 21:20, 22 March 2009 (UTC)[reply]

Fork bombs[edit]

(source: Fork_bomb#Example_fork_bombs)

I have two questions. First, will someone please explain why

:(){ :|:& };:

is a fork bomb and how it accomplishes it's task. Next, why is

format ELF executable
entry start
start:
	push	0x2       ; Linux fork system call
	pop	eax       ;
	int	0x80      ; Call to the kernel
	jmp	start     ; Loop back to the start


used instead of

format ELF executable
entry start
start:
	mov	eax,0x2   ;
	int	0x80      ; Call to the kernel
	jmp	start     ; Loop back to the start


TIA, Ζρς ιβ' ¡hábleme! 07:37, 21 March 2009 (UTC)[reply]

1. [1].
2. Maybe the push/pop combination takes up less space? --wj32 t/c 08:47, 21 March 2009 (UTC)[reply]
2. I can't see how that would be as you would have to change both push and pop to their hex equivalents and change eax and 0x2 to their hex equivalents. In the second example, you just have to translate mov and the two operands. Ζρς ιβ' ¡hábleme! 13:47, 21 March 2009 (UTC) P.S. Is it something about null/non-ASCII bytes?[reply]
It is indeed done to save space. PUSH has a form with a one-byte operand that's sign extended to four bytes, but MOV doesn't. The register operands are part of the opcode byte, so the PUSH/POP pair takes three bytes and the MOV instruction takes five. -- BenRG (talk) 18:46, 21 March 2009 (UTC)[reply]
For the first question (which I assume is about shell commands), this code:
:(){ :|:& };:
is evidently deliberately obfuscated. First it declares a function called ':'. That's an unusual name for a function - but it's not illegal. It's clearly deliberately confusing, so let's call it something less confusing like "bomb". Now we have:
bomb(){ bomb|bomb & } ; bomb
. That's really two commands separated with a semicolon:
  bomb(){ bomb|bomb & }
  bomb
In other words, it's declaring a shell script function - and then calling that function. So what does the function do? Well, the innards of the function are:
 bomb | bomb &
- which means run the function 'bomb', pipe the output of it into another instance of the function 'bomb' and run the whole mess in background. So this is a recursive function that calls itself twice each time. So you run 'bomb', and it creates 'bomb|bomb' which results in (bomb|bomb)|(bomb|bomb) ...and so forth - doubling the number of backgrounded threads each time. SteveBaker (talk) 18:43, 21 March 2009 (UTC)[reply]
Oh, dear, that a cue for a song:
(bomb|bomb)|
(bomb|bomb)|(bomb|bomb)|
(bomb|bomb)|(bomb|bomb)|(bomb|bomb)(bomb|bomb)
there's no limit! (sorry) 87.114.29.204 (talk) 21:10, 21 March 2009 (UTC)[reply]

question[edit]

Can I have a links to free chess games for my windoes computer? thank you. -jonty —Preceding unsigned comment added by 194.80.240.66 (talk) 11:34, 21 March 2009 (UTC)[reply]

How about WinBoard, or another game from Category:Free chess software? In future questions, please use a more descriptive header than "Question". Xenon54 (talk) 11:40, 21 March 2009 (UTC)[reply]

ah well thank you, i am still learning all the pages to this site, it seems there is a page for everything it's finding them that is the problem -jonty

Windows Vista comes with a quite nice 3D chess game. --Andreas Rejbrand (talk) 12:22, 21 March 2009 (UTC)[reply]
By "free chess games" do you mean you want to replay famous chess games, play chess against the computer, or play against a human opponent ? For the last case, Pogo.com does a decent job, as does Yahoo!. StuRat (talk) 05:19, 22 March 2009 (UTC)[reply]
You could also try the chess game that comes with the Microsoft Entertainment Pack. JCI (talk) 01:58, 24 March 2009 (UTC)[reply]

XP windows update: how to start custom install[edit]

Resolved
 – Certes (talk) 20:33, 21 March 2009 (UTC)[reply]

I run Windows XP (legally but without WGA) and would like to install some but not all of the updates I have downloaded. Shutting down the computer only allows me to install all or none. What I need is the Custom Install available via the yellow shield in the taskbar or something equivalent. Is there any way to make the yellow shield or the dialog it opens appear on demand? Obviously XP contains a program or service to do this but a long trawl through the search engines fails to reveal its identity.

If not, is there any other way to run a Custom Install? I realise that I can use Microsoft's website if I install lots of unwanted software first but would rather avoid this. I have tried stopping wuauserv, amending registry keys within \HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\WindowsUpdate\Auto Update (not necessarily the right ways) restarting wuauserv and running wuauclt /detectnow, but nothing visible happens.

All suggestions gratefully received. Certes (talk) 12:07, 21 March 2009 (UTC)[reply]

For the benefit of anyone else with this problem, someone kindly pointed me at [2] where "ulogic" has a solution. Certes (talk) 20:33, 21 March 2009 (UTC)[reply]

Stickam.com[edit]

How would I go on about & making my stream title/display name like this guy: http://www.stickam.com/tweetybirdluvsu, "hidden"?

I've seen it done on various people's profiles, on this site but I just don't know the exact coding nor fully get it.

Also...

"You may enter HTML/CSS. JavaScript is not allowed." —Preceding unsigned comment added by Kvltgrinder (talkcontribs) 13:08, 21 March 2009 (UTC)[reply]

Turning things off and on[edit]

It seems that there's an old adage somewhere that says to "save electricity", you have to turn things off when they're not in use. So, my question is, would it be more beneficial for me to leave something on, rather then turning it off and then back on after a few seconds? For example, if I were using my calculator for homework, I tend to turn it off after every question (just a habit), but then turn it back on after I finish reading the next question. Does this waste more battery life than if I just leave it on completely? Another thing, my mother used to tell me to turn lights off when I leave a room, even if it were for a short time (usually just a few minutes). Would this cause more electricity to be used than just leaving it on for that short duration? Furthermore, when turning something on, does it use more electricity then it does when it's currently in use? I'm hoping someone can answer these questions. Thanks a lot! 141.153.217.159 (talk) 15:58, 21 March 2009 (UTC)[reply]

This has been covered in many (MANY) areas. There is a lot of anecdotal evidence that turning off something for a very short time will save less energy than the amount it takes to get it turned back on. The key thing to remember is that the plural of anecdote is not data. The data, produced by testing, even on rather pathetically scientific shows such as MythBusters, indicates that electrical devices do not require a lot of energy to turn on. Therefore, the rule "leave it on" only applies if you plan to turn it back on within less than a second of turning it off. -- kainaw 16:04, 21 March 2009 (UTC)[reply]
The problem with repeatedly turning things on and off isn't that it wastes energy, it's that it subjects the items to unnecessary wear and tear. For example, most switches are only good for a finite number of uses, based on what they expect the typical usage to be, and this doesn't include the type of duty cycle you seem to have in mind. StuRat (talk) 05:09, 22 March 2009 (UTC)[reply]


They did this on Mythbusters for lamps and for them it showed that it saved the lightbulbs if you turned them off everytime you left the room. chandler · 10:09, 22 March 2009 (UTC)[reply]
The OP hails from New Jersey, so this is not relevant, but incandescent bulbs are being phased out in the EU, commencing in 2010. The aim is to reduce CO2 pollution by 30 million tons. This brings up the question if CFLs (compact fluorescent lightbulbs) and other options like halogen or LEDs require more power when being switched on repeatedly.
According to one site Mythbusters gave these data:
  • Incandescent: 0.36 seconds
  • CFL: 0.015 seconds
  • Halogen: .51 seconds
  • LED: 1.28 seconds
  • Fluorescent: 23.3 seconds for a switch off / switch on to break even. Presumably this refers to energy usage and not wear and tear on the switches. --Cookatoo.ergo.ZooM (talk) 10:53, 22 March 2009 (UTC)[reply]

How is the case with turning the vehicle engine off while at a traffic signal where the light is expected to turn green in half a minute or so? Does re-starting the engine burn more petrol than keeping it running for half a minute. Not a computing question, but fits in this discussion thread. Jay (talk) 08:18, 23 March 2009 (UTC)[reply]

Generally no (though it does depend on the car). Some vehicles are designed to do this automatically as a feature (mostly hybrids), but frequent restarting in vehicles not designed to do so could cause excessive wear on the starter and electronics. – 74  14:28, 23 March 2009 (UTC)[reply]

Solaris 9 in VMWare[edit]

I installed Solaris 9 inside VMWare Workstation 6.5, and I have everything working except audio and ethernet. I'm using the CDE. Whenever I type ifconfig -a no ethernet adapters show up, just lo0. I tried typing ifconfig hme0 up, just for the heck of it, but it said there was no such interface. My NIC is a Broadcom Gigabit Ethernet, and I've tried using NAT and bridging the connection, to no avail. I also typed dmesg, but nothing stood out to me in the output. I'd love to paste the output from that command here, but shared folders don't seem to be working in the guest, even though I installed VMWare tools. Whenever I run the Perl config script I get an error message:

Guest vmxnet fast network interface: failed
Unable to start services for VMWare Tools

That's the only error I see, and I can see the driver:

prtconf -D | grep 1022
pci1022,2000 (driver name: vmxnet)

—Preceding unsigned comment added by () John 44 fhm (talkcontribs) 17:26, 21 March 2009 (UTC)[reply]

blocked mediawiki export, api.php - gfdl freedoms[edit]

One wiki (not wikimedia related) blocks mediawiki export page, api.php and does not provide any database dumps. How can I download whole wiki (it has hunderdeds of articles) with full edit history? Person who hosts it says that he will block ip range if anyone tries to download too many articles. If a wiki is on gfdl, mustn't it provide a way to download it as a whole, not just parts? 217.75.59.5 (talk) 19:51, 21 March 2009 (UTC)[reply]

Not all wikis are GFDL licensed, but wikis licensed under the GFDL must provide a machine-readable version of published materials. See WP:GFDL for text of license. --h2g2bob (talk) 20:34, 21 March 2009 (UTC)[reply]
I think having per-article source available is enough for that though. Algebraist 20:36, 21 March 2009 (UTC)[reply]
The hoster may be worried about performance or cost. You will have to reduce your impact, possibly by downloading the articles slowly, perhaps one per minute, rather than one per second! Graeme Bartlett (talk) 21:24, 21 March 2009 (UTC)[reply]
There's nothing in the GFDL that requires you to allow people to download a complete database, or to download it as a whole rather than as parts. --98.217.14.211 (talk) 23:11, 21 March 2009 (UTC)[reply]
OK, thanks. So i guess we can only make sure everyone of the editors know that wiki is 100% owned by evil admin (who won't give dumps), and persuade at least some of them to move to another wiki. (last free dump is not too much old). It is not really possible to dump automatically, because admin would easily make little changes to obstruct dump-tool. 217.75.59.5 (talk) 23:58, 21 March 2009 (UTC)[reply]

Is 10-to-1 still valid?[edit]

Back in the days when a 1200-baud Silent 700 was the leading edge of technology, all of us in the Time-sharing business knew that even though a byte contained 8 bits, asynchronous communication added a bit on each end of the data (start and stop bits), so that "1200 bits per second" really delivered only 120 characters per second rather than 150 as might be naively assumed. On the other hand, if you had a big-ass synchronous modem, you could send and receive blocks of 400 or 800 characters at 4800 baud, true 8 bits per character minus some block-level overhead (but much less than the 20% of async "interactive").

Are today's communication rates on the phone line, over DSL and cable and so forth, still 10 bits per character, start + 8 + stop? (I assume the answer is Yes using 56K dialup.) As I sit here browsing the web and downloading photos, does 120KB/s still mean 1200Kb/s?

--DaHorsesMouth (talk) 20:17, 21 March 2009 (UTC)[reply]

For ADSL things are rather complex, as the spectrum used is divided into "bins" that have different characteristics; the ADSL article notes "The total maximum capacity derived from summing the bits-per-bins is reported by DSL modems and is sometimes termed sync rate. This will always be rather misleading as the true maximum link capacity for user data transfer rate will be significantly lower because extra data is transmitted that is termed protocol overhead, a reduced figure of around 84-87% at most for PPPoA connections being a common example". Much of that is due to the structure of ATM frames, which have 5 bytes of overhead for every 48 bytes transmitted. Over and above that (in ADSL, cable, and the earlier dialup solutions you mention) have additional overheads for IP and TCP packet headers (overheads that are per-packet, which means end-to-end efficiency differs greatly between bulk transfers and smaller packets). 87.114.29.204 (talk) 20:49, 21 March 2009 (UTC)[reply]
The encoding schemes on DSL and cable don't remotely resemble those old dialup modems. Modern dialup modems (that use V.42 and so forth) don't work like that either. They delete stop/start bits and do fancier data compression as well. 75.62.6.87 (talk) 10:35, 25 March 2009 (UTC)[reply]

Intel Core 2 Duo processors vs Centrino 2 Duo[edit]

Good Morning and thank you for being available to ask questions. Laptops are being sold with either Intel Core 2 Duo processors or Centrino 2 Duo processors. Could you please tell me which is better in a way that an end user will notice. I did look at the article http://en.wikipedia.org/wiki/Comparison_of_Intel_processors. It lists processor features but not Centrino or how each new model is better than the last. Many thanks Cadbury29 (talk) 21:38, 21 March 2009 (UTC)[reply]

Centrino isn't a processor, but rather Intel's chipset for laptops - the Centrino article says "There is usually a lot of confusion with Core 2 Duo and Centrino Duo. Centrino Duo is the combination of Core 2 Duo and Intel Next Gen Wireless Systems." 87.114.29.204 (talk) 21:47, 21 March 2009 (UTC)[reply]

My apologies for missing this on the Centrino article. Does the combination improve the user's computer usage experience whilst improving battery life or does it slow the computer down in comparison with a standard Core 2 Duo? Thank you again. Cadbury29 (talk) 21:57, 21 March 2009 (UTC)[reply]

You're still not seeing what Centrino is. It's a brand of chipset. All PC CPUs need an accompanying chipset; you can't run a Core 2 Duo without a chipset; if you didn't install Centrino you'd have to have the equivalent offering from nVidia or VIA (or you'd have an AMD processor with an AMD chipset). So you can't run a "standard core 2 duo" alone; there's no comparison to make. We don't have articles comparing mobile chipsets from different vendors, it seems, but there's generally not a lot in it between Intel and nVidia at least. 87.114.29.204 (talk) 22:14, 21 March 2009 (UTC)[reply]
It's like saying, a Windows PC vs. a Dell PC, which one is better? --76.167.241.45 (talk) 23:42, 21 March 2009 (UTC)[reply]
Assuming, of course, that the Dell PC runs Windows (they do have some Linux versions now). Thanks, Genius101Guestbook 15:26, 22 March 2009 (UTC)[reply]

Centrino is simply a marketing "certification" All it means is the devices has an approved A) CPU, B) Chipset, C) Wifi Card. All Intel laptops will have core 2 processor and perform the same, however some may include a dell other brand wifi card instead of an intel one, for example, or maybe use a different chipset, perhaps one from nVidia. That one change will make the laptop no longer "centrino" but other than that it's all the same. 204.16.236.254 (talk) 22:28, 27 March 2009 (UTC)[reply]