Wikipedia:Reference desk/Archives/Computing/2011 November 13

= November 13 =

Tech support IRC
Does anyone know any IRC channels (preferably but not necessarily on Freenode) that provide good general tech support for computers? --121.98.230.200 (talk) 02:48, 13 November 2011 (UTC)


 * ##windows, ##linux, ##hardware, ## {insert OS name} ¦ Reisio (talk) 04:45, 13 November 2011 (UTC)

SSH connection
Suppose you enter this over a SSH connection, on one line:

/etc/init.d/network stop ; /etc/init.d/network start

and press Enter. What happens? Does the network restart? Does the host receive the first command, disconnect, and fail to receive the second? Does the result depend on how fast the network is and how fast the host executes commands? --140.180.3.244 (talk) 07:24, 13 November 2011 (UTC)


 * When you type the first "/", that character is sent over the ssh connection and then echoed back and displayed on your command interpreter window. Then you press "e" and the same thing happens. Etc. So when you finally press enter, and it gets sent to the remote host, the command interpreter already has received all the characters of both the commands. The commands are sent over the network as each character in its own network packet, not as an entire line or two separate commands.
 * However, when the shell executes the first command, and the network connection goes down, the command interpreter will probably get a hangup (SIGHUP) signal (assuming a regular Unix-like system). I'm not sure but I think this will cause a race: whether the shell manages to run the second command, or whether SIGHUP kills the shell first. I'd guess if the init scripts take a couple of seconds to run, the SIGHUP has a good chance of winning the race. 88.112.59.31 (talk) 08:18, 13 November 2011 (UTC)


 * The connection won't necessarily be lost. A TCP connection shouldn't even notice a momentary loss of connectivity in the lower network layers. If you run the above command and don't get your shell prompt back pretty much immediately, something went wrong. And even if something does go wrong, it won't necessarily result in the connection being broken immediately. TCP could keep retrying indefinitely, even if the IP layer is dropping the packets because the route to the destination is down. When the connection comes back, the connection resumes.
 * After running the above command, you should get your prompt back without much delay, assuming the new network configuration works and doesn't change the IP address of the remote machine. 67.162.90.113 (talk) 00:40, 14 November 2011 (UTC)


 * Even if the network goes down, the second command will execute because it's sent as one command to the shell, notwithstanding the fact that there's a ; indicating two separate commands. I think 88's assuming that bash will somehow drop the 2nd half of the command because the it's separate. I don't think that's the case. bash only begins to interpret the command after the carriage return, and once it's in its queue (I don't know bash internals... but you get the point) it should continue to execute it. The only exception I could think of is if it kills processes by logged out users, and if the first command caused the user to logout. The disown command could solve this problem too by making it irrelevant if the instigating user logs out.


 * The point is, the whole payload doesn't start executing until you hit the carriage return and at that point bash has the whole line, even though it's 2 commands. If bash doesn't do the second because the user logs out, that might cause the second half to fail, but that may depend on particulars. Shadowjams (talk) 05:13, 15 November 2011 (UTC)


 * I actually asked because I saw this article: http://thedailywtf.com/Articles/Remotely-Incompetent.aspx
 * which implies that the above command would kill the remote host's connection without resuming it, but doesn't explain why. I was skeptical, and couldn't figure out why the second command wouldn't execute.  --140.180.3.244 (talk) 06:30, 15 November 2011 (UTC)

Inexplicable iTunes crashes
I am attempting to rip a CD I bought at a concert to my iTunes, but the songs on the CD don't have any ID3 tags. When I view the album in iTunes, I can open the album info, but when I try to change any info, iTunes suddenly becomes unresponsive and crashes. Can anyone explain this? If it helps, I'm using iTunes 10.5 on Windows XP. – PeeJay 16:45, 13 November 2011 (UTC)


 * Have you tried reinstalling iTunes? You won't lose your music, but you will have to reauthorise your computer, and if you have an iPod, you might need to wipe it and resync with the new library. Also, do you have another computer with a CD drive? If you do and the last solution doesn't work, rip the CD with iTunes on the 2nd computer, input the ID3 info, then go to C:\Documents and Settings (or just Users if the other PC runs Vista or 7)\ \My Music\iTunes\iTunes Media\Music\ and copy the folder with the same name as the artist. iTunes adds info to the raw files when you edit their ID3 tags in iTunes. HTH, User:Bodman456 | Come talk to me or ask me a question! (I don't bite ;D) 23:00, 13 November 2011 (UTC)

Couple of website based questions
1) Firstly, if I wanted to run a small online business, buying up just something small and common, for example toys, to have boxes full of whatever delivered to my flat where I can then repackage them for delivery to my customers, would I be able to just set up the site and start selling stuff, or are there various forms to fill in, legal requirements to get dealt with and so on before I can officially start such a business?

2) Secondly, suppose I find that I share a particular interest with a group of other people and I want to start a website, with discussion forums and a section where members can upload and share short videos they have made, and so on, roughly how much would that cost to run and how long would it take someone with little experience to create such a thing?

148.197.81.179 (talk) 17:04, 13 November 2011 (UTC)


 * I numbered your questions for ease of reference.


 * 1) This depends on the jurisdiction. In most states of the US, for most products, I don't believe any paperwork is required before starting such a business.  Of course, there are regulated products, like medications, which do require quite a bit of paperwork.  Toys might also have some regulations to beware of, like lead content.  Some of those regulations may not apply if your business is below a certain size.  The other thing to consider is what type of business you want to have.  There's a sole proprietorship, partnership, S corporation, limited liability corporation, etc.  The choice can have tax implications and also can make a difference for legal liability in case you are sued (say if a kid chokes on one of your toys).  The sole proprietorship is the simplest and doesn't normally require any paperwork beyond normal tax filing at the end of the year, in the US, but also offers no tax advantages or legal liability protection.  Another way around the legal risk might be to work as an agent for other toy companies.  That is, you might not keep any in stock, but might just order toys on behalf of others, and take a cut.  A legal disclaimer should then be posted saying you take no responsibility for ensuring their safety.  This also avoids the risk of ending up with stock you can't sell (although you might want to buy one of each, so you can evaluate them and photograph them for your website).  You will lose the ability to get bulk discounts, though, but avoiding double shipping costs should make up for that.  StuRat (talk) 17:21, 13 November 2011 (UTC)


 * 2) Sounds like you want a wiki, like Wikipedia. I believe they can be set up fairly easily, getting the software for free.  The biggest expenses would be a server, which can just be your PC, and a web address, which you can get for like $10.  However, if your website attracts many visitors, then you would need multiple servers.  You could get those yourself, at a cost of thousands of dollars, or you can rent servers for a monthly charge.  Also note that nobody will know about your web site unless you advertise it in some way.  If you can do so by word of mouth, great, but paid advertising will cost quite a bit. StuRat (talk) 17:45, 13 November 2011 (UTC)

How many is 'many visitors'? Would say 500 be too many? What about 5000? 148.197.81.179 (talk) 17:52, 14 November 2011 (UTC)


 * I believe most ISPs will charge extra for an internet connection that is used for commercial purposes. It isn't just a matter of having a server.  And you can't really keep it secret, since you need a fixed internet address to do this. Looie496 (talk) 17:52, 13 November 2011 (UTC)


 * My wife did exactly that.
 * 1) She got a business license to do business. There is a lot of paperwork involved. A lot of tax headaches. She opted to pay a lawyer $300 to take care of everything for the first five years. Then, once she had years of examples of how everything worked, she could do it herself.
 * 2) She was repackaging hedgehog food. This spawned a sister website (http://hedgies.com). Even though she stopped repackaging and selling hedgehog food, the sister site still exists.
 * So, it can be done. The biggest headache is handling sales. You will be taking credit card payments. The credit card companies steal a large chunk of each payment. So, you need to find a means of gathering enough of your own sales that it is worthwhile to keep in business. -- k a i n a w &trade; 17:59, 13 November 2011 (UTC)

If you'd like some real-time help: http://webchat.freenode.net/?nick=newPerson&channels=#websites ¦ Reisio (talk) 22:08, 13 November 2011 (UTC)


 * For a beginning/small site, renting server space from a hosting farm is pretty economical. Dreamhost and Bluehost are among the two biggest names in that business right now, and plenty of small websites run their platforms on them. They're around $10 a month depending on how long you buy in advance, which is pretty cheap. I've used Bluehost for years and set up lots of other sites for other people on Bluehost; it's about as easy as it gets, and is pretty fully-featured from the outset. I'd recommend that over purchasing and setting up your own servers unless you happen to be a sysadmin in your spare time. If at some point your site gets too big for those options, it will not be hard to pay someone to migrate your files to dedicated servers. --Mr.98 (talk) 17:27, 14 November 2011 (UTC)

Compile Ubuntu kernel for ARM, while adding modules
Hi, I want to compile the current Ubuntu ARM kernel for Oneric Ocelot, but I need to add specific modules, more specifically, the ones from the CyanogenMod 7 kernel for the HP TouchPad so that the kernel will run on the HP TouchPad. Then end goal for this is to produce a working build of Ubuntu 11.10 that can run natively on the HP TouchPad. If anyone could point me in the right direction, I'd be very grateful. User:Bodman456 | Come talk to me or ask me a question! (I don't bite ;D) 23:36, 13 November 2011 (UTC)


 * Presumably you have an Intel Linux workstation already (you can do this on cygwin, but why bother). Firstly you'll need a GCC crosscompiler and binutils toolchain appropriate for the target which you can get premade  or you can build yourself ; you'll need to find which arch-specific settings to use from people who've already built for that specific platform.  Then you need to configure and build a kernel tree for the specific board - ARM's example (for a different board) are here - again you'll need to know specific settings (devices, memory, modules) for the platform. -- Finlay McWalterჷTalk 00:28, 14 November 2011 (UTC)


 * Sure you need to go through this? Cross-compiling is a PITA. Any reason this won't work for you? Nevard (talk) 04:46, 15 November 2011 (UTC)


 * It's not so much of a "need" as a "want". I know about the Ubuntu chroot, but I would prefer to have a native system, which means you don't have Ubuntu and WebOS running at the same time. User:Bodman456 | Come talk to me or ask me a question! (I don't bite ;D) 09:07, 15 November 2011 (UTC)

Windows 8 ARM & x86
Will future Windows 8 ARM programs run in x86 Windows 8 and the other way round without recompiling? --Cerlomin (talk) 23:53, 13 November 2011 (UTC)


 * This article says no. They could have chosen to support this (as Apple did when they migrated from PPC to Intel) with a translating emulator akin to QEMU, but it seems they haven't. -- Finlay McWalterჷTalk 23:59, 13 November 2011 (UTC)


 * Thank you. --Cerlomin (talk) 09:29, 14 November 2011 (UTC)