Wikipedia:Reference desk/Archives/Computing/2009 July 25

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Computing desk
< July 24 << Jun | July | Aug >> July 26 >
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.


July 25[edit]

Internet causing computer crashes 2[edit]

Posted this earlier (just search for a question without the 2 at the end of title--) and someone suggested using Malware Anti-Malware so I finally got around to doing that (the computer I'm trying to fix isn't mine...) and it hasn't resolved my problem. Here's the original problem:

My computer crashes sometimes when I use the internet- completly crashes. Whatever site I'm on is open, nothing looks different, except everything is frozen and must turn off power using power button. It happens when I use Firefox, IE, on different websites (some don't even have flash or anything- last time I used the computer it crashed on espn.go.com)

I've ran Malware and it found a RougeDriveCleaner which I doubt caused the problem. Before Malware I ran I ran A-2 (free) squared and it found 2 viruses (Riskware.gen.Nero!IK, Trojan.BAT.Delfiles). After running A-squared I could use explorer.exe; before it would give me DEP and explorer would die everytime I tried to open it (even to view a folder.) Not sure if any of those are really problems but I deleted them anyway since they were in files and programs I didn't need. Ran some others-Avira(free), SuperAnti-Spyware (free) (it found 2 virus and about 300 tracing cookies also but unfortunatly its log somehow got erased..) I could post the HijackThis! log but I'm not so sure it would help. Given that it's sporadic I'm guessing that the actually virus (assuming its a virus) isn't always running. When it does run my computer crashes so I need to restart. Could there be a reason outside a virus? Computeridiot34 (talk) 00:22, 25 July 2009 (UTC)

I would run MalwareBytes AntiMalware and Spybot Search and Destroy (both free) and a free online antivirus scan such as Trend Micro HouseCall. This will least rule out the possibility of a virus or malware problem. Next I would then look at what programs are set to run on start-up using msconfig and services.msc (just use start --> run to execute these) as you may find there are some junk applications such as free screensavers etc that are causing the crash. Also, use Crap cleaner (free) to clear out all temporary files too. Rjwilmsi 09:56, 25 July 2009 (UTC)
It has 8 gb of temp. internet files but I'm not sure what that means. It has alot of work files downloaded as PDFs from online journals, etc. Things I can't delete. So are these 8 gb of temp. files causing the crashes? Computeridiot34 (talk) 23:35, 25 July 2009 (UTC)


I once had mine slow down to almost freezing, 6 times a day, at the same time almost to the minute. I traced it to a "normal" ad updater looking for (and finding) the website but being unable to update it (as I had it saved on the hard drive but not running in the browser). Some similar endless search may be using up all your CPU to the point of not letting any commands from you get through.
First delete your browser cache and cookies. Get and run [www.safer-networking.org Spybot]. Let it delete temporary files (these are internet cache type files, not things you've downloaded and saved with files names). Go to Advanced>Tools>System Startup. Clicking the right arrow bar will give details of known or suspected malware programs that start with the machine. Untick any suspicious ones, then reboot the computer and see if it behaves better. (If it does, go back to Spybot, untick them again and choose Delete from the top toolbar).
THEN boot into safe mode and run the anti-virus programs suggested by Rjwilmsi above. Delete anything marked BHO (browser help object) as many of these try redirecting to other sites, a good cause of internet problems.- KoolerStill (talk) 14:15, 28 July 2009 (UTC)

transferring files - tricky[edit]

What would be the easiest way to transfer a ~7 GB dmg file from a PC (Windows XP, desktop) to a Macbook (Tiger)? I have a 4 GB flash drive and a 2 GB SD card (which fits in a camera, conveniently).

I am assuming that splitting the file using 7-Zip and then transferring it would be the straight-forward solution. However, I was wondering if there is an "easier" way out. Is it possible to "open" the dmg on a PC as on a mac? I mean reveal the contents of the package so that I know what I need and what I don't ... thanks. Kushal (talk) 02:25, 25 July 2009 (UTC)

How about connecting them to the same network, and then transferring the files over AIM or something like that? Our Apple Disk Image article talks about the .dmg file format; in particular, you might be interested in the Apple Disk Image#Non-Macintosh section. --Spoon! (talk) 04:13, 25 July 2009 (UTC)

Thanks for the ideas. I got the ISO using the article you showed me (the Java program is awesome). I was wondering if I could change contents of the package and repackage it so I can run the dmg off of a 4 GB flash drive. Any ideas? Kushal (talk) 19:23, 25 July 2009 (UTC)

Archivers (7zip, WinRAR, etc.) can split archives into user-defined sized chunks which can then fit on your flash drive. --antilivedT | C | G 00:26, 26 July 2009 (UTC)

Antilived, thank you for the reply. I used 7 zip to create zip files and I used split and concat to join them but the resulting zip file was unusable. I am trying to do the split and join process again to see if it was some error on my part. I wish 7 zip was available on the Mac OS. Well, I will be back to report on the progress. Kushal (talk) 16:07, 26 July 2009 (UTC)

Still could not get it to work. I am looking at alternate solutions. Kushal (talk) 21:57, 29 July 2009 (UTC)
Got on the same (wired Ethernet) network and used Skype to transfer the file as per suggestion. Thanks, everyone. (New question coming up soon.) Kushal (talk) 18:09, 31 July 2009 (UTC)

Jobs-for-life in the computer industry[edit]

Let's say you are a top-dog 3D CG programmer. To keep you from getting a pink slip, you have to learn lots of things during your career (i.e., new hardware, new instructions, advanced algorithms, improved software languages/architecture ...). Most abilities in the computer industry are becoming obsolete or inadequate in a few years thanks to the progress. What are the kinds of knowledge/expertise that have withstood the changes for the longest period of time?

I went to a public library last night. I noticed that some Z-80/8080 books published in the late 1970s and early 1980s are still being borrowed by people (due dates!). I think they are useful because many 8-bit processors are used in today's embedded systems. People are making honest money coding these 30-, 40-year-old things for a very large number of diversified employees.

Let's say you're a computer technician driving a time-machine to 2009 from the 1960s or 1970s. If you rule out those single-employee, limited-demand repairing jobs, are there any today's hard- software jobs that may demand your input? -- Toytoy (talk) 05:33, 25 July 2009 (UTC)

Many major organisations still use legacy systems developed in the late 1970s. They might have been ported to newer equipment or even newer operating systems, but if the program code is stable and it still does what is required then it could easily still be in operation. Those organisation are reluctant to lose their investment in years of development, for something new which might be years before it is as stable. Therefore you will still find organisations using a program written in COBOL and running on IBM mainframe systems; and they will have one or two people already on staff who are probably the only ones who know anything about how it works. Astronaut (talk) 09:26, 25 July 2009 (UTC)
Another thing to consider is to position yourself as the only one who knows how a particularly critical system works. For example, I used to know the guy who developed part of the communications system that is used in around 60% of the world's lotteries. It is a small role, but he is needed whenever a new lottery system is setup or an existing system undergoes a major upgrade. Astronaut (talk) 09:35, 25 July 2009 (UTC)
I think Z-80 programming is unlike COBOL. Being a COBOL programmer today is like working for a museum -- generally you don't make new things. In contrast, a new model of microwave oven may require new functions written in Z-80 assembly language from scratch. You probably can change your job because many industries still use these 8-bit chips. You job is not repairing a particular legacy system. -- Toytoy (talk) 15:25, 25 July 2009 (UTC)
You should think about computer theory (properly, "computer science", but that has become a loaded term). Programming has changed very significantly, and will continue to do so, but the structure and methodology of computer system design has actually not changed very much since its first incarnations. From the standpoint of hardware, the things which used to be reserved for very expensive mainframe computers are now available on every desktop, so this has enabled designers to rearrange the way they do computing; but overall, the same basic concepts of system analysis are the same. Being a critical thinker is more important than knowing the ins and outs of this week's latest fad scripting language. Nimur (talk) 16:07, 25 July 2009 (UTC)

Tracing of an email address[edit]

In the recent case of Darryn Walker, who was unsuccessfully prosecuted in the UK for writing an "obscene" story on a text-based story website, how was the author traced? The article says that he was "reportedly traced" through his email address, but I believe this was an anonymous Yahoo or Hotmail address. There was no IP address visible on the site. So how is it possible to trace an individual through their use of an anonymous email address? --Richardrj talk email 08:06, 25 July 2009 (UTC)

Even if the user has an anonymous e-mail address in which they provide no personal details when they sign up, if they access that e-mail address from their home computer the e-mail provider can log the IP address of their home computer, and the ISP that provides their Internet service can link that IP back to the bank account that pays for their Internet connection. In that way the user could be traced. Rjwilmsi 09:50, 25 July 2009 (UTC)
Thanks for that. So what are the privacy policies of email providers and ISPs in this regard? Do email providers log the IP addresses of their users, and if so, do they pass on those details to ISPs? And, to add the final link in the chain, do ISPs pass on those details to the police when they are asked to do so? --Richardrj talk email 13:52, 25 July 2009 (UTC)
They all do log the IPs and they pass them on to law enforcement agencies with subpoenas, usually. --98.217.14.211 (talk) 14:07, 25 July 2009 (UTC)
Depends on the country you are querying about, and our answer depends on whether the police/government behave according to law or not. Tempshill (talk) 05:41, 26 July 2009 (UTC)

Man day in software[edit]

If a software project is said to be completed in 20 man days, does it mean the entire project was completed by 1 person in 20 days? Or is it that x number of people worked on this project for 20 days? I would appreciate a quick reply. Thanks!--117.196.133.19 (talk) 09:21, 25 July 2009 (UTC)

It means the equivalent 20 days of work, if carried out by one person. Whether it was in fact 1 person for 20 days, 2 for 10, or 40 people for half a day is not specified. See man-hour for more. Rjwilmsi 09:44, 25 July 2009 (UTC)
In project planning, a man-day is the amount of work done by one person in one day (I'm surprised we don't have an article about that). The day is usually defined as 6 - 7.5 man-hours depending on the hours of work and the length of the lunch break. A man-year is typically 240 man-days (ie. it excludes weekends, vacations, sickday). In your example above, it means 1 person worked alone for 20 days, or 2 people worked together for 10 days, or 4 people worked together for 5 days, or 20 people worked together for 1 day. Astronaut (talk) 09:46, 25 July 2009 (UTC)
We have man-hour, and man-week and man-month redirects to it (it's more generally written). So I've redirected man-day to man-hour; that article could do with a more generic name, but hypothetical worker abstract work units doesn't trip off the tongue. -- Finlay McWalter Talk 13:52, 25 July 2009 (UTC)
I vote for hypothetical worker abstract work units to be made an article, or at least redirect to the other one.
Nobody has linked to the Mythical Man-Month yet? This famous book, written by an IBM software project manager, blasted the idea that (1 person x 20 days) = (20 people x 1 day), or that software projects can even be measured in these terms at all. Nimur (talk) 16:09, 25 July 2009 (UTC)
The classic rebuttal is the case of pregnancy. Technically it might be ~9 man-months of effort, but 10 men (or even 10 women ;-) are not going to be able to produce a baby in under a month, no matter how hard they try. -- 128.104.112.87 (talk) 18:28, 25 July 2009 (UTC)

Hooking speakers up to laptop[edit]

I just got a set of 5.1 speakers, because I had "points" to redeem through Westlaw and it seemed like the best thing of all the crap in their little online store. I don't particularly need 5.1 sound, just basic stereo with the sub would be fine with me. The speakers have three different colored male cords to hook into your computer/TV, but my laptop has only one output thing for speakers/headset. Can I just plug the black cord into my laptop or could that cause problems somehow? I don't want to buy an external sound card or anything fancy. If these speakers won't work I'll just sell them or give them away. Calliopejen1 (talk) 14:16, 25 July 2009 (UTC)

The "Green" is color code for primary (stereo) speaker. Plug that into the "headphone jack" for the standard sound experience. You can also buy external stuff (for example, a USB attachment) which will have a breakout for each speaker. Remember always, though - the sound quality is dictated by the weakest link on your audio chain - your source material, your laptop's audio card, your cable, your speakers, etc. If you're watching web TV, there's only stereo sound anyway, so you aren't missing anything. Nimur (talk) 16:13, 25 July 2009 (UTC)
None of those plugs should cause problems, but you probably want to plug in the primary speakers, not the side-speakers. Nimur (talk) 16:15, 25 July 2009 (UTC)
Ok thanks! 209.6.22.105 (talk) 18:29, 25 July 2009 (UTC)

server blocked by nat[edit]

How can I access a web server on a computer which is connected to my local network via NAT? Is there like a free program which you could run on both computers to let them speak to each other?

Do you have administrative access on the NAT? It sounds like you need port forwarding. Otherwise, you could try setting up a VPN and tunneling connections, but this is not easy. (My lousy experience with OpenVPN last summer led me to conclude that commercial VPN software is worth the exorbitant costs that they charge large organizations). Nimur (talk) 16:17, 25 July 2009 (UTC)
Wait, now I'm confused - you are already on your local network? Is the computer already running web server software, like Apache (server)? Then, you should be able to use the local IP address or DNS name of that computer directly. The NAT should only matter if your client is outside the local network. Nimur (talk) 16:19, 25 July 2009 (UTC)
I tried entering "localhost" into the browser on the computer without the server but it didn't work. The other computer with the server has the same ip address as this one, I checked on http://showip.net Reg556 (talk) 16:46, 25 July 2009 (UTC)
You should read Network address translation. Because of the NAT, your computers appear to have the same IP address, but they actually have two separate, local addresses. If you are using a home router, these will usually be something like "192.168.1.100" (but it depends on your DHCP settings and the type of NAT/router/DHCP server you have). You need to find the actual, local IP addresses. If the systems are running windows, you can check your ip settings from the command line: Start Menu > Run > cmd, and type ipconfig /all, which will list your local IPs. On Linux/Unix, you can type ifconfig in a terminal. "localhost" will not work - it always refers to the current machine, and does not perform any transactions over the network. Nimur (talk) 17:04, 25 July 2009 (UTC)
What kind of "access" do you require? Can't you just point your browser to the address you get from your showip.net or whatever? --Spoon! (talk) 18:09, 25 July 2009 (UTC)
No because I get the same ip address from showip.net for both computers. I've tried all these suggestions. Is there not a simple program I could install on both computers that would talk to each other and make the link for me?
I think your problem is still that you have not found the local ip address. Any "what-is-my-ip" website will be totally unable to find this for you. You need to check the machines or the router to see how the NAT is mapping individual computers to local IP addresses - note that this is not the same IP you see on the IP look-up website. That is the external IP address, assigned to your cable-modem or DSL modem. The NAT maps that back to different, unique local IP addresses for each computer on your network. You must use this local IP address to connect; or you can set up port-forwarding on the router to do the mapping. I described a method above for finding the local IP - do you need more help with that? Nimur (talk) 22:19, 25 July 2009 (UTC)
I think I do need more help, I tried entering ipconfig /all and it returned a list of ip address for things like gateway and stuff. I tried entering each ip address into the browser but none of them connected to the server on the other computer Reg556 (talk) 05:22, 26 July 2009 (UTC)
We have an article on ipconfig. Take a look at this screenshot. You want to find your IP address (located directly above the Subnet). (There is a lot of other information you do not need). The IP listed here is the local IP address of the machine. While on the local network, this is the best way to access the machine. Find the IP address of the server; try connecting to that address using the browser on the other (client) machine. If it is still not working, verify that the server is actually on (both the server hardware and the server software). Also check if there is a software firewall on the server, which might be blocking access. Check if you can access the server while you are sitting at the server's terminal. Sometimes a network ping is the best way to check connectivity; that lets you isolate networking trouble from server-software trouble. Note that software firewalls can also interfere with Ping. Nimur (talk) 16:31, 26 July 2009 (UTC)

What is this system called?[edit]

[1] Under "Customize results". In Firefox 3, you can adjust the properties of the address bar in an about:config setting. It is only one value but that one value can be used to make many different combinations of properties. The system (the blog author calls it a bitmap, but I can't find any other references to it) uses numbers which are powers of 2 to represent the options. Once you choose which options you want, you add up the values to get your combination. 24.6.46.177 (talk) 16:42, 25 July 2009 (UTC)

Or, even better, you "or" them together (so that noting bad happens if you "add" an option already "on"). This technique is very common in programming. But I do not know of any name for it... --Andreas Rejbrand (talk) 17:01, 25 July 2009 (UTC)
Yeah, it's either called a bitmap or a bit field. It's an extremely common pattern in systems programming and computer-hardware, where one bit (or several adjacent bits) represents a single value, and so a given byte or word compactly stores a bunch of values. -- Finlay McWalter Talk 17:03, 25 July 2009 (UTC)
Yes also see Flag word - "word" here means a collection of bits (0 or 1), and flag - see Flag (computing) - presumably from the ability of a flag (realworld) to be either up or down. (Potentially "flag word" and "bit field" should be merged - if anyone want to tag them?)83.100.250.79 (talk) 17:22, 25 July 2009 (UTC)
Yes, flag word is better, although it seems to talk only about cases where each logical unit is a single bit. It's common to find cases like:
              bit0   - output enable/disable
              bit1-3 - output gain (0..7)
              bit4   - output buffer interrupt enable  
              bit5-6 - output buffer interrupt threshold (0..3)
              bit7   - reserved
... and we don't really have an article that matches this (very common) pattern. I don't think bitfield and flag word should me merged. Flag word is the hardware thing, bitfield a software construct. While you'd often use a bitfield to address a flag word, you could also do it just with masks and shifts, and equally a bitfield can be used as a purely software construct (as in the case the OP is talking about). -- Finlay McWalter Talk 17:39, 25 July 2009 (UTC)
I can't find the article either, I've added each to the 'see also' section of the other and left it at that.83.100.250.79 (talk) 18:39, 25 July 2009 (UTC)
While we're naming alternative terminology, these are sometimes called bit vectors or registers, especially in hardware description languages like verilog and VHDL that support nonuniform word lengths. Nimur (talk) 03:22, 26 July 2009 (UTC)

Subscriptions to a blog[edit]

Is there a way to set up a Blogger blog so that people can just click a link to be notified of new posts to the blog via email? I see that a reader/subscriber can do this if they use Google Reader but that may not be what everyone uses for their email. And it means that everyone has to go through the trouble of signing up with a special service just to find out about updates to the blog when they could just visit the blog and find out. Alternatively, is there a blogging service that does allow users to subscribe in a way like what I want? Dismas|(talk) 17:43, 25 July 2009 (UTC)

It's hard to imagine a practical implementation of that that wouldn't require a bunch of steps from the user, which would amount to a signup. First they have to type in their email address, then probably fill out a CAPTCHA (to avoid the system from being used to mailbomb people), and then reply to a confirmatory email (again to avoid the system from being abused to sign up unwitting third parties to huge lists of unwanted chatter). That's pretty much the same sign-up process for a regular RSS aggregator like Bloglines or Google Reader, so they might as well just use one of those and get all the associated flexibility. Now you could do this as a client side feature (with an RSS reader program or a firefox plugin), but that's obviously not a zero-install-zero-setup option either. -- Finlay McWalter Talk 18:02, 25 July 2009 (UTC)
Okay, so submitting, going through CAPTCHA, and confirming is fine. If the instructions included going through a signup or download of an aggregator, then I see that as too much complication that people may not be willing to go through for a single blog. Dismas|(talk) 20:09, 25 July 2009 (UTC)
Doesn't Google Reader allow you to configure an RSS aggregator to send email updates? This only requires that the blog supplies RSS, and that you are willing to use Google's service. Nimur (talk) 03:29, 26 July 2009 (UTC)

converting raw fils[edit]

So I've got a dslr and am shooting in RAW (good), and am taking pictures of steam engines. I'm having a problem with the exposure - or rather converting the RAW file to jpeg. The problem is that the subject is pretty dark, but the sky is full of nice fluffy clouds. By moving the exposure slider up and down I can either get picture of black engine with completely washed out sky, or a nice picture of the clouds but an underexposed subject. I can try hdr, but that seems to produce dramatic images rather than realistic looking images. Ideally I want to open the raw file, select different areas of the and apply different exposure values to those areas only but ensure that the joins between areas are smooth. I only have ZoomBrowser EX and Digital Photo Professional. But do I need Photoshop Elements to do this? —Preceding unsigned comment added by 86.144.124.80 (talk) 19:51, 25 July 2009 (UTC)

The HDR images you see online are mostly gaudy examples where they've taken things to an extreme - it is possible to use the technique with a degree if subtlety (e.g. some of the photos in this set - http://www.flickr.com/photos/phototoasty/sets/72157594225970774/) and it seems almost everyone oversaturates their HDR photos (all the better to emphasise their bad trip appearance). Anyway, if you want to be editing the files yourself, you'll find that The Gimp and dcraw will read the various RAW formats for many popular cameras (RAW isn't a format per se, just whatever data comes out of a particular camera's sensor, so you need software that knows about your camera). More stuff is listed at raw image format#Software support; with those you should be able to convert your camera's RAW (assuming it's supported) into TIFF or PNG and you can edit that in whatever you're comfortable with. -- Finlay McWalter Talk 20:50, 25 July 2009 (UTC)
Another good software to do this is hugin (software). Export your RAW file to different exposures (eg. +1, 0, -1), and follow this tutorial (ignore the stitching part, you don't need it). It creates some quite subtle HDR images but tends to require some contrast enhancement. --antilivedT | C | G 00:21, 26 July 2009 (UTC)

Can GIMP open RAW files? GIMP is a free, free alternative to Photoshop. Nimur (talk) 03:33, 26 July 2009 (UTC)

The OP needs a program which can process the colour depth present in the RAW files, which is more than 8 bits per channel. The GIMP is limited to 8 bits per colour channel, whereas Photoshop can process 16 bits per channel. I don't think the SE version of Phohotoshop does 16-bit per channel color, though. An option would be to convert the RAW files to high-color-depth TIFF files with dcraw, and read these with Photoshop. There is a GIMP fork called CinePaint which can handle high colour depths, but it's only available for Linux and Mac, not for Windows. I've tried it on a Debian machine (at SteveBaker's recommendation), but found it rather unstable, and it was later removed from the repositories because of its bugginess. The dcraw page has a link that might be of interest, I followed it, ended up here, and read about the programs Zero Noise and Perfect Raw. I haven't tried them, but intend to when I get the time. Unfortunately, some of the info appears to be in Spanish only. --NorwegianBlue talk 07:55, 26 July 2009 (UTC)
This article http://www.wikihow.com/Take-High-Dynamic-Range-Photographs and this http://qtpfsgui.sourceforge.net/about.php program may also be of interest. --NorwegianBlue talk 21:43, 26 July 2009 (UTC)

Broken link[edit]

http://en.wikipedia.org/wiki/Sonoran_Desert - this page was accessible up until about a week ago. Now it just generates page load errors. —Preceding unsigned comment added by Clhowson (talkcontribs) 20:14, 25 July 2009 (UTC)

Works for me. Try clearing your browser cache. -- Finlay McWalter Talk 20:52, 25 July 2009 (UTC)

Core 2 Quade 9400 and Vista 64 SP 1[edit]

Recently, I have upgraded my hardware. My new processor Core 2 Quad 9400 requires Vista 64 bit SP1. I found this info from its official. Right now I'm using XP SP 2. Should I install Vista 64 bit SP1 for better CPU Performance or stick with current XP SP 2. I have not used Vista before. My motherboard is 750i SLI Nvidia Geforce and Graphics card is Sapphire ATI 4890. Any suggestion? —Preceding unsigned comment added by 119.30.36.44 (talk) 21:05, 25 July 2009 (UTC)

Surely it doesn't require vista; IA64 will run 32bit OSes perfectly fine. For an ordinary user, the most obvious advantage of switching to a 64 bit OS is that it allows you to address more than 4GB of RAM; while some operations are certainly faster, some others are (in practice) slower, as the wide IA64 instructions are larger, making for poorer use of cpu cache. Assuming everything is working for you now, and there's nothing that you need Vista for, personally I'd stick with XP until Windows 7 comes out (it seems to have all of Vista's improvements and few if any of its demerits). -- Finlay McWalter Talk 21:24, 25 July 2009 (UTC)
If you have something that requires Vista, get something else. HalfShadow 21:28, 25 July 2009 (UTC)
To the best of my knowledge, Intel Core 2 CPUs implement x86-64 rather than IA-64. -- Meni Rosenfeld (talk) 21:04, 26 July 2009 (UTC)
Your CPU does not require Vista. You can run XP, Windows 2000, or Windows 1.0 if you feel like it. Tempshill (talk) 05:38, 26 July 2009 (UTC)
It's possible that the Core 2 Quad 9400 "requires Vista 64-bit SP1" in the sense that it will not work properly in the original 64-bit Vista because of some compatibility bug that was fixed in the service pack (SP1). In principle you might have hardware (not the CPU) that only has drivers for Vista 64-bit, but almost certainly you will be able to find 32-bit XP drivers for everything. -- BenRG (talk) 13:16, 26 July 2009 (UTC)

any way to delete iPod content from WITHIN iPhone?[edit]

So my iPhone is full but I wanna download some apps. Is there any way I can DELETE some of my music from WITHIN iPhone (ie without connecting it to anything).

Thanks!!!

ps. the reason i'm asking is that i just have an old computer with usb 1, all this content is from someone else's computer but I don't want to lose it. thanks. 82.234.207.120 (talk) 21:22, 25 July 2009 (UTC)

Don't think so. The whole point of having to connect to a computer is so nothing gets deleted accidentally. HalfShadow 21:24, 25 July 2009 (UTC)
Can you do it if you have a shell on the iPhone?F (talk) 09:35, 26 July 2009 (UTC)

Fixed /dev/ path for external USB drive[edit]

I'm using Kubuntu. Each time I plug in my external USB drive, it gets a new path in /dev (e.g. /dev/sda1 one time and /dev/sdb1 the next). How do I set it to always have the same path? NeonMerlin 21:34, 25 July 2009 (UTC)

You can hack something in udev, but the easy way is not to refer to it by its /dev/sdXN name, but by its UUID, which is fixed. Try this - with the USB disk absent, ls -l /dev/disk/by-uuid > /tmp/foo. Then insert the disk, wait until it's recognised, then type ls -l /dev/disk/by-uuid | diff /tmp/foo - . This will show you the UUID of the disk and its dev-uuid. So now, rather than refer to it as /dev/sdc1, refer to it as /dev/disk/by-uuid/ABCD-1234 or whatever. You can also statically mount devices with mount or /etc/fstab by UUID (search their respective man pages for UUID for the syntax). I mount all my local hard disk partitions this way - this way I can rearrange the SATA cables or whatever and the disks all still mount correctly in the places I want. -- Finlay McWalter Talk 22:37, 25 July 2009 (UTC)
Hey, I didn't know you could mount drives like that. Thank you, that's a great tip. APL (talk) 17:56, 26 July 2009 (UTC)