Wikipedia:Reference desk/Computing

Welcome to the computing section
of the Wikipedia reference desk.

skip to bottom

Select a section:

• Computing
• Entertainment
• Humanities
• Language
• Mathematics
• Science
• Miscellaneous
• Archives

Want a faster answer?

Main page: Help searching Wikipedia

   

How can I get my question answered?

• Select the section of the desk that best fits the general topic of your question (see the navigation column to the right).
• Post your question to only one section, providing a short header that gives the topic of your question.
• Type '~~~~' (that is, four tilde characters) at the end – this signs and dates your contribution so we know who wrote what and when.
• Don't post personal contact information – it will be removed. Any answers will be provided here.
• Please be as specific as possible, and include all relevant context – the usefulness of answers may depend on the context.
• Note:
  • We don't answer (and may remove) questions that require medical diagnosis or legal advice.
  • We don't answer requests for opinions, predictions or debate.
  • We don't do your homework for you, though we'll help you past the stuck point.
  • We don't conduct original research or provide a free source of ideas, but we'll help you find information you need.

Ready? Ask a new question!

How do I answer a question?

Main page: Wikipedia:Reference desk/Guidelines

• The best answers address the question directly, and back up facts with wikilinks and links to sources. Do not edit others' comments and do not give any medical or legal advice.

See also:

• Teahouse
• Help desk
• Village pump
• Help manual

March 28

SolidWorks

Applying a cosmetic thread to a shaft part, should it be visible as such ? Swactive (talk) 04:03, 28 March 2009 (UTC)

I don't know about SolidWorks, but in solid modeling in general, you would need to add the thread material with a union, or remove material with a subtraction. It would be more complicated if you needed to do both to create the thread (in that case you might do better to start with a smooth shaft which is either smaller than the entire thread or larger than the entire thread, so a single union or subtraction can do the whole job). Now, as for visibility, I would expect it to show unless one of the following applies:

1) You are in a simplified display mode that doesn't show that level of detail.

2) It's an internal thread and can't be seen from your current point of view. This would also depend on the display mode. A hidden line removal view, for example, would hide those threads when viewed from the far side. StuRat (talk) 04:57, 28 March 2009 (UTC)

Dispose and Finalize

In the .NET Framework many classes have a Dispose method but do not have a finalizer or their finalizer does not call Dispose. This is especially true for GUI components like Button or MenuItem. Programmers should remember to free resources by calling the Dispose method, but if they forget the resources would be leaked. Why do GUI components not implement a finalizer? Thanks in advance. --wj32 ^t/c 06:04, 28 March 2009 (UTC)

A finalizer shouldn't be used to free resources, because an object might never be finalized, or even if it does, it might be a long time until it does. --76.167.241.45 (talk) 06:42, 28 March 2009 (UTC)

No. Programmers should always call the Dispose method, but if they forget there will be a resource leak. Even Microsoft says [1]:

Note that even when you provide explicit control by way of Dispose, you should provide implicit cleanup using the Finalize method. Finalize provides a backup to prevent resources from permanently leaking if the programmer fails to call Dispose.

The problem is that Microsoft does not follow its own guidelines; they don't implement finalizers for most GUI components. --wj32 ^t/c 07:24, 28 March 2009 (UTC)

I've never understood why Java and friends go to all the trouble of providing nice garbage collection but don't make it deterministic so you can do proper resource management with RAII. It's like they consider RAM to be the only garbage that matters. --Sean 14:44, 28 March 2009 (UTC)

To make finalization deterministic, the runtime environment either would have to run the garbage collector nearly every time a reference was reassigned or went out of scope, or would have to implement some form of reference counting. Frequent execution of the garbage collector would lead to unacceptable performance. Reference counting was used successfully in Visual Basic 6 and earlier, but it has its limitations. It does not detect inaccessible objects that mutually reference each other (e.g., A references B and B references A), and it does not work well in a multithreaded environment due to the need for mutex locks when accessing the counters. These were some of the primary reasons Microsoft used to justify the switch to garbage collection in .Net. See Reference counting#Advantages and disadvantages. The end result is that if we want deterministic freeing of resources in a garbage collection environment, we are stuck with manually tracking the object's lifetime using the dispose pattern. The advantage is that if we fail or elect not to do so, the resources will eventually be freed up. -- Tcncv (talk) 05:26, 29 March 2009 (UTC)

Flash and Java on 64-bit Linux

Flash and Java plug-ins do not load for me on Firefox. Tonight I started Firefox from the command line and noticed the following output, which doesn't normally appear when I start from a K-menu or quick-launch icon:

LoadPlugin: failed to initialize shared library /usr/lib/adobe-flashplugin/libflashplayer.so [/usr/lib/adobe-flashplugin/libflashplayer.so: wrong ELF class: ELFCLASS32]

LoadPlugin: failed to initialize shared library /usr/java/jre1.6.0_12/plugin/i386/ns7/libjavaplugin_oji.so [/usr/java/jre1.6.0_12/plugin/i386/ns7/libjavaplugin_oji.so: wrong ELF class: ELFCLASS32]

LoadPlugin: failed to initialize shared library /usr/lib/firefox-addons/plugins/libflashplayer.so [/usr/lib/firefox-addons/plugins/libflashplayer.so: wrong ELF class: ELFCLASS32]

LoadPlugin: failed to initialize shared library /usr/lib/adobe-flashplugin/libflashplayer.so [/usr/lib/adobe-flashplugin/libflashplayer.so: wrong ELF class: ELFCLASS32]

How do I fix this error? I'm using the 64-bit version of Kubuntu 8.04, and my processor is a Core 2 Duo, so I suspect I need to set the plug-ins to run in a backward compatibility mode. These are closed-source plug-ins of which no 64-bit version exists. NeonMerlin 07:15, 28 March 2009 (UTC)

If really no 64-bit versions for those plugins exist, then you would have to use a 32-bit version of your browser to use those plugins (you can install 32-bit programs in a 64-bit system, you know). However, this is not the case for those plugins. You can download the 64-bit Flash plugin for Linux here. And for Java, if you get the x64 version of Sun JRE (Java Runtime Environment) 6u12 or later, it will have a Java plugin. Alternately, other open-source implementations of Java like IcedTea also have plugins. --76.167.241.45 (talk) 07:31, 28 March 2009 (UTC)

It looks like those plugins are 32-bit. --wj32 ^t/c 07:42, 28 March 2009 (UTC)

I think the OP already knows that they are 32-bit, as he says "of which no 64-bit version exists." I guess the thing he needs to know is that 32-bit plugins cannot be used with 64-bit browsers; they have to be the same number of bits. --76.167.241.45 (talk) 18:29, 28 March 2009 (UTC)

The solution is to install the 32-bit firefox, which works fine on 64-bit linux and runs the 32-bit plugins without a problem. If we are lucky, there will be 64-bit plugins around the time 256-bit computing is standard. Well, if we are really lucky, nobody will be stupid enough to make websites that depend on flash and pdf by then. -- kainaw™ 22:00, 28 March 2009 (UTC)

I have 64 bit Firefox - and Flash and Java are working just fine. The Firefox Help/About... window says:

Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.1) Gecko/2008070400 SUSE/3.0.1-0.1 Firefox/3.0.1

The plugins I have:

• GCJ Web Browser Plugin (using IcedTea) 1.2
• Helix DNA Plugin: RealPlayer GD Plugin-compatible version 0.4.0.626
• Shockwave Flash 9.0 r124
• Shockwave Flash 8.0 r99. Gnash 0.8.2, the GNU Flash Player

Before there were 64bit plugins available, I used to use NSPluginWrapper to use the 32 bit plugins. Didn't work for Java, but it did work for Flash just fine. I suggest using the 64 bit plugins though, if they're available. -- JSBillings 01:11, 29 March 2009 (UTC)

Wow! Adobe released an alpha version of the 64-bit flash player on Feb 24. That is utterly amazing. I gave up on them well over a year ago. At this rate, we'll likely see a beta version by 2011 and a full-functioning version by 2013. I know that is a long time, but it is a lot faster than many other companies provide support for Linux. -- kainaw™ 21:36, 29 March 2009 (UTC)

What poor support for Linux? There's still no 64 bit Java plugin (there is a JRE but it doesn't have a plugin AFAIK) nor 64 bit Flash plugin for 64 bit Windows browsers (either IE or Firefox) as far as I am aware. :-P Nil Einne (talk) 00:43, 30 March 2009 (UTC)

Installing Microsoft Office without a CD/DVD drive

My friend has bought a new netbook (an Advent with Windows XP) to replace their old, broken laptop. They now want to install Microsoft Office but the netbook doesn't have a CD/DVD drive. Is it possible to copy the contents of their MS Office install disk onto a USB stick and install it from there? Astronaut (talk) 11:50, 28 March 2009 (UTC)

Safest approach might be to make an ISO of the CD, put that on the USB stick, then mount that with a drive emulator... --98.217.14.211 (talk) 13:57, 28 March 2009 (UTC)

Astronaut, I think you can do that. I know for a fact that you can do that with the enterprise version of Office, but I am not sure with any other versions. Try it and let us know. ---J.S (T/C/WRE) 15:40, 28 March 2009 (UTC)

You can do it easily - there is no CD-check in any version of MS Office, so you can install it from CD- or DVD-drive, flash drive, over network or even from your hard disk drive. —Preceding unsigned comment added by 194.99.216.135 (talk) 16:27, 29 March 2009 (UTC)

IIT

Moved from Miscellaneous

I am doing my third year Bachelors in technology in mechanical engineering in some crappy college in Hyderabad,India.I have a lot of technological aptitude but due to my low thinking nature and laziness I couldn't make it to Indian Institute of technology.Now my Btech aggregate may be about 50%.I may at the most raise it to 65% by the time I complete my fourth year.I dearly want to work hard for GATE and get into IIT and do my MTech in IC engines.But will IIT accept a candidate with such a low score in BTech even if I pass GATE with an excellent score.Will it sideline me for a candidate with a good BTech score and a good GATE score? —Preceding unsigned comment added by 202.63.109.14 (talk) 11:12, 28 March 2009 (UTC)

To help you out, yes it is possible, I think (but I am not sure 100%). But why do you want to go to IIT anyway. There are far better universities in the world doing far better research in IC engines. You can try to go to other country and do research (Masters) in a better university. Try to do some good projects. It might help. - DSachan (talk) 23:54, 28 March 2009 (UTC)

This was probably best left at Misc IMHO or perhaps moved to Humanities. But regardless, although I don't know how difficult it is to get in to IIT I would presume it would be rather difficult to get in to these 'far better universities with far better research research in IC engines', particularly given that for a foreign student (as I would presume is the case), the OP would almost definitely want a scholarship (else it would be rather expensive) and scholarships for Masters are usually hard to come by anyway. I'm not saying to the OP it is hopeless, rather that I wouldn't expect too much from other universities and shouldn't set your sights too high, unless you're capable of self-financing. An alternative would be to try and work in the field first before you start off you Masters since that way you would have something else other then your grades Nil Einne (talk) 00:51, 30 March 2009 (UTC)

TRS vs TS jacks for headphones and amplifiers

I was checking out some guitar headphone amps and noticed that all of them have TS, rather than TRS connectors. Is it that headphones having 1/4" TRS connectors do not work with amplifiers and musical instruments, only ones having TS connectors do? An example of the scenario I would face would be to connect a headphone having a TRS connector (not a headphone amp) to my guitar amplifier's emulated headphones jack. Would this headphone work? --Leif edling (talk) 15:21, 28 March 2009 (UTC)

• Read TRS connector; it discusses such mismatches. --jpgordon^∇∆∇∆ 19:16, 28 March 2009 (UTC)

Root access for brain in a vat

If I'm a brain in a vat connected to a computer, then how do I gain root access to this computer and open a command-line interface? NeonMerlin 17:37, 28 March 2009 (UTC)

Accessing data files ... one moment please .... //full stop. We require more input to complete your request. Please specify any peripheral devices available to you. Please specify connection mode. Please specify the desired operating system you wish to access. You may wish to consult our CLTO (Command Line Technical Officer). BIOS unavailable at this time. Assimilation may not be possible without further input. /root function disabled /root access denied - Terminating connection ..... — Ched ~ ^(yes?)/_© 18:44, 28 March 2009 (UTC)

You'd need a way to send data to the computer from the brain. I imagine something like Morse code would be doable today. That is, if the brain thought a certain thing, like "raise my (now absent) right leg", it would be detected by an electrode and register a "dot". If they try to raise their left leg, it's a "dash". Using this method the brain could send Morse code signals to the computer to get it to do what the brain wants. Perhaps we could then improve on this system and recognize 128 different signals the brain can send, and we would then have a 7-bit ASCII code, which would be just like typing at the keyboard. StuRat (talk) 23:19, 28 March 2009 (UTC)

• by the way .. I love your user page Neon - absolutely great work! :) — Ched ~ ^(yes?)/_© 00:05, 29 March 2009 (UTC)

i would like my brian to control the computer, i imagin i could surf the intert super fast and become super intelligent —Preceding unsigned comment added by 194.80.240.66 (talk) 08:06, 29 March 2009 (UTC)

There has been a very small amount of work done in this direction - Kevin Warwick at the Reading University in the UK has actually gone so far as to have simple neural implants performed on himself. He has the means to control simple external devices using his neural system and the ability to 'feel' external impulses supplied by a computer. When the computer activates the implant, he feels a tingling feeling run up and down his fingers. When he thinks about moving his hand - the computer can register that and do pretty much any simple control task. He demonstrated this by operating a robot hand somewhere in the USA from his lab in the UK over the internet. In principle this would enable him to do what you're thinking of - but right now, the best you'd get would be some kind of low speed 'morse-code' type of interface. Far from becoming super-fast or super-intelligent - it's just a very slow replacement for mouse and keyboard. But we ARE getting there - and probably in 20 to 30 years we'll have something a bit more dramatic and useful. SteveBaker (talk) 17:44, 29 March 2009 (UTC)

Though, frankly, with enough experience with a keyboard and mouse-like device, one can already surf pretty dang fast. As for super-intelligent... people do not seem to grow any more intelligent just because they can look stuff up quicker. Superficially they can appear to know more facts, but that's about it... intelligence is about raw processing. You probably can't upgrade that... --98.217.14.211 (talk) 00:02, 30 March 2009 (UTC)

Series of tubes

How many tubes are in a typical user's own personal internets, if they don't have a website number? 209.42.106.46 (talk) 17:39, 28 March 2009 (UTC)

I'm sorry, 'Senator' Stevens, you should do your own homework... uhh, legislative research... Freedomlinux (talk) 19:04, 28 March 2009 (UTC)

over 9000 —Preceding unsigned comment added by 82.43.88.87 (talk) 21:50, 28 March 2009 (UTC)

WHAT 9000?! --Alinnisawest,^{Dalek Empress} (extermination requests here) 00:03, 29 March 2009 (UTC)

Usually they at least have an IP number of 127.0.0.1: the loopback, otherwise they could have uninstalled tcp/ip, it could be done on windows 98 if there was no network adapter. Graeme Bartlett (talk) 21:14, 29 March 2009 (UTC)

I can has moar trux plz? --70.54.192.144 (talk) 08:55, 3 April 2009 (UTC)

Random intractable all-browser failure on non-infected machine

I have spent many hours looking on forums for a solution, but to no avail. After about three years of smooth functioning, my T5010 eMachine suddenly developed this problem: After a period of time, all browsers will fail to bring up websites. Opening and closing a few programs will sometimes get browser functionality temporarily. Changing to a different browser does not help. Ultimately, rebooting is necessary. This is especially likely to happen when I view media, including YouTube video, or PDF files, and especially quickly if I bring up Windows Media Player.

At the same time, there is never any problem with FTP, Skype, or IM. There is no Internet connection problem.

The typical solutions (reboot the router, scan for infections) have had no effect. I have seen discussions of DNS resolution problems, but the solutions I've seen involve working with registry entries that don't match my hierarchy. There is no other problem on my machine, and no other problem with any browsers (e.g., they do not freeze or lose any options). I'm using Google Chrome and IE8 currently. I installed Cacheman on the machine some time prior to this happening, but they say there are no reports of problems like this. Uninstalling did not help. I have reinstalled it.

Can anyone point me to a resolution to this kind of problem?

Thanks for any ideas!76.212.141.50 (talk) 19:26, 28 March 2009 (UTC)

If the problem only affects your web browsers, but not other internet programs, something may be causing problems with port 80 connections. Consider checking firewalls or security settings on your computer or router. --Bavi H (talk) 01:38, 29 March 2009 (UTC)

Well, first you should probably narrow down whether it's a DNS problem or an HTTP problem. When you type in a website address that you haven't visited lately, does it fail immediately, or does it spend a while trying to connect before failing? If it fails immediately, it's probably a DNS problem. As a first step, I'd try changing your DNS servers in the IPv4 settings on your network connection. On the other hand, if the browser takes a while before it fails to display the website, then you probably have an HTTP connectivity problem, where your http traffic is being dropped for some reason. If this is the case, it's possible that https connections might still work. Try going to https://mail.google.com (or another https site you know) and see if that works.

I have to say, this probably won't be an easy problem to figure out. There's no reason I can think of that your media programs should break your web browsers. It could end up being a corrupted .dll, which you'd never be able to figure out without reinstalling everything on your computer. Still, don't give up, maybe it'll be something easy to fix. Indeterminate (talk) 01:59, 29 March 2009 (UTC)

Uninstall any firewall software, then from the run window:

• netsh winsock reset catalog

Reboot. 24.76.160.236 (talk) 06:36, 29 March 2009 (UTC)

Uninstalling the firewall is generally a bad idea. --h2g2bob (talk) 16:12, 29 March 2009 (UTC)

Thanks for the ideas. Based on your comments and what's happening, then it appears to be a DNS or DLL problem. I tried DNS fixes without result. So maybe I'll be reformatting the HD. But does this effect ever come from Internet providers squeezing bandwith when users download a lot of video or other media? I was watching a lot of YouTube and DemocracyNow, etc.76.192.217.35 (talk) 19:02, 1 April 2009 (UTC)

March 29

Out-of-office auto-replies from people I've never emailed

Lately I've been receiving a lot of out-of-office auto-replies from people I've never emailed. I also receive tons of spam, btw, the spam folder contained more than 10,000 messages a while ago. I use a gmail account, and Windows XP most of the time, but have a couple of linux distros installed as well. Apart from these auto-replies, I have no reason to believe that my PC is malware-infested. Ad-aware and AVG report no problems. I've checked some of the auto-replies (from linux), and they appear to be genuine. So my question is, what's going on? Are spammers sending emails that appear to come from my email address? And if so, is there any benefit in using a real email address for spamming instead of a nonsense one? --NorwegianBlue ^talk 12:06, 29 March 2009 (UTC)

Yes, spammers are faking your address on their emails. They probably do find some advantage in using a real address, but they also know that the backscatter (undeliverables, out-of-office) goes back to a real address (you), and you're probably more likely to read an undeliverable-bounce, as you'd worry that it's a genuine one. To my mind email systems should be configured not to quote any (or at most a line or two) of an undeliverable message when bouncing it, but some quote the whole thing (which means they're essentially acting as unwitting spam relays). 87.114.147.43 (talk) 12:31, 29 March 2009 (UTC)

There is a principle at the heart of this - which is that while email is not a reliable delivery mechanism. it shouldn't ever irretrievably lose a message. So it bounces the entire message back to you in order that you may re-send it without having to retype it. Of course that's a rather ancient principle - from back when people used the command-line 'mail' program in UNIX. These days, most (if not all) mail clients keep copies of outgoing mail - so (in principle) it could work as you suggest. However, if you are suggesting a new system, something with end-to-end permissions would be better - where in order to send someone a message, you first have to obtain revokable permission to do so (eg via a web site). SteveBaker (talk) 15:18, 29 March 2009 (UTC)

A truly random "From" email address is very unlikely to be on a recipient's whitelist, and many people would potentially disregard random strings of characters (humans like to discern meaning); an actual person's email address is very slightly more likely to be whitelisted and conveniently avoids the "random nonsense filter" mentioned above. The reality is that it costs spammers almost nothing to use your email address in the "From" line, so from their perspective even a minute gain is worth it (this being an industry that defines a 1% conversion rate as "successful"). – 74  15:56, 29 March 2009 (UTC)

Resolved

Thanks, everyone! I did suspect it was like that, but it's nice to get my suspicion confirmed. --NorwegianBlue ^talk 17:27, 29 March 2009 (UTC)

Sending SMS via the computer.

Cheers. I got a Motorola Razr V3xx here, and a MacBook with Bluetooth. I want to send SMS from my computer, via the phone, to the number i specify. I have goSerial installed (which let's me connect and send AT-commands to my phone), as well as the ruby lib's to make this automatic later on. When I understand how to.

I know there is this post ( http://en.wikipedia.org/wiki/Motorola_Phone_AT_Commands ) about how to do it, but it just won't work. I need to see a real example, perhaps I'm doing it wrong.

Thanks in advance for help. 81.231.234.120 (talk) 15:43, 29 March 2009 (UTC)

Screen savers and tooltips don't work.

I have a Dell Inspiron 1501 running Windows Vista Home Basic. For several months, the screen saver has not come on and tool tips will not work on several programs (most notably Windows Live Messenger and Pidgin, where they show up for about half a second before disappearing, and Opera, where they don't work at all). A few days ago, both of these things randomly started working again. I have no idea what changed, and today, neither the screen saver nor tooltips work. I think it has to be a problem with the keyboard or touchpad, because even when I just click "preview" on the screen saver, it disappears immediately. But when a USB mouse is plugged in (and the touchpad is therefore disabled), the same thing happens, so that just leaves the keyboard or something else really weird. Does anyone have any idea what the problem could be and how to fix it? I really don't want to reinstall Windows, but having the tooltips and screen saver back for a couple of days made me realize how much I missed them. 99.245.16.164 (talk) 17:54, 29 March 2009 (UTC)

PC slowing down

I'm sure this will sound familiar to a lot of you and thought you might have some advice for the layman. My old PC was really slow by the time I got rid of it and when I got this new one I was happy with the way it just seemed to start up nice and quick and easy. But now I've had it for a couple of years, it seems to be getting really slow. I'm mainly talking about start up and login. Actual operation is not too bad, unless it's performing several tasks.

I'm sure this will sound familiar to a lot of you and thought you might have some advice for the layman. My old PC was really slow by the time I got rid of it and when I got this new one I was happy with the way it just seemed to start up nice and quick and easy. But now I've had it for a couple of years, it seems to be getting really slow. I'm mainly talking about start up and login. Actual operation is not too bad, unless it's performing several tasks.

So I've tried 'disk cleanup' and 'defragging' but neither seem to make much difference. I'm still running it on the 512MB RAM that it came with so I've just ordered 2GB which will hopefully notice.

I've just gone through the 'add or remove programs' thing and have removed some stuff like all the Java updates that seem to have built up. I guess I've still got Java and if I need to get any new updates then I'll soon be informed? I was cautious about removing anything else as I'm not sure exactly what I'm doing. I've been into 'msconfig' and de-selected certain programs under the startup section, so it's not loading up my Kodak/iTunes/Quicktime at startup. But was unsure about changing anything else. Is there anything else I can try to clear it out a bit. Thanks.91.109.199.184 (talk) 19:22, 29 March 2009 (UTC)

For most of three decades I have primarily been a Windows user (reality of work environment), and I've been amazed at what a difference a "clean install" can make - basically backing up data, ensuring that you have all of the licenses and preferred settings for important software, then wiping the hard drive and reinstalling Windows, applying all critical updates, installing needed software, restoring data, and setting preferences. It's a long process, but the detritus that accumulates over time really turns Windows to molasses. I've seen many people replace machines, only to find that the one they replaced is quite good once it's had a clean install. There is an (free) application called "Crap cleaner" (googling this yields the right thing: ccleaner.exe from filehippo) that has helped me reduce the accumulation of detritus in the registry, etc. ymmv --Scray (talk) 20:04, 29 March 2009 (UTC)

The PC came with windows already installed so I don't have the original installation disks or anything. I guess that makes it hard to go down this route? Is there a way of manually removing the detritus? (apart from crap cleaner)91.109.199.184 (talk) 20:30, 29 March 2009 (UTC)

If your machine came without install disks, and without a system restore disk, then it will have one of two other ways of doing a system restore. Either it will have an option to burn a system restore disk (Acers, for example, generally do this) or it will have a system restore partition on the disk (information about booting into that will be briefly displayed as the system boots). Note that such a system restore will blank your machine, so you'll need to take a backup first (but as your system is clearly in trouble, that should be your first order of business anyway). 87.114.147.43 (talk) 21:07, 29 March 2009 (UTC)

One strategy here (for deskside computer at least) is to buy a cheap second hard drive. Make a system restore disk - take your old hard drive out of the computer and put it someplace safe - then put the new one in - then use the restore disk to install the OS utterly cleanly. Once you know it boots and everything works, you can plug in your old hard drive as the second (non-bootable) drive on the system and you'll have all of your files there (including all of the junk of course). If something horrible happens and your new installation doesn't work or the restore disk is faulty or something - then you can put the old drive back as the bootable drive again and be back where you were. Now you can transfer files across to the new drive at your leisure - and only when you are satisfied that everything is working - reformat the second drive. This creates a bunch of free disk space - but it's a much more relaxed process than the backup/reformat/restore approach - and you can take your time doing it, with the computer still usable throughout the process. Once you have a second drive - you can do this again at any time in the future. It's worth the $50 or so that low-end drives cost these days. Of course, if you have a laptop - it's not so easy. SteveBaker (talk) 22:33, 29 March 2009 (UTC)

i think ccleaner will be one best option for ur problem.try this link to download this free software [2] also try start-run-%temp% &delete all unwanted temporory files.—Preceding unsigned comment added by Jawhar.vt (talk • contribs) 11:46, 31 March 2009

can my laptop write to dvds/cds

my laptop says it has "Variable speed Blu-ray Disc reader and CD/DVD multidrive" does this mean i can write to dvds and cds? or no?--Zcrazy0 (talk) 21:22, 29 March 2009 (UTC)

Usually "multidrive" or "superdrive" means it can both read and write CDs and DVDs. --98.217.14.211 (talk) 21:30, 29 March 2009 (UTC)

ah sweet thanks! i wasnt given a burning program though, so i guess ill have to get one myself and try.--Zcrazy0 (talk) 21:32, 29 March 2009 (UTC)

Try InfraRecorder. 87.114.147.43 (talk) 21:41, 29 March 2009 (UTC)

Windows (at least XP and Vista) can burn CDs and DVDs for you. (If you want more options, however, then you need another application.) --Andreas Rejbrand (talk) 00:09, 30 March 2009 (UTC)

Hibernate

As far as I can tell when a computer is put into hibernate it doesn't use any power at all; everything in memory is put onto the hdd and the computer turns off completely. Hibernate, however, also allows the computer to start up much quicker than if you were to shut down. So my question is are there any reasons why you wouldn't/shouldn't just put your pc into hibernate instead of shutting down? (Apart from, of course, certain times when shutting down really is needed) --212.120.248.41 (talk) 22:40, 29 March 2009 (UTC)

You are correct in your understanding on the way hibernate works and I see no disadvantage to using it as your normal power off mode. A big advantage is that you can leave your documents and applications open and running, although as a precaution I would suggest saving any work in progress. However, if you are using Microsoft Windows, it is also advantageous to occasionally do a full shutdown and restart. Windows has a reputation for not being a stable platform for long term continuous activity, except in limited one-application server type environments, so an occasional clean slate is good. I don't think this is needed for Linux or the Mac. -- Tcncv (talk)

So more reasons to switch OS? I've been quite fed up with how slow my computer has been lately (Windows XP, SP3), to the point I start cursing out Bill Gates each time it slows down or freezes on me. --Whip it! ^{Now whip it good!} 23:49, 29 March 2009 (UTC)

That comment was slightly POV. To "compensate", I can tell you that my Windows Vista PC is almost never shut down, and it always works fine and responds fast. --Andreas Rejbrand (talk) 00:11, 30 March 2009 (UTC)

This already was discussed right here. You have to reboot for updates and driver installs, otherwise it shouldn't matter. -- Consumed Crustacean (talk) 04:10, 30 March 2009 (UTC)

I have no malware or any suspicious software on my computer (at least that's what my anti-virus software and Spybot tell me) and everything is up-to-date. --Whip it! ^{Now whip it good!} 06:24, 30 March 2009 (UTC)

And if you're in that situation but are still having slowdown, you've got some terribly written software on there. Either a driver, a service, or something else that tends to stay in memory. As I said in the last thread, even if there's a memory leak a modern OS frees the memory when the process is terminated. Simply leaving the computer on should cause no issues. I leave my computer on / sleep it, I leave the computer at work on, there've never been issues. -- Consumed Crustacean (talk) 02:34, 2 April 2009 (UTC)

I have a Macbook and when it isn't being used it's just closed, I can't remember the last time I turned it off. You open it and it works - sometimes it's a little sluggish for a few seconds, but a good 95% of the time it just gets back on with being a great little laptop. To be fair I don't ask a huge deal from what is quite a powerful machine (just internet browsing, music storage, photo editing and some basic word-processing/spreadsheets). It must take up power being in sleep-mode though as the battery does drain a couple of % overnight. ny156uk (talk) 21:51, 30 March 2009 (UTC)

March 30

What's the most recommended ANSI/ISO C++ compiler for Windows Vista?

I'm a beginner C++ developer, I work on Windows Vista but I need to develop cross-platform application which compiler do you recommend for me?.. Are there free ones? thanks in advance Supersonic8 (talk) 00:05, 30 March 2009 (UTC)

I suspect there are more, but two come to mind.

• Microsoft Visual Studio Express – Although well known for developing GUI and web applications, it can also be used to develop console applications.
• The GNU Compiler Collection - This provides command line compile capabilities, but I believe there exist integrated development environments that can be used with GCC. Perhaps someone else can add recommendations.

I believe both of these compilers support various extensions to the ANSI/ISO C++ standard, but these can be disabled and standard compliance can be enforced with an option. Also, a few years back, I don't think any available compilers were 100% standard compliant, but the shortcomings involved some of the very technical implementation details for templates, partial specialization, and separate compilation. This may no longer be the case, but if it is, I suspect you will not notice the difference at the introductory or intermediate programming level. -- Tcncv (talk) 02:44, 30 March 2009 (UTC)

While I'm sure MSVC's strict-standards mode is fine, you might enjoy being able to use the exact same compiler on all your platforms, which you can probably do with GCC, but not MSVC. Also, you'll want to stay away from single-platform APIs and use Boost or PoCo or some other cross-platform library. --Sean 13:11, 30 March 2009 (UTC)

As mentioned, for cross-platform, I suggest GCC. If by "cross-platform" you mean "Windows Vista and Windows XP", then I suggest using Microsoft's compiler since it is optimized for Windows. -- kainaw™ 15:15, 30 March 2009 (UTC)

MSVC is pretty much Windows/XBox-only. If you really want cross-platform (by which I mean being able to compile everything from Windows to Linux to MacOS to OLPC to Sun workstations to Nintendo DS/GameBoy to Arduino to iPhone to AndroidPhone to Lego robots) - then GCC really is the only game in town...and it's free. If you plan to use is under Windows - I recommend downloading the 'CygWin' package - which includes a GCC setup. If you plan to use Linux - then GCC is likely to be installed already. SteveBaker (talk) 02:51, 31 March 2009 (UTC)

Does GCC support compiling apps as native Windows x64 yet? Nil Einne (talk) 06:05, 31 March 2009 (UTC)

I have no clue - from a quick search - I don't think so. But I don't think I've seen an actual, practical, working Windows machine running x64 yet! I mean - they obviously exist - but every Windows system I ever encounter is 32 bit. GCC certainly can compile 64 bit binaries (I've been running 64 bit Linux exclusively at home for a couple of years now - and GCC compiles every line of code for that without any problems) - but whether there is some Windows interface issues in 64 bit, I don't know. Certainly Cygwin doesn't run natively in 64 bit Windows yet - it has to use 32 bit "WOW" emulation. But whether you could compile for a 64 bit Windows target using cross-compilation under Linux or 32 bit Windows machine or something, I have absolutely no idea. SteveBaker (talk) 03:26, 1 April 2009 (UTC)

GCC will compile to x86-64 [3] [4] but not Itanium. Dog Day Today (talk) 10:10, 1 April 2009 (UTC)

InternetCalls Software

I'm referring to internetcalls; the software where you pay money and can call some countries real cheap and others free. You can store phone numbers in this software so its just a matter of clicking the green phone once the desired phone number is on display. Once the green phone is clicked, the number on display is called; the other party will then respond. After thirty seconds, I want to disconnect the phone by clicking the red phone button (which is next to the green phone so you have to move your mouse). Then, within a short time, I again want to move my mouse to the green phone and call the number. Now the recipients are a real angry bunch so if there is no voice by the first call, the phone is kept engaged on their side. So if the green phone button is clicked again, you will get a engaged tone. Ensure that within 30 seconds of the call, you disconnect by moving the mouse to the red phone button and clicking it. Repeat this for the whole day. The recipients will eventually un-engage their phone (within an hour) so since you are continually calling them, they will eventually un-engage their phone and receive the call - they will then engage their phone again.

Now this is a lot of work - to spend the whole day calling again and again. Is it possible to design something, so the procedure just designed happens automatically? I'm just interested to know whether it is possible to get the computer to do this automatically. I mean, I could have some device which presses the mouse so that the green button is pressed and the call continues - but remember, I want to call and disconnect the phone within 30 seconds of the call (to minimise money loss). Therefore, the mouse would have to move to the red phone button and back but you've also got to keep the timing in mind - the mouse has to move back and forth every 30 seconds. Does anyone know of some intelligent method to achieve it (not that I want to do it, of course, but it is some strange idea that I thought off and wondered whether it could work). No human can be at the computer. Thanks for the help and great respect for anyone who can think of a practical way to achieve this. —Preceding unsigned comment added by 58.161.138.117 (talk) 05:07, 30 March 2009 (UTC)

Could someone please inform me on this? I was interested to know if such a thing could be done (i.e basically moving and clicking the mouse automatically). I just gave an example from a particular source. —Preceding unsigned comment added by 58.161.138.117 (talk) 09:53, 30 March 2009 (UTC)

Yes, it's doable. A practical solution for this wouldn't involve the mouse at all -- it'd be accomplished with a script program. And for the record, it'd be an idiotic crime, as tracking down the culprit wouldn't be very difficult -- you'd pretty much be guaranteed to get caught. -- Captain Disdain (talk) 10:05, 30 March 2009 (UTC)

Thanks. As I said, I don't plan to do it, but just wondered whether such a prank could be accomplished. By the way, how could you do this using a script program? (What is a script program?). —Preceding unsigned comment added by 58.161.138.117 (talk) 10:45, 30 March 2009 (UTC)

In this context, a script program is basically a list of commands for a specific piece of software. In this instance, the commands would be along the lines of "dial number, wait until answer, disconnect, start the script all over again after thirty seconds". I believe Skype supports scripts, for example. In practice, the script would need to be more complex than that, of course, but you get the idea. -- Captain Disdain (talk) 12:45, 30 March 2009 (UTC)

(digressing discussion removed – 74  16:08, 31 March 2009 (UTC))

convert text to numbers

Is there any formula or facility to convert number in figures to number in words(ie. 112 should be converted to one hundred and twelve)in word or excel?. (There is a formula to convert normal numbers to roman numbers in excel.) thank you 124.43.42.154 (talk) 06:59, 30 March 2009 (UTC)

There's a site here (http://www.ozgrid.com/VBA/CurrencyToWords.htm) that shows you how through a bit of code, depends on how good you are at editing code but seems simple enough 194.221.133.226 (talk) 08:33, 30 March 2009 (UTC)

The trick is to work backwards. Start with an easier number: 123. Your output begins as "". The last 3 becomes "three". Then, the 2 is in the tens place, so you add "twenty" to make it all "twenty three". Then, the 1 is in the hundreds place, so you add "one hundred" to make it "one hundred twenty three". There are special cases to account for, such as your example of 112. Initially, the 2 will make your output "two". When you see the 1 in the tens place, you don't append to the output, you overwrite it with "twelve". Then, add "one hundred" to it to make "one hundred twelve". -- kainaw™ 14:15, 30 March 2009 (UTC)

I forgot to mention the use of functions in this. Consider 123456. After doing 456, you will have "four hundred fifty six". Now, you have 123 to work with. The function that you used to turn 456 into text is called again with 123. It will return "one hundred twenty three". You stick "thousand" on the end of that and append it to your output to get "one hundred twenty three thousand four hundred fifty six". -- kainaw™ 14:17, 30 March 2009 (UTC)

Basically - you have to get it to work reliably for 1, 2 or 3 digits - then chop your number into groups of three (starting at the right) then insert the appropriate "trillion"/"billion"/"million"/"thousand" words between each group. However, if you want to get fancy, you can turn 1,300 into either "one thousand, three hundred" or - if you are clever: "thirteen-hundred". SteveBaker (talk) 02:40, 31 March 2009 (UTC)

Help me with some download things?

I am at college and behind a proxy that does not allow me to download executables, but my personal hard drive (we get one each to use) has a virus on it which Sophos (the colleges chosen virus protection) cannot remove. I would like to download Hijack This ( [5] ) and Spybot Search and Destroy ( [6] ) to figure out what is causing it, but as I previously mentioned I cannot download exe files. I can however run them just fine and have full admin rights on the OS. Could someone please help me by downloading these two programs, putting them in a zip file (I don't have winrar), renaming the zip file, and uploading them to somewhere like Mediafire. This would be much appreceated. Gunrun (talk) 08:59, 30 March 2009 (UTC)

With our (stupid) school proxy I can simply append ?random-stuff-goes-here to the URL. As you know those are treated as parameters and are ignored by the web server (as far as I'm concerned). --wj32 ^t/c 09:45, 30 March 2009 (UTC)

OK, here's HijackThis: [7] --wj32 ^t/c 09:47, 30 March 2009 (UTC)

Much thanks Wj32. Worked like a charm although it didn't turn up anything out of the ordinary, which is worrying, as I'm almost certain I DO have a virus or something. Gunrun (talk) 10:06, 30 March 2009 (UTC)

In case you need Spybot: [8] (I couldn't use Mediafire for some reason). I'm also curious as to why HijackThis would help you. If you have a virus on your removable hard drive HijackThis wouldn't find anything because it only scans the local system configuration. Plus, it's an anti-spyware tool, not an anti-virus tool. --wj32 ^t/c 10:11, 30 March 2009 (UTC)

It's not a removable hard drive. I have a drive with its own OS on it, in a caddy that I place in the front of the PC. However I forgot to take the drive out one day and someone else used the PC. I use the word Virus to just mean something I don't want on the computer, adware, virus, worm, or spyware. But thankyou for Spybot. If I can't find this I will borrow a teachers install of AVG or something, but since a teacher isn't here right now I figure I'd try this first. Gunrun (talk) 10:23, 30 March 2009 (UTC)

To use the word "virus" to refer in general to any sort of threat can cause confusion. Use the term malware instead. :)--Xp54321 ^{(Hello! • Contribs)} 18:28, 31 March 2009 (UTC)

monochrome TV transmitter and reciever

please help me with the block diagram and description of a monochrome TV transmitter and reciever —Preceding unsigned comment added by 117.206.32.67 (talk) 12:33, 30 March 2009 (UTC)

This looks a hell of a lot like a homework questions. If so, the correct answer is based heavily on how much detail the teacher expects. Just saying "block diagram" isn't helpful. Do you need to explain exactly how the picture is interlaced? Do you need to explain how the sync is added? Do you need to discuss the tuner on the receiver side? Because it is such a general question, I have to assume the textbook and class lecture provided the boundaries for the question. -- kainaw™ 14:11, 30 March 2009 (UTC)

transmitter
(monochrome TV)

– radio waves →

receiver

    – 74  16:58, 30 March 2009 (UTC)

This same question was asked in the science desk and in the help desk. Please don't do that Nil Einne (talk) 04:56, 31 March 2009 (UTC)

capital "I" in MS Word 2007

hi,

basically when I type an "I" on MS word 2007 on vista it doesn't go straight to a capital "I" when I press space. How do I fix it?

thanks, --84.69.32.176 (talk) 18:24, 30 March 2009 (UTC)

Click the Office logo in the top-left of the window > Word Options > Proofing > AutoCorrect options... > Replace i with I > Add > OK > OK — Matt Eason ^{(Talk • Contribs)} 18:29, 30 March 2009 (UTC)

You should not need to do this manually (although it should work fine). Perhaps you simply has changed your language from English to another language? "i" is only replaced by "I" if the current text is in English. --Andreas Rejbrand (talk) 19:23, 30 March 2009 (UTC)

remote desktop

I need some advice on setting up a remote desktop system. I have a computer at home which I would like to be able to control via the internet from another computer. I would like to be able to see and run programs just as I would if I were at home. How can I go about this? My home computer has a dynamic ip address, and the second computer from which I would be controlling the first does not have administrate rights, if that makes any difference. Thank you for your help —Preceding unsigned comment added by 82.43.88.87 (talk) 19:20, 30 March 2009 (UTC)

VNC seems to be the first choice for Windows and Linux. Accessing your home computer remotely is complicated by the dynamic IP address, but a services like DynDNS gives us a useful target for remote access. My fave VNC client for Windows is (TightVNC, but I stick to static IPs and have not used DynDNS. / edg ☺ ☭ 19:50, 30 March 2009 (UTC)

Comparison of remote desktop software might be useful. I have used RealVNC and switched to the Remote Desktop feature of Windows for reasons that had nothing to do with performance. They're both excellent. You'll have to open some holes in your router firewall to let yourself in. Tempshill (talk) 20:30, 30 March 2009 (UTC)

Simple OSS

Through reading books I have learned the basics of C and C++ however there's only so much I can learn from books and I would therefore like to start looking at open source code to learn good programming practices etc. However another problem is finding an open source project which is isn't too complicated for me to understand at this point. So can anyone tell me about any relatively simple open source software I can look into? Thanks --212.120.248.41 (talk) 19:45, 30 March 2009 (UTC)

What you actually need to do is start writing code. That is the only way to learn; you will never learn by studying it (whether it's in books or reading source code). You can get a free version of Visual C++ at this link (or Visual C#, which I would recommend for beginners). Tempshill (talk) 20:27, 30 March 2009 (UTC)

Need to set up a generic printer but I don't have a driver

I was gifted with a dell pentium(?) running win98se. I did not receive any disks with it. I would like to set up a generic printer so that I can print to "FILE", but the machine keeps on asking me for the original disk. I can't "go to the printers manufacturers site to get an updated driver", because there is no manufacturer, and microsoft won't deign to support a legacy machine. The file I need is TTY.DRV, but I can't find it. the other files it asks for (dll and hlp files) would be nice, but I can ignore them. I've beeen looking all over the net for TTY.DRV, but I can;t find it. Any ideas? Phil_burnstein (talk) 20:13, 30 March 2009 (UTC)

CutePDF Writer is a free Windows printer driver compatible with Windows 98 and newer, which outputs PDF files. I bet it will suit your needs. You'll also need to find and install Ghostscript for it to work. Tempshill (talk) 20:51, 30 March 2009 (UTC)

Usually *any* printer driver will work (though a PostScript driver might provide more useful output). So just pretend you have an HP whatever-jet, download some PostScript drivers for it, and insist that Windows installs them despite the fact that the printer cannot be found. – 74  01:55, 31 March 2009 (UTC)

I also suggest you consider a PDF writer printer driver.

From previous experiments, I've found the "Generic / Text only" printer driver is designed for dot-matrix printers and is really only suitable for printing plain text documents. It will not print any images or graphics. Most programs aren't designed to cope with text-only printers: The proportional-spaced fonts on the screen are then poorly converted into monospaced printer text by printing letters on top of each other.

The output file is really a very simple set of printer commands and is unsuitable as a text document. (The output file may contain FF characters to start a new page, solo CR characters to return to the beginning of the current line for overprinting, and solo LF characters to go down a line without going to the beginning of the line. These control characters will work as intended when the file is transferred to a dot-matrix printer, but will have different effects in text editors.)

--Bavi H (talk) 07:11, 31 March 2009 (UTC)

PostScript is a standardized language used for all sorts of things (including printing). The resultant printer file from a PostScript driver can later be sent to most any PostScript printer, converted to a PDF, or opened in several image processing applications. I have personally used this method in the past and it works. (A PDF writer would also work, of course.) – 74  16:28, 31 March 2009 (UTC)

World Tour / Rock Band Instrument Compatibility

Hello, is there anyone out there using these instruments with the alternative game in Europe, specifically the UK. I am aware that there is a patch about to do this, I have seen joystiq's handy chart, but I heard someone moaning that it was a US patch only. Is this true?

Merci FreeMorpheme (talk) 20:21, 30 March 2009 (UTC)

Cloud Computing Confusion

Hello. I've been trying to figure out the whole Cloud Computing business, but there's something I still can't get my head around. As I understand it, all you basically need in order to use services from "The Cloud" is a web browser. If that's the case, then you could use your browser to "emulate" just about anything the "Cloud" will let you, like a game console (e.g. OnLive.com). Does this mean that html is able to "emulate" anything, as long as you have the computing power of the Cloud (not to mention a ridiculously fast Internet connection)? How is this possible? Aren't there limits to what html (or a web browser) can or cannot do? Also, is it really completely independent of the device you're using your web browser on? Does that mean that I could use, say, extremely old computers to do what even cutting-edge computers can't, thanks to the Cloud? I'm sure my questions sound pretty silly to you, but that's basically my confusion. I hope to get answers I can understand... :P Thanks in advance, Kreachure (talk) 23:09, 30 March 2009 (UTC)

The browser can only do browser things. That means HTML, Javascript, and plug-ins (Flash, etc.). So no—a browser can't emulate everything, at least, it wouldn't be able to without some heavy-duty plug-ins that spoke to it in browser-speak. --98.217.14.211 (talk) 00:27, 31 March 2009 (UTC)

There are many things that can be accomplished with HTML, javascript, and flash; and many of the things that cannot be accomplished are of questionable value in the first place (advanced 3D graphics, transparent interfaces, etc.). Keep in mind, however, that the more complex tricks in javascript and flash will require moderate processing power to run on your local machine. But the idea is that the "cloud" (a.k.a. server) provides the entirety of the functionality—no installation is required and "applications", "documents", and "sessions" can be shared across multiple machines, users, and connections. It might be considered the ultimate extension of "thin client" computing. Because resource requirements are limited, older computers and new lower-powered netbooks are quite capable of keeping up with advancements in "The Cloud" (and a true "cloud" computer could dispense with many of the costs and problems of modern computers). Ideally, when you are required to upgrade your "cloud" computer it would be as simple as pulling a new one out of a box and plugging it in—all your programs, documents, sessions, etc. on the "cloud" would be available automatically on the new computer. Such an environment would allow true commoditization of the personal computer. – 74  01:36, 31 March 2009 (UTC)

This doesn't sound like "the ultimate extension of thin client computing", it just sounds like ordinary thin client computing to me. This kind of thing has been around forever; the IBM 3270 is another example. The problem is that the user experience is almost always improved by pushing as much of the logic as possible to the client side. The latency and bandwidth limitations of real-world broadband connections are such that there's not much that benefits from being offloaded to a server, no matter how underpowered the client (except of course for big shared databases like Google's or Wikipedia's). I suspect that the appeal of this setup for companies like OnLive has to do with the fact that it's essentially an unbreakable form of DRM. You can't crack or copy software that's never present on your machine in the first place. -- BenRG (talk) 18:13, 31 March 2009 (UTC)

By "ultimate extesion of thin client computing" I meant that it's hard to get more abstract than connecting to a "cloud" of servers. Your "improved user experience" includes viruses, malware, crashes, slowdowns, a complicated upgrade cycle, and significant costs – I have no doubt that some users wouldn't feel that was an improvement. I don't think cloud computing will suit everyone (I personally would not use it) but it does provide significant advantages to a large portion of the population who want a computer that "just works" like a TV. – 74  18:28, 31 March 2009 (UTC)

It's not just bandwidth that's important but in some instances latency. OnLive in particular which has been treated with scepticism with many people asking whether it will work or latency will be too great an issue for most action games. Latency arises from two areas, one is from the connection, one is from the cloud computers and the processing (including compression in the case of OnLive) they have to do. However while many cloud computing apps do just use a web browser and normal plug ins, others e.g. OnLive will I'm pretty sure use specialised (but I presume multiplatform and open source and obviously free) software or plugins (What is the OnLive Game Service?. Nil Einne (talk) 04:53, 31 March 2009 (UTC)

Off topic, but why do you assume OnLive will open-source any of their software? They're a dot-com that claims to have 5,000 pages of patent filings. Tempshill (talk) 16:31, 31 March 2009 (UTC)

I'm a game developer - and while I can see where these guys are coming from - that latency issue is a severe and insurmountable problem. The speed of light is a harsh mistress. For 'twitch response' games - this flat out isn't going to work. Doubtless there are some clever tricks - like motion compensation done locally - that'll help with some classes of game - but I'm deeply skeptical. Also, when you see the number of computers and the bandwidth bill that YouTube need to keep all of those custom movies streaming more-or-less efficiently - and consider that these guys need to get all of that AND actually run game-quality graphics for thousand to hundreds of thousands of people simultaneously...it's a tough call. Then we have to consider resolution...I want my 1600x1200 60Hz display - I don't 640x480 - I don't want compression artifacts - I don't want lower frame rates or fuzzier images. I'd really like to use multiple monitors - or true-3D displays. I want "ClickBoom!" shooting - not "Click"...wait...wait..."Boom!". SteveBaker (talk) 02:58, 1 April 2009 (UTC)

March 31

Quick 'n easy wireless router question

Can you use a 802.11n router with an 802.11g wireless card? I ask because I'm trying to help my aunt hook up a wireless internet connection. I know that two of her computers have wireless n airport cards, but one of her laptops only supports G. So will that old laptop work with an N router, assuming it's close enough? DaRkAgE7_[Talk] 00:43, 31 March 2009 (UTC)

Yes. The two devices will operate at the maximum ability common to both. In this case, they will use 802.11g protocols. -- Tcncv (talk) 01:42, 31 March 2009 (UTC)

According to our article 802.11n, a "mixed-mode" network (at 2.4GHz) provides backwards compatibility with older 802.11 devices. Maximum throughput will be limited in such a configuration, but it will probably still be sufficient for most applications. (Note that you do *not* want 5GHz operation—you might have to configure the router specifically for 2.4GHz.) – 74  01:49, 31 March 2009 (UTC)

It's worth remembering that there's no such thing as a 802.11n router or wireless card yet as the standard has not been finalised. All current equipement is draft/prerelease n. From a look at the talk page, it sounds as if, at least as of mid 2008 that a lot of the draft equipement performs very poorly sometimes even worse then g equipement. As such, it may be better to just get a 802.11g router. Nil Einne (talk) 09:41, 31 March 2009 (UTC)

A lot, perhaps, but far from all. My Apple Airport Extreme (802.11 draft-n) gives me ~160 Mbit/s - way, way faster than the ~26 I got with the previous generation Airport Express (802.11g). Heck, it's fast enough that I only bother to plug in the 1 Gbps cable when I transfer more than 20GB, which doesn't happen very often. -- Aeluwas (talk) 17:33, 31 March 2009 (UTC)

I've heard that a lot of routers will default to G in a mixed-mode environment, effectively cutting the theoretical speed in half. You can get routers with multiple transmitters that will give you a separate G and N network, though. -- JSBillings 23:50, 31 March 2009 (UTC)

downloading

which is the efficient and powerfull downloading software that improves the downloading speed? —Preceding unsigned comment added by Jawhar.vt (talk • contribs) 11:15, 31 March 2009 (UTC)

There are many download managers. Check out List of download managers and Comparison of download managers. --Sean 12:04, 31 March 2009 (UTC)

Note that while many of these will help with organizing downloads, you probably wont get dramatically improved speeds with ordinary HTTP downloads. Download managers that promise that are mostly snake oil. 195.58.125.53 (talk) 16:00, 31 March 2009 (UTC)

Anon is right, "downloading acceleration software" will probably not improve your downloading speed. Tempshill (talk) 16:30, 31 March 2009 (UTC)

Right. But some also seem to surmise otherwise, so also not grokking the first answer. -- Fullstop (talk) 21:23, 31 March 2009 (UTC)

How does an operating system change HTTP download speeds? I would like to meet these "some" that can prove it. Upgrading to high speed internet gets cheaper and more available every day. Livewireo (talk) 21:35, 31 March 2009 (UTC)

Switching operating systems will not improve your download speed, unless the new OS has better-configured network settings than the old one, in which case you could get the same effect by fixing the settings on the old OS. -- BenRG (talk) 17:34, 1 April 2009 (UTC)

The questioner is likely asking about dial-up services that offer high-speed downloads over dial-up. There is nothing special going on. Assuming you download their "high speed program" on your end, the internet traffic is compressed (just a zip file) and sent to you. You decompress it on your end. The result is that you send less data over the slow telephone line but do more work on each end compressing and decompressing the data. -- kainaw™ 22:01, 31 March 2009 (UTC)

The very significant problem with those claims is that they must use 'lossless' compression - and as such, they cannot compress already-compressed data. Hence, if you are downloading JPEG photos or movies or MP3's (which are already close to optimally compressed) - these services don't (mathematically: cannot possibly) compress them significantly further. The things they do get a speedup with is text - HTML files, that kind of thing. But because an entire bookfull of text is less than a single large photo, this is speeding up the one thing that's already pretty much not a problem. So unless you are in some very specialised situation where most of what you are reading is text with very few photos (like this Wikipedia page for example) - these compressing-dialup services are a con-trick. SteveBaker (talk) 02:48, 1 April 2009 (UTC)

I think many of these services lossily recompress inline images in web pages, which can make pages load substantially faster (if you don't mind the quality loss). I can't imagine they'd get much benefit from zip-type compression over dialup lines, since dialup modems have built-in compression already (V.44).

Also, JPEG is much farther from optimally compressed than you might think. PAQ8P and Stuffit can losslessly compress JPEG files by nearly 25%. You can get similar gains on compressed PDF, and probably on MP3, ZIP, JAR (Java) and SWF (Flash) as well. Those algorithms are extremely slow, though, so slow that I think a modern CPU would have trouble with realtime decompression even at dialup speeds. They're also patent-encumbered. I seriously doubt that any of the compressing-proxy services do this kind of compression. -- BenRG (talk) 17:34, 1 April 2009 (UTC)

3DS Max 9

Does anyone know how to put different colors onto a compound object in this software, thanks. BigDunc^Talk 19:55, 31 March 2009 (UTC)

April 1

Power failure

A few years ago after a power outage, a desktop computer of mine refused to turn on. I've been leaving it in a garage until a week ago, when I decided to try to fix it. What is the most likely cause of the problem? I had no uninterruptible power supply and likely no surge protection. I've read a few blogs on the Internet that suggest the power supply unit may be fried. In that case, how hard is it to get a compatible new power supply unit? I think the one in this computer is quite unique; it connects to the motherboard with a twenty-pin connector, and the motherboard supplies the AMD CPU with power. --05:14, 1 April 2009 (UTC) —Preceding unsigned comment added by 99.237.234.104 (talk)

So you're saying it looks like this? Then it's an ATX, and should be fairly easy to replace, unless it's proprietary. (Someone else may need to correct me, I've not dealt with hardware in a while.) But if it's been in your garage for several years, it could be numerous things that have either died completely (due to dust, heat, water, etc.), or will need a little work to get back to working order. –  LATICS  talk  05:52, 1 April 2009 (UTC)

Are you fixing it to retrieve data or to recover it as a functional pc? If just to recover data you will find life much easier by simply taking the Hard-drive out of that computer, setting it to 'slave' and adding it to your existing pc provided it has input for IDE hard-drives (i'm presuming based on the age it will be IDE rather than SATA). Alternatively you can find a product online that will be quite cheap that allows you to turn any hard-drive into a 'portable' hard-drive, though not sure what they call them. ny156uk (talk) 22:26, 1 April 2009 (UTC)

Both, actually; I want it to be operational, but if that were not possible, I'd be satisfied with data recovery. Thanks for the info!

About the power supply unit: No, it's not an ATX. Here's a photo: ftp://99.237.234.104/DSC_5408.JPG The medium-size connectors have 4 large pins each, and the largest connector (the one proving power to the motherboard) has twenty hole, each with a wire leading to the power supply unit. I think in an ATX, one of the holes is not connected to the power source. --99.237.234.104 (talk) 05:55, 3 April 2009 (UTC)

The worm

Alright, I'm sure a fair amount of tech-savvy users around here have heard of the conficker worm. Here's want I want to know: First, why are they doing this on April Fool's Day, of all days? Second, why do they target Windows, as usual? Has there ever been a major worm or virus specifically targeted at Apple? I have all the latest updates installed, I have anti-virus software installed (Windows Live Onecare) and I even followed these instructions (I used Microsoft's Fix It For Me) via this blog, a link I clicked on through here. When they mention USB devices being used to pass around the worm, should one be worried when they connect a music device like an iPod to their computer? And I also read that merely googling 'conficker worm' can put you at risk, when will the danger pass? Will logging in to websites requiring email and password (such as this one) put you at risk? And how do you avoid having a panic attack, especially when you hear the French and Royal Navy have fallen victim to this?--Whip it! ^{Now whip it good!} 05:58, 1 April 2009 (UTC)

If your system is fully updated then your PC isn't at risk, but even if it wasn't googling would be safe, however clicking links from google might not be. The worm could easily pass from an ipod onto the machine, if the ipod is set to act as a hard drive. The worm targets Windows because it is the most popular OS, and since most of the security companies seem to think the worm is attempting to create a botnet they'd want as many machines in it as possible. As for the April First thing I have no idea.

Incidentally I'd suggest ditching Live Onecare, as it isn't actually very good at detecting viruses. The big 3 free antiviruses ( AVG, Avast!, and Avira) generally score better on detection rates and similar. If you want a paid for anti virus I'd highly recommend NOD32 Gunrun (talk) 10:26, 1 April 2009 (UTC)

A source on my claims about onecare by the way. Gunrun (talk) 10:29, 1 April 2009 (UTC)

I have to say that I too do start to panic when I read that military forces depend on Windows. -- Hoary (talk) 15:35, 1 April 2009 (UTC)

I would say that April Fools is the absolute best day to do it. It makes warnings seem like hoaxes. Not that the day matters too much. APL (talk) 17:38, 1 April 2009 (UTC)

The danger from USB devices comes from AutoRun, a huge security hole in the default Windows install which has been used over the years as a vector for all kinds of malware. You should disable AutoRun on all Windows machines you administer, then you won't have to worry about plugging in USB devices any more. As far as I can tell all that's happening today is that existing Conficker "installations" are switching to a new mechanism of random domain name generation. The worm has been around for months, and this change doesn't make it suddenly more dangerous. If you can access www.microsoft.com, you're not infected. -- BenRG (talk) 17:57, 1 April 2009 (UTC)

Conficker file size

How big is the Conficker worm? Dismas|^(talk) 10:42, 1 April 2009 (UTC)

According to this page one version is 58,368 bytes. -- BenRG (talk) 18:44, 1 April 2009 (UTC)

Where to purchase/Improvise a missing jack for a DVB-T stick?

I recently won a DVB stick in an ebay action. It seems to be in a good condition except the fact that a small jack that connects antenna jack to the stick is missing. Please have a look at this image from an online shop where I have marked this part: http://img24.imageshack.us/img24/8384/r970.jpg.

The diameter of the antenna jack is about twice the size of the DVB female as seen in this photo:http://img6.imageshack.us/img6/9960/42761281.jpg . If you could please suggest a way to connect these two it would be nice.

131.220.46.25 (talk) 10:56, 1 April 2009 (UTC)feld

It's hard to tell from the photos, are you looking for something like his? —Preceding unsigned comment added by 82.43.88.87 (talk) 17:27, 1 April 2009 (UTC)

Conficker-generated domains

OK, so I did a Google search on something, and some results come from domains that have a random number, e.g. "foo.bar.345.pl". Are these the Conficker-generated domains the article is talking about? Blake Gripling (talk) 12:05, 1 April 2009 (UTC)

No, Conficker uses long domain names that change every day. -- BenRG (talk) 16:49, 1 April 2009 (UTC)

To expand a bit on this, Conficker used to generate 250 unique domain names per day and now generates 50,000 per day, or around 18 million per year. I don't know how long they are but it must be more than 3 digits in order to produce that many unique names. I'm sure the names include letters as well as digits. Also, those domains aren't registered by Conficker. Rather, whoever's controlling the botnet registers a domain when they want to send instructions to the botnet. Only a few will ever be registered, and even those won't show up on web search results (since they won't be hosting a public web site, I would assume). -- BenRG (talk) 18:17, 1 April 2009 (UTC)

Need a redirect or new article/section for premium rate SMS

I'm hoping somebody here will have an answer. Basically I'm looking for the correct article to redirect the above red link to. Value-added service seemed promising, but there is no mention of VAS SMS's being charged at a premium (probably not all VAS's are charged at a premium in all countries/networks). I'm talking here about services like SMS-in competitions (Idols (South Africa)), as well as airtime recharging and other payment and subscription services (e.g. ringtone and wallpaper downloads) that are conducted by sending "premium rate" SMS's as payment. Any ideas? Is "premium rate SMS" the correct term for this concept, or should it just fall under value-added service? Is there even a standard name for this animal internationally? (In South Africa all such adverts and services are required to carry the disclaimer "SMS's charged at VAS rates".) Zunaid 12:15, 1 April 2009 (UTC)

Premium SMS redirects to Short Code, so I'd suggest you do the same thing. There is a section SMS#Premium-rated short messages, but Premium SMS and Premium rate SMS are so similar it doesn't make sense to redirect to different things. --Maltelauridsbrigge (talk) 13:41, 1 April 2009 (UTC)

Short Code doesn't provide any information on the topic, so redirecting to it would be incorrect. SMS#Premium-rated short messages comes closest, so a redirect there would seem to be best. – 74  19:31, 1 April 2009 (UTC)

Networking

Hi just let me know which one to opt is it MCSE or CCNA. which is good in IT field.

As far as I know MCSE you have to complete 6 or 7 exams to get it but with CCNA it is 1. They are both entry level if you are looking for a job in the Networking field I would say in my country Ireland that CCNP is the qualification that is most in demand. BigDunc^Talk 18:25, 1 April 2009 (UTC)

You'll obviously need to make your own decisions on this, but ... you may want to consider the A+ certs as well. — Ched :  Yes?   : ©  05:55, 2 April 2009 (UTC)

Autorun file

Hey guys

I no guru I just want to know what exactly a autorun file and what it is made of Please show where I can find the most thorough info about it

Sorry for the trouble and tanks for your time —Preceding unsigned comment added by 196.210.196.121 (talk) 15:01, 1 April 2009 (UTC)

See AutoRun 87.115.166.150 (talk) 15:11, 1 April 2009 (UTC)

Visual Studio: Convert or include C# form for VB project

I'm using Visual Studio 2005. I accidentally created a project as a C# rather than Visual Basic project, and built a form in it. How do I either convert the form into VB or allow a VB project to use it? NeonMerlin 22:34, 1 April 2009 (UTC)

Maybe try here and here. --Whip it! ^{Now whip it good!} 22:44, 1 April 2009 (UTC)

Keeping two time machine backups

I want to keep two Time Machine backups because it is more reliable. What are my options for doing so? For example, synchronisation software might be able to keep both of the backup drives up to date. If it does, then I need it to be able to keep the entire drive synchronised - the backup only comprises one of several partitions on the disc.--78.33.187.186 (talk) 23:07, 1 April 2009 (UTC)

Put it another way, could I use RAID 1 to routinely backup to either disc; and then maybe once every day or two, plug them both in by which point Leopard would automatically synchronise them? I think. I don't know much about RAID, I'm guessing :-) I have actually read the article, but it's quite a confusing concept to a newcomer. --78.33.187.162 (talk) 20:25, 2 April 2009 (UTC)

Why does Linux need a higher desktop system specification than Windows Millenium?

Especially it seems regarding graphics - see here for Ubuntu: https://help.ubuntu.com/community/Installation/SystemRequirements

The graphics on my old WinMe computer were perfectly adequate and high definition, why does Linux require more resources? Only a cut down version of Ubuntu would work on that computer according to the link above - why? What does Linux do that requires more resources? As far as I recall it even requires more than my current XP desktop computer.

Wouldnt it be sensible for Linux to get out of the loop of the rat-race of demanding greater and greater specifications for computers? 89.242.107.39 (talk) 23:37, 1 April 2009 (UTC)

The 'absolute minimum installation' at the bottom of that page has similar requirements to WinMe. If that's not small enough, then there are plenty of other Linux distributions to choose from. Ubuntu is designed to run on fairly high-end machines, but Damn Small Linux, for example, has very low requirements. Algebraist 23:47, 1 April 2009 (UTC)

Ubuntu needs a relatively recent graphics card for its "visual effects", which I find pointless and distracting (example). If you turn that stuff off, it should run on the same graphics cards as WinMe. -- BenRG (talk) 11:11, 2 April 2009 (UTC)

Also, you do realize that Ubuntu's first release came 4 years after Windows ME's release, right? To properly compare whether Ubuntu is getting itself out of the rat-race of demanding greater and greater specifications, you'd need to compare it a more recent OS like Windows XP. While not requiring the latest and greatest components seems to be wise for Ubuntu, you have to assume that the average consumer's computer is going to increase in power (measured in virtually every important metric). Thus to make a comparable OS, you will need to increase the minimum requirements. BenRG's comment about turning off the visual effects is also important. If your computer has unusual requirements (like being underpowerwed or requiring extra security), then customization is going to be required for virtually any OS.--droptone (talk) 11:46, 2 April 2009 (UTC)

Not Windows! It'll remain reliably slow and insecure no matter what customization you perform! (j/k) – 74  01:59, 3 April 2009 (UTC)

What an utterly bizarre claim! Linux excels at running on vastly older, simpler hardware than Windows. Especially for graphics - Linux still runs just fine on machines with no graphics cards at all! My family file server sits in my garage - it has no graphics, no keyboard or mouse whatever - an ancient 600MHz CPU and runs OpenSuSE 10.3...no problems. SteveBaker (talk) 01:52, 3 April 2009 (UTC)

April 2

Automatic recording on VCR

I've got a VCR and plan to go away on holidays for 2 months. Now my VCR can only record one time in advance - you can only set one future time for recording. But my favorite T.V series comes once a week on television so out of 40 shows, I will only be able to record one. One cannot access the shows online and nor is it feasible to buy a HDD recorder. Can I invent a practical device to record? Of course, the device would have to press the record button at a fixed time each day, and move over and press the stop button half and hour later (each day). This would have to go on everyday. I know that you can buy timers which can automatically start a device and stop it (start and stop at whatever time you please). For instance, you can automatically keep a lamp lit from a fixed time til later by setting the times on your timer. I have two timers. Can I achieve this recording? (My idea was to exploit Faraday's law: use the timer to start a current going through a coil (hovering above the record button). If a magnet was hung above the coil (a very thin magnet; similar in shape to a pencil), this would push the record button. Have a similar system set up above the stop button. The only problem is that you would have to re-set the system at the end of each day in-order-to prepare it for the next day but unfortunately no one is at home. Thankyou for any practical suggestions and note that the space on the video is not a problem. —Preceding unsigned comment added by 58.161.138.117 (talk) 01:16, 2 April 2009 (UTC)

A spare computer and a couple old CDROM drives, perhaps? (My preferred solution would be to purchase a DVR or download the shows from The Pirate Bay.) – 74  06:24, 2 April 2009 (UTC)

Thanks but would my device work? Or can a similar device be used and would it be practical? I appreciate your help but unfortunately I don't have access to these things. I'd like to know how one could use a device similar to mine to achieve this. —Preceding unsigned comment added by 58.161.138.117 (talk) 09:53, 2 April 2009 (UTC)

Great story about the CDROM drive! Jørgen (talk) 13:22, 2 April 2009 (UTC)

You said that the space on the video tape is not a problem, but I don't understand how that can be. If this is an hour length show, that's 40 hours. For a half-hour length show, that's 20 hours. I don't think even the longest tape on the lowest speed/quality setting will last that long. So, in your case, perhaps a low tech approach is in order: have a friend or relative record it for you. Also, the show may very well be available where you're going for vacation, so you could possibly watch it there. What's the show and where are you going ? StuRat (talk) 10:49, 2 April 2009 (UTC)

Thanks but as I said, none of your options are feasible in my case (don't ask me why but if I were to tell you, you would know what I mean). I know for sure that the show will be on in the place I am going to - the only problem is that this place is a non-English speaking country. I can speak the language but not very well and certainly not well enough to understand the show. I mean, it is not going to be the end of the world if I miss it - I initially thought so and thought long and hard as to how I could record it. I worked out the method using Faraday's law but then I realized the space problem. So really, my main interest now is how I can alter my method to record my show (practical alterations allowed only). If I can find an alteration, it'll be a bonus as I can record my show. Otherwise, it's not a big deal because I can live without watching the show anyway. Thanks for any help and the suggestions I have got so far. —Preceding unsigned comment added by 58.161.138.117 (talk) 11:58, 2 April 2009 (UTC)

Well, because of the length of the recording, you can't use unattended VHS anyway, right? If your computer has a "video in" port (uncommon on PCs unless you knowingly bought a TV card (in which case you wouldn't be asking this question) but I know these used to be around on old Macs, for example) you can use the TV tuner on the VCR to feed the correct channel into the PC (via "video out" on the VCR), and then set up a timed recording-to-harddisk on the computer. Other than that your best solution would be to leave the VCR with a friend and kindly ask them to turn on the timed recording each day and switch the tapes now and then. Jørgen (talk) 13:27, 2 April 2009 (UTC)

I'm not sure that those timers you get for turning lights on and off have enough accuracy to tape a TV show. The ones I've used are pretty crumby. APL (talk) 14:18, 2 April 2009 (UTC)

I feel compelled to caution against creating an automated device using mains power without sufficient understanding of the principles of electricity. At best your design would work chaotically; at worst you could do serious damage to yourself, your house, etc. So far, the best advice presented is to have a friend/neighbor record the shows for you. If none of the advice above is acceptable then I suggest you resign yourself to missing your show while you're gone and have a good vacation. – 74  01:55, 3 April 2009 (UTC)

• Buying a HDD recorder may not be feasible, but you could certainly upgrade your VHS recorder. Before I recorded DVDs I set my VHS recorder while on holiday. It could set up recordings a month in advance and had slots at least 8 recordings. The limiting step in my case was the size of the tape. - Mgm|^(talk) 12:02, 3 April 2009 (UTC)

• That's a good point. You could probably buy a much better VCR cheaply, say over eBay, as that technology is rapidly becoming obsolete. You might either find used VCRs or new ones that just don't sell any more. If they are really cheap, you could pick up several, so you can set each to record maybe 10 hour-long (or 20 half-hour) episodes on a tape. Also be sure you have enough antennas for all those VCRs. StuRat (talk) 18:29, 3 April 2009 (UTC)

My Firefox is lazy...

Using Firefox 2.0.0.20 in Windows 98, it complains of an "Unresponsive Script" and asks if I want to continue whenever I refresh my Netflix queue. It wouldn't be so bad if it asked just once, but it asks several times when I try to refresh the page. Is there a way to get it to only ask once per action on my part ? Or, perhaps to give it a bit more patience and try a bit more before it bothers me with it's complaints ? What I'd like it to do is go on working to load the page, without bothering me, while I go to another tab, like here, and work on something else. Note that I don't necessarily want to disable this warning entirely, since I do sometimes cancel an unimportant action (like loading some page that's full of annoying ads), when I get this warning. StuRat (talk) 10:32, 2 April 2009 (UTC)

To increase the javascript timeout, go to about:config. In the search, type dom.max_chrome_script_run_time. I believe it starts out at 20 (seconds). You can increase it to any value you like. I looked, but didn't see any setting that looks anything like "show script error". -- kainaw™ 16:06, 2 April 2009 (UTC)

That sounds like what I want, what directory should I search to find it ? StuRat (talk) 18:19, 3 April 2009 (UTC)

I noticed a suggestion for Netflix... Make multiple profiles and break up your queue among the profiles. This will decrease the script time. -- kainaw™ 16:37, 2 April 2009 (UTC)

That's an interesting idea. I get 3 Netflix mailed to me at a time and also use the streaming video service through an LG set-top box. Is there some way to have 3 different profiles, each of which will have one mailed to me at a time ? In that case, how would I select which of the 3 streaming queues shows up on the set-top box ? StuRat (talk) 18:17, 3 April 2009 (UTC)

Have you considered updating to the latest version of Firefox? Tempshill (talk) 20:33, 2 April 2009 (UTC)

Or even Windows? 98 is not designed to work with anything new, like Firefox (ok, Firefox is not new, but Microsoft have only just acknowledged its existence).--KageTora (talk) 21:07, 2 April 2009 (UTC)

Just steal Windows Xp, holy moly. --70.54.192.144 (talk) 08:59, 3 April 2009 (UTC)

If your computer is running too slowly, upgrading to a newer version of Windows is pretty much guaranteed *not* to be the answer (because of the significant increase in required resources in each new version). – 74  12:15, 3 April 2009 (UTC)

Yes, upgrading FireFox further likely requires upgraded the O/S, which requires tossing my ten year old computer out and getting a new one. Not an option. StuRat (talk) 18:13, 3 April 2009 (UTC)

Atomic operations on NFS

Certain POSIX file system operations are normally guaranteed to be atomic — if several processes attempt conflicting operations, exactly one of them will succeed, and all will be properly notified of their success or failure. Examples include appending to (with O_APPEND) or creating (with O_EXCL) a file, making directories and links (of both types), and renaming files (except when it isn't). These become dicey over NFS: append doesn't work, and O_EXCL depends on the NFS version and kernel. mkdir() is apparently how CVS deals with the issue, and link() is the documented way of dealing with it. The system calls link() and rename() may have false failures on NFS; symlink() makes no such disclaimer, but I can't imagine how it would defend against the same problem. mkdir() just says "here there be dragons", and unlink() is even scarier (I suppose it is referring to the potential for a local unlink() on an NFS server to break processes on clients that (think they) have the file open — is this better understood by anyone?).

Meanwhile, a narrative of solving this problem suggests that the documented approach using link() doesn't work, and that there can be false successes from these functions (see the discussion of move())!

So: what's the whole story? Are NFS versions that don't support O_EXCL vanishingly rare yet? (I doubt it.) Are mkdir(), link(), symlink(), and rename() really guaranteed to be atomic even over NFS (bugs notwithstanding)? When these functions falsely report failure, with what errno do they do so? (I should think EEXIST except for rename(), which should be ENOENT.) Of course, the retransmitted RPC (are there other false-failure mechanisms?) could also fail in any of a variety of ways if other processes have manipulated the file system (a new symbolic link could make the destination path for link() now be on a different filesystem for EXDEV, for instance). So perhaps the point is that any errno could mask a success (almost: EFAULT, ENAMETOOLONG, etc.)?

Finally, is there a standard for how the corresponding user commands (mkdir, ln, link, and mv) behave in the presence of NFS and false errors? Looking at, for instance, mkdir.c and its subroutine in mkdir-p.c, I see only a single call to mkdir(), so it would seem that a shell script cannot assume that a failure of GNU mkdir is real. --Tardis (talk) 18:02, 2 April 2009 (UTC)

This answer over at Stackoverflow quotes the Callaghan book as saying that in NFSv2 only symlink and rename are atomic (and not mkdir or unlink), and that v3 adds creat as atomic too. Yes, the rename(2) manual does look scary (although it says if the server "crashes"; an ordinary network error shouldn't cause this, so you could be horribly unprofessional and pretend it'll never happen). I agree that if rename has that problem then surely symlink must too. The "solution" is (tee hee) to use NFS locking instead, but of course that has its own "what happens if the server crashes" problems. What's needed is real transactions, but I see no guaranteed safe way of implementing them on such a shaky foundation. 87.115.166.150 (talk) 19:19, 2 April 2009 (UTC)

Ah ha! This must be why some locking schemes (including the one emacs appears to use) use symlink(2) and use a fake string for the link target which uniquely identifies the requesting process across the network. Emacs appears to use the format $username@$hostname.$pid:$somenumber (where that $somenumber might be unix time or maybe just a random number). With that you can:

   while True:
     symlink (lockfile, <lengthy name using above recipe>)
     if success:
       return SUCCESS
     else: // fail
       if the lock exists:
         if it's our PID etc:
           return SUCCESS
         else
           return FAIL
       // if the lock doesn't exist, loop around and retry

This probably still has failure modes, particularly when locks are held briefly, but it's progress. 87.115.166.150 (talk) 19:35, 2 April 2009 (UTC)

There are of course multiple levels of "usable" here. Anything must be atomic and controlled in its effects in order to be safe for locking. Anything that additionally has a visible, client-specific side effect can be used for locking: the stat() check works without needing link()'s return value at all, and similar tricks work for symlink() and rename() [EC: like Emacs']. Of course, anything that always reports its success properly is also sufficient.

I find it hard to believe that link() is out, since open(2) specifically recommends it. I also have trouble comprehending how mkdir() could be damagingly non-atomic: its problem is that it (I suppose) provides neither a reliable status nor a client-specific side effect. I guess unlink() is bad because it could, after a lost response, hit again and break someone else's lock — but that applies to rename() just as badly(!), and doesn't seem to fit with the use of NFS handles. (Conveniently, for my present purpose I only need lock, not unlock.) The proposed solutions you linked assume that their primitives are reliable — if ln falsely declares failure, no one will ever get the lock. (These issues can clearly be worked around by using a unique target for the symlinks and checking for it.) If my fears about rename() are justified, clearly the proposed unlock routines are broken regardless of any added checking; we could make (with no error checking) a directory and (with checking) a lock file in that directory and then rename() the directory away to unlock: any replay would fail because the destination would be a non-empty directory.

I think, with all that, it's possible to make everything bullet-proof (so long as the clients don't crash and keep their locks!) — but my inability to see Callaghan's reasoning gives me pause in claiming victory. (I'll try to look at the book sometime.) Moreover, I'd like to know what the full set of working solutions is for the sake of having choices and understanding. --Tardis (talk) 20:17, 2 April 2009 (UTC)

I think the long and short of it is "NFS is broken", that its semantics fail to properly mirror POSIX, and that without server-site state these can't be fixed in NFS. Even if later NFS versions are better, you probably wouldn't be well advised to code just to them (unless you control the deployment environment). That leaves you having to code as if all the "guarantees" of atomicity are false (which it seems they kinda are), so you're left with the above scheme or something like it. That should work (as long as the underlying filesystem is sane), on the assumption that NFS' caching isn't so broken that it would allow two processes to "successfully" create the same symlink. It's no shock that there are so many rude alternate-expansions for the acronym NFS. 87.115.166.150 (talk) 20:34, 2 April 2009 (UTC)

Ah, more issues. To verify that symlink you'll need to call readlink(2), but it seems that readlink itself has issues, not least an inconsistent size when compared with the value reported by lstat(2)on NTFS (not NFS). GNU solves this with areadlink - that's the function that ls(1) uses. 87.115.166.150 (talk) 12:15, 3 April 2009 (UTC)

Playing Japanese games on a British PC

I lived in Japan for a while, and came back with a few games (among other things) and have recently bought a new PC here in the UK, because my two Japanese laptops have packed in now. So, my question is, how can I get these games to work on a British PC, but actually showing the Japanese Kanji, etc., as all it has done for me in the past is show a bunch of '?' and I have to guess what the option was. I read Japanese, and this would be a massive help. Just games like 'Cossacks' or 'Age Of Empires'. Cheers.--KageTora (talk) 20:06, 2 April 2009 (UTC)

For starters (assuming you're running XP): control_panel -> regional_and_language_options, and in the languages tab check the two suplemental features boxes. I don't know if that actually installs Japanese fonts (or if it's just the necessary infrastructure for Japanese rendering and input-method); the East Asian sans-serif typeface has some links to fonts, if that doesn't do it. 87.115.166.150 (talk) 20:17, 2 April 2009 (UTC)

Thanks, I have actually done all that because I work as a translator from Japanese and I use this computer for that. It doesn't seem to work for games.--KageTora (talk) 20:35, 2 April 2009 (UTC)

They may be relying on a specific font that Japanese XP installs; if you can get the disk from one of those Japanese laptops to work in your UK PC, copy over the .TTF files by hand. 87.115.166.150 (talk) 20:44, 2 April 2009 (UTC)

It sounds like the games are non-Unicode. If that's the case, changing your setting for non-Unicode programs should fix the problem. As long as you have all of the Japanese fonts installed, that should take care of the question marks. Microsoft has instructions here. Daram.G (talk) 15:03, 3 April 2009 (UTC)

Yes, this is the most likely thing to fix the problem, but it does have potentially undesirable side effects like making \ sometimes display as ¥ and causing display problems with the many programs that assume a Latin-1 locale (though if you've previously used Japanese Windows I'm sure you're used to this). You could try AppLocale instead, though it doesn't always work (and I suppose games are especially unlikely to work). -- BenRG (talk) 21:27, 3 April 2009 (UTC)

Google calendar search results

When search for a specific word in my Google calendar, I says it finds 41 results, but only displays the 6 recent instances of that word. I suspect that results that are displayed are the ones that were created since I started using Google calendar, and the ones that aren't displayed are events that I imported from my old Yahoo calendar. Any hints on how I can get this to work correctly? I'd really like to be able to search my old events, which is of course the whole point of saving old calendars. ike9898 (talk) 20:48, 2 April 2009 (UTC)

• Why aren't you asking Google? - Mgm|^(talk) 11:58, 3 April 2009 (UTC)

Because places like the Ref Desk are usually quicker to response than massive companies providing services for free? --98.217.14.211 (talk) 13:10, 3 April 2009 (UTC)

I searched for the answer fairly thoroughly before I came here. And yes, I find the Ref Desk is full of very smart people that can often to you to the right answer very quickly. BTW, since I posted my question, I've discovered that if I go to one of the events that doesn't appear in the search results, open it (ie click 'edit details') and then re-save it, it will then appear in future searches. That doesn't solve my problem, but maybe it is a clue. ike9898 (talk) 13:49, 3 April 2009 (UTC)

Will I Need a new Power Supply?

My current PC has an Intel Pentium D processor. I have a (pipe) dream of upgrading to a Core i7 (for which I assume I'll also need a new motherboard, new ram, and a whole bunch of other wallet busting stuff.) As you can imagine, I'm trying to fit this whole thing into an almost non-existent budget, so I'm wondering, would I also need a new power supply? A few months ago I upgraded my graphics card to a GeForce 8800GTS, and I bought a new 550W power supply for that. And with it my computer would spontaneously shut off. I had to return it and get a 650W power supply. Now I'm thinking that buying "just enough" power wasn't the best idea. Is 650W going to be enough for an 8800GTS, a Core i7, and 3 hard drives? (Oh yeah and by the way, I have 3 hard drives.) Or am I going to have to scrape together enough money for yet another power supply?

Digger3000 (talk) 21:06, 2 April 2009 (UTC)

My experience, which may be completely different from yours, is that hard drives are not the power problem (until they are breaking and suck tons of juice just to get the disk up to speed). The biggest problem is the video card. I had a computer with a 300W power supply and no problems. I got a fancy dual-head video card and suddenly my computer rebooted all the time. I read the manual for the card and it said it required 100W all its own. So, since 300W was OK before, I assumed a 500W power supply would be fine (200W more). Well, I continually got random rebooting issues. I returned the 500W power supply and got a 650W. To this day, I still get random rebooting problems - but they are very rare and always happen when there is a lot of video conflict going on, such as dragging a window with some video playing in it from one monitor to the other. So, I have no reason to assume the new CPU will require a better power supply. -- kainaw™ 22:00, 2 April 2009 (UTC)

I use an 850W PSU and I have an 8800GT, an FX-60, and 3 hard drives. I believe the required specs for the 8800 series cards are around 500W, but I could be mistaken. Useight (talk) 02:36, 3 April 2009 (UTC)

Follow-up question: I bought the 8800GTS because I thought it would give me fantastically better performance than my old 7600GS. Well...it didn't. And right now my theory is that CPU, rather than the graphics card, was holding me back. In other words, the 8800GTS is "too much card" for the Pentium D processor. (Hope that makes sense.) I'm worried, though, that if I did get a Core i7, I would run into the reverse problem. The Core i7 would be "too much processor" for the 8800GTS card. Is it likely that the 8800GTS won't allow the i7 to be used to its full potential? (I don't know if I phrased that in the best way, I hope you understand what I'm asking.) Digger3000 (talk) 02:58, 3 April 2009 (UTC)

What you're asking is what part of your computer is the bottleneck. The good news is that you can't have "too much processor for your graphics card" -- having one component that's more powerful than the others won't make your computer run more slowly. Chances are that the 8800 GTS would indeed be holding your computer back from its full potential, but you should still see a marked increase in performance compared to your old setup -- combining the Core i7 with the 8800 GTS won't slow anything down, but you'll certainly get even better performance with a newer graphics card -- something from the GeForce 200 Series, perhaps. The thing is, though, you're never going to be right on the cutting edge unless you spend pretty massive amounts of money and upgrade on a very frequent basis. -- Captain Disdain (talk) 06:17, 3 April 2009 (UTC)

Vairy Touch Mobile Phone

On a Vairy Touch, how do you keep the screen from going off every 10 seconds? It's a PITA when you are trying to look through your music playlists.--KageTora (talk) 22:03, 2 April 2009 (UTC)

Don't worry. Done it. It's apparently on 'LCD Backlight', which would be irrelevant on a mobile made by an English speaking company. Anyway, thanks for your time. --KageTora (talk) 22:13, 2 April 2009 (UTC)

Be careful, I have a "Back-light time-out" setting on my cell phone, which I set to "Off", meaning I don't want there to be a time-out at all, and it should only turn off the back-light when I close the phone. However, what it actually did was turn the back-light off entirely, which was a royal pain as I couldn't see to turn it back on. It was like some practical joke the manufacturers liked to play on unwary customers. StuRat (talk) 18:05, 3 April 2009 (UTC)

Resolved

StuRat (talk) 18:05, 3 April 2009 (UTC)

April 3

Please correct the sentence

This line in Lightweight Directory Access Protocol is quite perplexing to me: "The client may also send a certificate to prove its identity. After doing so, the client may then use SASL/EXTERNAL to have this identity used in determining the identity used in making LDAP authorization decisions."

Is this a grammatical error or a technical one? Probably not the latter as the line was added by Kurt Zeilenga who is the author of several RFCs and was the founder and architect of the OpenLDAP project. Jay (talk) 08:42, 3 April 2009 (UTC)

I think it's probably a grammatical error -- or perhaps more to the point, a brain fart. I think the sentence is supposed to read, "After doing so, the client may then use this identity with SASL/EXTERNAL when making LDAP authorization decisions." (Or, perhaps, "...may then have SASL/EXTERNAL use this identity when making...") I'm guessing, of course -- that is the kind of a sentence that tends to send the reader's parser into conniptions. -- Captain Disdain (talk) 09:08, 3 April 2009 (UTC)

Are you sure the sentence is incorrect (though it is most certainly unclear)? To me, it parses as:

The client supplies a certificate to prove its identity. The client then uses SASL/EXTERNAL (a mechanism for sharing authentication information) to request its (newly established) TLS identity be added to the list of identities used in making authorization decisions for this connection on the server.

Presumably, the client has no control over *how* the server decides which identity to use for further authorization decisions (and may supply credentials that the server disregards entirely). For example, all LDAP users are given access to "common" directories, but access to private directories is limited to admin users with a certificate. A client who has already authenticated as a user (identity:user) enters a TLS and supplies the admin certificate, which is validated. The client then requests that the new TLS authentication (identity:admin) be added to its session for further authorization requests. The client then requests a private directory; the server verifies the client has an authorized identity (admin) and processes the request. The client then requests a common directory; the server verifies the client has an authorized identity (user) and processes the request. (The preceding is speculation; I have no experience with LDAP.) – 74  12:07, 3 April 2009 (UTC)

Can this be done?

How could i connect an IDE hard drive to my computer that has only one IDE header on the motherboard? currently the only thing connected to it is my DVD Burner. I am not planning to dual boot or anything but just use it as a storage thing. How do i connect it so that it doesn't screw up my configuration?  Buffered Input Output 13:04, 3 April 2009 (UTC)

A given IDE host interface can support two IDE devices (known as the master and the slave). Either your current IDE cable has three connectors (one for the host, and one each for those two devices) or you should get one that does. It's generally necessary to configure both devices with jumpers, to make one master and one slave. If you need to support yet more IDE devices, and your motherboard doesn't have another IDE host connector, you can find inexpensive IDE RAID cards (which, contrary to what the name RAID suggest, will also allow each disk to work normally - standalone, not part of a RAID volume). Dog Day Today (talk) 13:33, 3 April 2009 (UTC)

Increase the passage of time

Is it posable to increase the passage of time within a virtual environment like Virtual Box Or VMware – This is me 22:11, 3 April 2009 (UTC)