Wikipedia:Reference desk/Archives/Computing/2007 April 16

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

April 16

Macs vs PC's

I frequently hear computer users boasting the superiority of Macs over PC's. Specifically, what traits of Macs makes it better than it's PC counterpart? If Windows and Mac OS X are run on two seperate similarly spec'd computers, which of the operating systems would be generally faster? Thanks. Jamesino 00:53, 16 April 2007 (UTC)

Neither Windows nor OS X is probably appreciably faster at anything (ignoring Vista). OS X is naturally somewhat more secure than Windows because it's UNIX-based, and doesn't have Internet Exploder, as well as simply being less popular and therefore less targeted (when was the last time a major email worm made it around with a Mac executable as an attachment?). It largely comes down to user interface differences and software compatibility. Windows supports a lot more software overall and OS X supports a few more graphics/video/animation/audio/art tools. Parallels Desktop and WINE/Darwine help close the compatibility gap, though they're not as successful with games. Now that they're using x86 processors, you can run Windows on the new Macs as well, which is nice if you want to play the occasional game, but dual-booting is a bit of a pain. -- Consumed Crustacean (talk) 01:07, 16 April 2007 (UTC)

The problem is a Mac is a PC, architecturally, and at least nowadays. What the real comparison is about is between Windows and Mac OS X. That is what is disgustingly misleading about those Mac vs. PC ads -- they are essentially Windows and Mac OS comparisons. There are a lot of advantages of Mac OS over Windows (Unix-based, hence more secure, better programming environment, etc), but not really a lot of advantages of Mac hardware over PC hardware. —The preceding unsigned comment was added by 129.78.208.4 (talk) 02:09, 16 April 2007 (UTC).

As a Mac fan and (rabid) Microsoft hater, I could go on and on about the virtues of the Mac and the bletcherosities of Windows PC's, but really, this is a flamebait question, guaranteed to scare up all sorts of hoary old claims and counterarguments, so perhaps I'll just nip that regression in the bud by not saying anything. (Whoops! Too late... :-) ) —Steve Summit (talk) 02:47, 16 April 2007 (UTC)

Why are all the comparisons between OS X and Windows? Why doesn't anyone consider Unix/Linux systems as PC and compare them against OS X? Unix/Linux are very much x86 compatible and thus as much a PC as any Windows system. --Wirbelwindヴィルヴェルヴィント (talk) 07:31, 16 April 2007 (UTC)

Yes, have the best of both worlds. Install Linux on an ordinary PC. That way it will be cheaper than either of them, you get all the security and virus protection like Mac OS X, but you don't need to pay 50% extra to get a Mac. --antilived^{T | C | G} 08:03, 16 April 2007 (UTC)

Then again Macs cost more because you get all the software a normal person really ever needs, with the extra purchase of Microsoft Office for Mac if you need it instead of iWork. [Mαc Δαvιs] ❖ 17:00, 16 April 2007 (UTC)

It's a shame that Parallels Desktop isn't around for any OS besides OS X. It's a pain to have to dual boot or use plain virtualization when an app doesn't run well in WINE. That said, WINE has made some tremendous improvements recently. And Cedega finally added proper SM2.0 support, though DX10 / SM 3.0 games are right around the corner. That is the largest thing holding me to Windows right now: games. Most everything else that I personally use runs fine in WINE or has an equivalent. Please excuse the nonsensical structure of this post. -- Consumed Crustacean (talk) 08:29, 16 April 2007 (UTC)

There are VMWare or Xen around for Linux though... --antilived^{T | C | G} 09:08, 16 April 2007 (UTC)

Bah.. "Mac OS X based on Unix" is marketing-speak. OS X isn't any more based on Unix than Linux is based on Unix (can 'o worms... how do you define "Unix"?). OS X has its roots in NEXTSTEP/BSD (the latter arguably is more Unix than UNIX, but who is counting?). Just being a Unix-like OS doesn't offer any assurance of security, though I think the "lacks Internet Explorer" point is a strong argument. Granted, nobody worth talking to really cares about "Unix name purity", but it's unwise to wake the neckbeards from their slumber.

On topic, nothing makes an OS better or worse than another except for the needs of the user. Try both OSes, use the one that you feel most comfortable with. For myself, I'd slightly rather use OS X than Windows, though I don't particularly like the WM of either (exposé is cute and all, but a decade of using Xwindows has indoctrinated me to the glory of virtual desktops). In practice, I use FreeBSD and sometimes Linux for day-to-day usage and Windows XP for work since MS Office is the de facto standard, most of the EDA software we use runs on Windows, and I'm not going to waste weeks fiddling with Wine... -- mattb @ 2007-04-16T16:27Z

Virtual desktops are coming in MacOS 10.5 (Leopard). For me, since I added a second screen to my Solaris system, I've felt a lot less need for virtual desktops, but I'll probably appreciate them again on my Mac laptop.

Atlant 12:40, 17 April 2007 (UTC)

I hadn't heard that. I'll be interested to see if the implementation is to my liking (I use edge flipping even though it seems to infuriate most people). My primary computer is a laptop, so a second screen isn't much an option. -- mattb 15:16, 17 April 2007 (UTC)

See here.

Atlant 17:09, 17 April 2007 (UTC)

Still can't tell whether it will support edge-flipping like functionality. Still, it at least looks well thought out (as one would expect). -- mattb 21:14, 18 April 2007 (UTC)

Yeah, I think you should just try both of them like mattb said. It really depends on who you are. Macs are generally better if you have trouble using the computer. If Linux is the choice for you, then you already know it and don't need our help. If it's neither of those then go for Windows. [Mαc Δαvιs] ❖ 17:00, 16 April 2007 (UTC)

Which of the two, Windows or Mac OS (excluding Linux), is generally geared more towards "computer geeks" and users with greater aptitude for computing?Jamesino 22:41, 16 April 2007 (UTC)

There was a time when most people's answer would have been windows, hands down. But, since OS X, this is a toss-up. Many people choose Mac because they think it's "simpler" - they've intentionally marketed themselves to this audience. But, plenty of propeller-heads like Macs these days too. Depends on the geek. You'll find that compiling unix source code (for example) is a piece of cake on an out-of-the-box Mac, versus not easy at all on Windows. Friday (talk) 22:50, 16 April 2007 (UTC)

Yup, I'm a poster-child for that observation. Love Unix like nothing else, the propeller spins so fast you can rarely see it, and yet I'm a Mac zealot, too. --Steve Summit (talk) 01:30, 17 April 2007 (UTC)

I always break it down to the question of whether you like to use your computer or dink around with your computer. If you like to dink, get a PC; it will prvide you with years of endless dinking opportunities. If you actually want to use your computer to accomplish work (with the computer minimising how much it gets in the way), get a Mac (or maybe Unix or Linux). And I say this as a person who, every day, uses PCs, Macs, and Sun Solaris (Unix).

Atlant 12:44, 17 April 2007 (UTC)

and then there's the security issue

Okay, I said I wasn't gonna do this, but I gotta dive in anyway. (The only thing more insatiably tempting than an OS flamewar is an OS security flamewar...)

Several people have said, "use whichever you like, use whichever works for you". And that's fine, up to a point. But the problem is that security is a sort of a "sleeper" feature. When it comes right down do it, most users -- and most vendors -- don't really care about security. You may say you want a secure system, and a vendor may claim to provide one, but how do you know? Security bugs don't (necessarily) mean that your system runs too slow, or is missing features, or gets blue screens of death all the time. That is, security problems don't manifest themselves in ways that are immediately obvious to assess.

So if you don't care about security, go ahead, choose the computer that works best for you. But if you do care about security, your definition of "works best" has to include being adequately secure.

Some people will disagree vehemently with me here, but Microsoft Windows is fundamentally less secure than Unix, Linux, and Mac OS X. Vista might be adequately secure; it's still too early to tell. (Initial reports don't look too good, however.)

Some people (the people who disagreed vehemently with me above) will tell you that there's nothing inherently wrong with Windows's security, it's just that the nasty evil hackers pick on it more and write more viruses for it. There are two flaws with this argument:

1. It's wrong. As I said, Windows is (or at least has been) fundamentally less secure. (I'll mention one reason below.)
2. Even if the argument is accurate -- even if Windows is, on balance, just as secure as its competitors -- the fact remains that there are thousands of times more extant viruses and other forms of malware for it than for all its competitors combined. That may not be fair, and it may not be Microsoft's fault at all, but it's a fact, and if you care -- really care -- about security, you'd do well to shy away from Windows for this reason alone.

Why is Windows so insecure? One fundamental reason is that Microsoft has always been interested in making it maximally easy for third parties to sell you new software which you can easily -- really easily -- install on your computer. Similarly, it wants to make it really, really easy for web pages that you visit to install keen new extensions to your browser, to improve your browsing experience. These installations -- of new programs, new web browser extensions, new device drivers for some new peripheral you just added, whatever -- require at most one click to install, and quite often zero clicks. Nothing could be more convenient.

The problem, of course, is that when you've made it this easy, there's no way to distinguish between the good and the bad. It's easy -- with little or no intervention from you -- for those programs you bought, and those extensions you want, to be installed. But Microsoft managed to make it precisely as easy for the email viruses and the adware and all the stuff you don't want to also be installed, maximally easily, with zero clicks, whether you want it or not.

Moreover, at this point and to some extent, Microsoft can't fix this problem. So many app vendors have gotten so utterly dependent on all the dangerous things that Microsoft used to let them do that Microsoft can't just turn them all off in the name of security, without risking not only a revolt from those vendors, but renewed antitrust accusations. Microsoft's trying to close some of the loopholes in Vista, but there's a new can of worms brewing with UAC...

--Steve Summit (talk) 01:27, 17 April 2007 (UTC)

And some of that is just rubbish. There are plenty of exploits, technological and social, to root or otherwise compromise the security of a multitude of OSs if you have local access. Many viruses and worms spread through social engineering, convincing people to open email extensions or files in instant messengers, which obviously grants that access. And if a program in Linux opens up a sudo window, and the user has no idea if it's malevolent, it's very possible that they'll enter in their password without thought. Of course it is less secure, but a technologically incompetent user will be as insecure with any OS which provides any measure of freedom. -- Consumed Crustacean (talk) 01:39, 17 April 2007 (UTC)

Don't say, "some of that is just rubbish". Please be specific.

• "There are plenty of exploits... to root or otherwise compromise the security of a multitude of OSs"

There are a very few. By comparison to the number of exploits for Windows systems, there are practically zero.

• "...if you have local access"

If you grant an attacker local access to your system, you are very stupid, and the problem of protecting your machine becomes exponentially harder. Yet well-designed systems give you some protection even here, by ensuring that a user-mode compromise can't escalate privileges to root.

• "Many viruses and worms spread through social engineering... which obviously grants that access."

This is not only not obvious, it is quite false, and it is at the heart of the PC virus epidemic.

Even if you convince a user to open an attachment, where is it written that the attachment, if executable, must be run? If there simply wasn't a way to easily run an executable from an untrustworthy source (from an attachment in an email, or from a random web page) the PC security landscape would be vastly, vastly different.

• "...if a program in Linux opens up a sudo window... it's very possible that..."

But this never happens.

It's all well and good to claim that these or similar exploits are "possible" on other systems. So why don't they ever happen? Because the evil nasty hackers all love the Unix and Mac and Linux and Solaris users so much that they decline to even try? Or because the task isn't actually so possible after all?

• "a technologically incompetent user will be as insecure with any OS which provides any measure of freedom"

Simply false. (Fortunately.)

Most users are technologically incompetent, so if we had to rely on them to keep our computers and networks secure, we'd have a huge problem. Fortunately, it's possible to engineer in lots of protections. (Unfortunately, in the Windows world, these protections haven't been engineered in, and most of the responsibility is left to the users, which is why we do have a huge problem.)

—Steve Summit (talk) 03:16, 17 April 2007 (UTC)

I was specific. You're the one stating uniquivocably that Windows is far less secure without actually giving any reasoning besides "OS X and Unix are more secure". On your points:

• Bullshit. Have you read any security websites? And even without root access a program can wreak havoc.
• How do you think email attachment viruses spread? There's one right now that's going around by claiming to be an antivirus program. User runs it, it has local access.
• So, umm, you want to make it hard to run executables? Just removing the executable attribute won't work; if you can convince users to run an executable because their computer is at risk if they don't, you can convince them to flag it and then run it
• It's all well and good to claim that something never happens based on a lack of evidence, but to simply disregard it as ever possibly happening because it disturbs your personal view is foolish. And then, to claim UAC has problems when they basically consist of convincing the user to do the same thing [1] (and yes, I realize the point that article is making)...
• Again, bullshit. I use Windows without an issue. I know some users who have actually intentionally installed Gator, though. The only reason they're saved right now on "alternate OSs" is because the number of people targeting those platforms is minimal.

I'm not saying that Windows is secure. It's not, and there are a plethora of issues. I'm just saying that most of the insecurity will always come from the user, and such things would affect most any OS. -- Consumed Crustacean (talk) 13:08, 17 April 2007 (UTC)

Tempting as it is to continue the point-by-point thing, it'll drift well past the scope of this RD if it hasn't already. (My fault for starting it, I know.) So I'll confine myself to one point here, and defer the rest to my talk page.

When you say "most of the insecurity will always come from the user", what do you mean by "most"? 51%? 75%? 95%?

The reason I ask is, do you imagine that the bulk of the solution to the current problem will come through user education? Do you believe there's any point in pursuing technical fixes, in making Windows more secure, or (if that's for whatever not an option) pursuing reactive band-aid solutions such as antivirus software?

If the only solution is user education, I despair. It's obviously not going to work if it hasn't by now. I despair particularly because even though I don't use the targeted OS, and even though I'm smart enough not to click on unsafe links, my inbox is deluged all the same with hundreds or thousands of spam emails per day, sent by the botnets made up of all the computers running the targeted OS and owned by users that weren't smart enough. So it's very much my problem, even though it shouldn't be my problem.

Second, if you believe (I'm not saying you do, but plenty of people seem to) that there's no point in pursuing technical solutions, because none of them will ever be 100% effective, or because stupid users will always figure out ways around them, this strikes me as being exactly equivalent to deciding not to put seat belts in cars because they can't protect against all accidents, and instead trying to educate users (drivers) into not crashing their cars so much. Or declining to put railings around balconies or other high places, because anyone with half a brain knows better than to jump off, and even if you do put up a railing, a suitably motivated/stupid/deranged person can always climb over it. —Steve Summit (talk) 03:08, 18 April 2007 (UTC)

I don't believe that. I believe that you're blowing the technical issues up to exceedingly large proportions while ignoring the contribution of the user. Yes, Windows needs to be fixed. Vista helps, but it probably needs more. However, even that will not take out the fool component. You can't fool-proof any OS which offers anything in the way of freedom.

Again, my example with the 'virus scanner' worm stands and has not been responded to. That's pure social engineering in action. And mail botnets, those which bother you so much, don't even require root access in many OSs. All they need is to be able to send mail. -- Consumed Crustacean (talk) 03:01, 19 April 2007 (UTC)

I'm not "ignoring the fool component". There's no way to absolutely fool-proof any system -- but that doesn't mean we shouldn't try to make systems mostly safe in the face of garden-variety fools. For example, most automobile accidents are caused by theoretically-avoidable human error. But we spend huge amounts of money on technical solutions to try to mitigate the consequences of that human error, because in practice and on a large scale, it's unavoidable.

I'm not sure what point you're still trying to make with the virus scanner worm. I thought I'd addressed it; I'll try again:

If your point is, "Some people can always be tricked into installing and running untrustworthy executables, so therefore there's no point removing the easy one-click executability features", I vehemently disagree. (And you said you didn't believe that, so I guess that isn't your point.) If your point is "removing easy one-click executability isn't a 100% perfect solution, since some people will still be tricked", then I agree. There are no perfect solutions -- but that doesn't mean we should ignore the 90% solutions. After all, three independent 90% solutions can give you 99.9% reliability.

If easy one-click executability were removed, I believe the virus problem would be reduced by at least an order of magnitude or two. Furthermore, it's much, much easier to educate people by saying "don't install untrustworthy programs" than it is by saying "don't click on untrustworthy links". If clicking on links could be made as inherently safe as it ought to be and as most people subconsciously assume it is, the education problem would be much more tractable.

Finally, I want to make sure what we're still arguing about. Earlier, you expressed concern about something "disturbing my personal view". Is the distressingly wrong personal view of mine, that you feel the need to correct, that (a) security problems can be addressed by technical means, or (b) Mac, Unix, and Linux systems are mostly secure, or (c) Microsoft Windows systems are dreadfully insecure? In particular, is part of the argument here that I've disturbed your worldview that Windows is adequately secure, or that it's fundamentally just as secure as the others? —Steve Summit (talk) 01:15, 20 April 2007 (UTC)

Running a basic router firewall has kept my PC free of spyware and adware for at least two years. The security issue is moot if you have half a brain and don't download..hey, what's that BIG PURPLE MONKEY? HE LOOKS SO NICE. -Wooty Woot? contribs 01:51, 17 April 2007 (UTC)

Well no OS can prevent the user from doing stupid things such as sudo rm -rf / or format c:. But then, I agree with Steve Summit, Windows is engineered to easily installation of softwares, including viruses. Just last Friday when my friend brought over a removable harddrive I caught a vbs script virus from it, just by having autorun automatically run the script. My Windows XP is even a customised version with nLite that stops autoruns, but Windows still defaults to autorunning whatever inside when you double click on it. Now I have no idea why you need autorun on a harddrive, but this is seriously bad engineering by Microsoft. --antilived^{T | C | G} 03:24, 17 April 2007 (UTC)

(edit conflict)

But obviously, the security issue is not moot. Obviously (unless we want to live with the current debacle), we have to find some mechanisms that don't place the bulk of the security burden on the user.

Out of respect for the 99% of computer users out there who aren't like you or me, I'm going to edit your statement to "The security issue is moot if you're careful not to click on stuff you shouldn't." It's not clear precisely what it takes to "not click on stuff you shouldn't" -- maybe it's "half a brain", maybe it's a basic mistrust that makes you decline to click on most things, maybe it's an innate knowledge of all the things that are safe to click on, versus all the things that are not. But whatever it is, it's clear that 90% or so of the computer users out there don't have it. So, again, the problem's not moot, and furthermore it's clear -- after years of trying, during which the problem has only gotten worse -- that waiting for those users to acquire their half brains (or whatever it is you think they need) is not going to work.

(Out of respect for the readers here who are despairing of ever seeing their Reference Desk again beneath this torrent of words, I'm not going to address the Bonzi Buddy issue, unless you insist.)

—Steve Summit (talk) 03:31, 17 April 2007 (UTC)

Adding internal speakers

I'm thinking of buying a nice 15 inch laptop, but all the ones I really like are supposed to have pretty crappy speakers. This is a big problem for me, because I listen to a lot of music out LOUD (not a big fan of wearing headphones all the time) and don't want to have to carry around and plug in external speakers. The hope is that I could buy one of these computers, take it over to a computer shop, or have a computer guy come over and replace the computer's internal speakers with kickass ones. Is this a valid hope? Would it work at all? How good could the sound get? I know part of it is the poisitioning of slits for letting the sound out of the laptop and other stuff. Thanks, 70.108.191.59 02:19, 16 April 2007 (UTC)

While it is possible to change out the actual speakers, you cannot expect to get good sound out of a tiny 1-inch or 2-inch speaker glued to the back of a plastic cover. The best sound I've heard in a laptop came out of a Vaio that had 4"x2" speakers down the side of the screen. It was kind of odd as Sony is not well-known for great sounding speakers. --Kainaw ^(talk) 02:23, 16 April 2007 (UTC)

I've heard good sound from a laptop. My mom has a 17" Dell Inspiron with Integrated Sound Blaster® Audigy™HD Software Edition, and it just blasts songs. I can wake people up across the house and downstairs with that thing on full volume. Really impressive. Do you know if I could buy something like that seperately and if it could fit into a 15" computer? 70.108.191.59 03:16, 16 April 2007 (UTC)

I suggest headphones. Then you can crank the volume up as much as you want, at least until the sound is dampened by the blood coming from your eardrums. StuRat 03:28, 16 April 2007 (UTC)

Dell laptops have built in in subwoofers. Though they are small, they make a world of difference. Having a soundblaster sound card helps too. --Russoc4 23:50, 16 April 2007 (UTC)

I agree. I have a Dell Inspiron 6400 with internal speakers, and it's really loud. And the notebook itself is pretty inexpensive too. You can try one of those. Hasanclk 12:09, 17 April 2007 (UTC)

New processor

I just heard that Intel is predicted to release a new type of processor on the 22nd of April. Is this accurate? Is this definitely worth waiting for if I'm thinking about buying a computer soon? How soon will it be in laptops? How soon will other processors' prices go down? If I wait, is it smarter to buy the old (not really) Intel Core Duo computers at a lower price, or is this new one going to blow it out of the water? Thanks, 70.108.191.59 03:19, 16 April 2007 (UTC)

The new processor coming out is the core 2 quad, yes it will be a great processor, but the price will also be very hefty, initially, like always. As to the other questions, it's all subjective and depends on what you do with (and expect from) your computer. For most things, like internet and email, you can get a $500 computer a few generations old that will already do most things just fine. If you play games, or something else 'special' then you need to research more then just asking one question on Wikipedia:) The advice from someone who knows a little, its almost never actually worth waiting for 'the next big thing' because there is a 'next big thing' every couple of weeks in this game. Just get what you can afford and you won't be disappointed. Good luck.Vespine 04:56, 16 April 2007 (UTC)

Core 2 Duos across the board are supposed to be dropping in price on the 22nd as well. -- Consumed Crustacean (talk) 05:19, 16 April 2007 (UTC)

There's rarely a reason to wait to buy a microprocessor for purely technical reasons. Moore's law stops for no man. I'm a big fan of buying a generation behind; you'll get a PC that is totally capable of doing what you want without breaking the bank. -- mattb @ 2007-04-16T16:33Z

What is with people and intel, why wait for an Intel quad core when you can get a much cheaper AMD chip. If your buying a new PC/Laptop look out for an AMD chip and SAVE$$$ SAVE$$ SAVE$$$. --Lwarf 10:40, 17 April 2007 (UTC)

Possibly because Intel has been thoroughly trouncing AMD in the performance/watt metric for some time. -- mattb

Uh...no! jajaja it doesn't matter when you've got arctic ice!!! and water cooling...with fishies in the water...oh wait...intel users are newbs that don't use this...or overclocking...or can't jajajaja...ATI is now with AMD as well..jajaja um...ok...my sugar high wore off...had to rant xD but anyways...this is an Intel question and as a fair contributor i did not get involved in this earlier because i don't believe in bashing...not here anyways...i think it's better just to state facts and not so much opinion...although that would be like an oxymoron of wikipedia being non opinionated when anyone can edit it...eh...ill go and think on that one for a while..but both companies are good...some users just prefer others..as is obvious with my ranting... 200.12.231.42 20:49, 17 April 2007 (UTC) Ag for MemTech

The interesting thing, is that when Prescott was around (still is), people didn't care about their CPU wasting hundreds of watts of power and runnning very very hot. Nowadays, when Intel finally fought back, they actually used that to bash AMD, such irony... --antilived^{T | C | G} 01:06, 18 April 2007 (UTC)

Very good point antilived i didn't remember that!!! jajajaaj xD 200.35.168.129 16:18, 18 April 2007 (UTC) Ag for MemTech

Sorry if I sounded like I was critasizing your choice of Intel, but my personal opinion is not to wait for the quad core and just buy the PC/Laptop that suits you the most, and if price is a problem take a look on web sites such as pricewatch --Lwarf 06:31, 19 April 2007 (UTC)

Soup's on

I spilled soup on my trackball and keyboard. When I rebooted my PC, it said it couldn't display my video mode, so I tried again in the last working configuration. That seemed to work, but when I rebooted again to check, it said ntoskrnl.exe was corrupt. In desperation, I tried experimenting and found that if I substitute a mouse, I'm ok. What the heck is going on and should I be worried? This is XP MCE edition by the way. Clarityfiend 04:12, 16 April 2007 (UTC)

Sure it's not the keyboard that's messed up? From google searches, keyboard errors can cause that error message when there's a short, which probably happened when you spilled soup all over your keyboard and trackball. My advice is to clean them if you know how to open up your keyboard and trackball, if they're models that can be opened and cleaned. After that, if you still get the error, you'll have to replace whichever one is damaged from the soup. THe soup was good at least, I hope. --Wirbelwindヴィルヴェルヴィント (talk) 05:06, 16 April 2007 (UTC)

OK, but how the **** would a keyboard corrupt the Windows kernel? --wj32 ^{talk | contribs} 11:36, 16 April 2007 (UTC)

Disk needs fscking, probably. —The preceding unsigned comment was added by 203.49.240.43 (talk) 12:21, 16 April 2007 (UTC).

It's definitely the trackball. My PC appears fine without it and refuses to boot with it. What is fscking? Clarityfiend 15:38, 16 April 2007 (UTC)

File System Check, a Unix command similar to Windows' scandisk. --LarryMac 17:31, 16 April 2007 (UTC)

It doesn't corrupt the kernel. Nothing is actually wrong with the kernel, just that there's a physical short in the wire, so it gives the driver errors. --Wirbelwindヴィルヴェルヴィント (talk) 00:59, 17 April 2007 (UTC)

I'm assuming this is a laptop computer, right? Do we have any evidence to suggest that your soup didn't make it all the way to the laptop's main board, creating any number of now and future troubles? Soup tends to besalty, and all those chloride ions left lying around can't be good for anything.

Atlant 12:47, 17 April 2007 (UTC)

Just an everyday desktop. Clarityfiend 23:12, 17 April 2007 (UTC)

Does your PC do something special if you boot with a mouse button (or trackball button) held down? Macs used to do that. Perhaps your soup is causing the trackball button(s) to appear to be depressed and that is provoking an unusual (to you) startup behavio[u]r? It would be well within reason that the soup could have either mechanically jammed or electrically shorted one of the microswitches that are actuated by the trackball buttons. If so, opening it up and cleaning (with distilled water, isopropyl alcohol, or both) may well fix you up.

Atlant 13:17, 18 April 2007 (UTC)

Inode conversions

Is it possible to take an inode number, feed it to the OS, and then get back some info on the file quickly, such as the path to the file? It's quite straightforward to get a filename/path and get an inode number, can we do the reverse? (won't find be rather slow?) —The preceding unsigned comment was added by 203.49.240.43 (talk) 11:00, 16 April 2007 (UTC).

Try inode-cat, which is cat except it takes an inode argument. There must be better tools though... --h2g2bob 11:47, 16 April 2007 (UTC)

I need something a little more portable: From the inode-cat manpage:

 BUGS
      icat should support more file system types. Right now, support is  lim-
      ited  to  ext2fs when built on Linux, and ufs when built on Solaris and
      BSD systems.

I'll check it out, though.

—The preceding unsigned comment was added by 203.49.240.43 (talk) 11:50, 16 April 2007 (UTC).

In general, the answer is *no*, since more than one path can point to the same inode (see hard link), and most filesystems don't keep an inode => path mapping, just path => inode. "find -inum N" is probably the fastest thing that's guaranteed to work on a vanilla Unix system. --TotoBaggins 13:23, 16 April 2007 (UTC)

The question is ill-posed, albeit for subtle reasons. You ask for the path to the file; never minding that there might be more than one, let's say you want a path to the file. The problem is, files do not in full generality have paths! A file can be opened and removed and still exist as long as the opening process does; on many systems with a /proc filesystem (and perhaps others) such "undead" files can actually be relinked into the filesystem. So the fundamental issue here is the notion (of a file) that you want, which doesn't match the Unix semantics. Now, to be sure, it would be possible to provide an ENOTQUITEDEAD errno code for int getpath(ino_t inode,char *buf,size_t size) that indicated the odd case of a file with no names, but for simplicity, performance, and perhaps security (what if the path returned wasn't accessible to your process?) no one has implemented such a function. --Tardis 15:03, 16 April 2007 (UTC)

In principle, given an inode number, it is easy to get to the associated file's contents, and to its "stat" information (size, mtime, owner, permissions, etc.).

However, under all Unix-like filesystems, it is arbitrarily difficult (and quite often impossible) to determine its path. As several people have already pointed out, a file can very easily have multiple pathnames, or zero pathnames.

As someone already pointed out, the only way to even try to find the path to an inode number is find -inum. This is, to be sure, horribly inefficient. But there just isn't any other way to do it. The filesystem simply doesn't contain the information that would make the search any easier.

The bottom line is that, for a Unix-like filesystem, if your question is "given only a pointer to an open file, or that file's inode number, how do I determine the path to the file?", the only answer is, "You don't want to do that. Find some other way of arranging things so you don't need to even try."

For similar reasons, it is difficult-to-impossible for a running program to reliably find the directory its executable file is in.

In terms of cat'ing or stat'ing a file given its inode number, which as I said is in principle possible, in practice, there is no way to do it portably. There aren't any standard system calls which take an inode number and which let you get a handle on the actual file. So any program such as the "icat" mentioned above pretty much has to be nonportable, either relying on nonstandard system calls or other hooks into the operating system, or knowing how to grovel over a particular filesystem implementation to track its inodes down. —Steve Summit (talk) 02:51, 17 April 2007 (UTC)

In the end, I think I want something more like UUIDs or GIDs to identify files, and use hash tables or something to map IDs to files. I reasoned that since inodes are somehow fundamental to the file system, I'd be able to do this quickly, but had forgotten that inodes aren't unique.

However, you can create something semi-portable that is filesystem dependent however -- you would create an abstraction layer and modules for each various file system.

Flying Windows XP

I've installed windows xp sp2 on a portable hard drive. All works fine through the install and copies the setup files. When it reboots to continue it shows a XP loading screen but then goes to a Blue screen of death, saying something about problems with new hard drives or other new hardware. Is this one of those things where microsoft have stoped usage of their stuff on a portable device? I've done it with 95, 98 and ME and they worked fine. Think outside the box 11:19, 16 April 2007 (UTC)

Was it the upgrade file from SP1, or the entire XP? If it was the free upgrade file, you need to use it on the computer that has SP1. --TeckWiz ^ParlateContribs_@(Lets go Yankees!) 11:22, 16 April 2007 (UTC)

Sorry about the confusion. It was a full windows xp install on a empty har drive. The XP disk came with SP2 built in. Think outside the box 11:40, 16 April 2007 (UTC)

I don't know the exact solution though I would suspect modifying boot.ini to be the solution. Splintercellguy 15:03, 16 April 2007 (UTC)

Generally it's a bad idea to do that with Windows, as it's not very flexible with the hardwares once installed. I would just stick with a Linux LiveCD or something. --antilived^{T | C | G} 03:12, 17 April 2007 (UTC)

It doesn't sound like modifying the boot.ini will help, since it is booting (and then crashing). It is possible to install XP on removable media, they don't have any mechanisms in place to block it as far as I know, but a plethora of things could be going wrong here (what with it being Windows and all). -- Consumed Crustacean (talk) 14:58, 17 April 2007 (UTC)

the problem is with the boot...i copied a Live CD to another small HDD i had laying around and it didnt work...then i cloned the exact partition onto the HDD and it did...the only thing that changed when i analyzed both is one had bootable marked in the MBR and it had a boot.ini different from an install...you might wanna look into how a Live CD is made and then apply that to your project. Hope this helps PS try to get a standalone Live CD many depend on BartPE to boot..and this will not help...for you will need BartPE on your HDD in order to boot If the boot.ini is not the issue then you run into a problem of having different hardware on your computer mainly between chipsets (Read VIA,etc) if you have this problem there is a workaround..i do not remember the program name or whatever but what it does is it cleans the saved hardware info on your Windows instalation so that it redetects everything everytime you put it on a new computer...just google it or something or look on the Live CD forums or the unattended CD forums i do believe they use the same on some of those 200.12.231.42 21:29, 17 April 2007 (UTC) Ag for MemTech

Thank you for all your help. I've decided to use a linux live cd called slax (150mb) and its sucessfully installed on a 250mb usb stick. I've tried it with a portable hdd but again had the same problem as before. seems neither liked the hard drive. Think outside the box 11:35, 19 April 2007 (UTC)

Regarding e-duc low cost Pc of South America

I heard of E-duc low cost PC. I think it is in Brazil but I am not sure. I want to know the configuration and price of that PC. But their website is not in English and I cannot read from that. I also found no articles for that PC in Wikipedia and also web search did not help. Anyone knows its

1)screen size? (if possible pixels, is it widescreen?)

2)configuration?

—The preceding unsigned comment was added by 59.92.105.38 (talk) 13:38, 16 April 2007 (UTC).

You can translate web pages from most common languages at Babelfish: [2]. The translation is not very good, but might be good enough for your purposes. If you post the web address on the Language Ref Desk you can likely obtain a better translation. StuRat 16:46, 16 April 2007 (UTC)

Their website is www.e-duc.com

The automated translator can't handle text inside a video like that, so you would need to manually type it into the translator. However, I didn't see anything like technical specs (I was expecting to see a page full of numbers) to translate, so I think you would need to use the e-mail option to ask them specifics. StuRat 16:20, 19 April 2007 (UTC)

distances

Is there a good program for finding the distances between cities and other locations? —The preceding unsigned comment was added by Ipmin (talk • contribs) 13:42, 16 April 2007 (UTC).

When you get directions from most map web sites such as google maps The directions will come with a driving distance. This may not be the exact straight line distance, but should be a good starting point. If you know the latitude and longitude of you locations you can get a distance calculation here LatLong Distance - Czmtzc 13:56, 16 April 2007 (UTC)

I'm not sure if they want the driving distance or the minimum (great circle) distance on Earth's surface. Even less likely would be the straight-line distance (through the Earth). StuRat 14:30, 16 April 2007 (UTC)

Well that is wy I answered both ways. Could you provide a reference to determine the through the Earth distance? -Czmtzc 14:36, 16 April 2007 (UTC)

You could use that Google pedometer thing. I saw it once, but had no need for it. Let lets you click one spot on google maps, then another spot. Then, it tells you the distance if you were to walk in a straight line between the two (or more) points. --Kainaw ^(talk) 14:53, 16 April 2007 (UTC)

I would assume that follows the curvature of the Earth, though, as few people can walk through the Earth (ghosts perhaps ?). StuRat 16:42, 16 April 2007 (UTC)

Which RFID chips are safe to implant in human bodies?

I have read about some people implanting RFID chips inside the skin, generally between the base of the thumb and index finger. I would like to know specifically which chips are specifically used for this purpose. —The preceding unsigned comment was added by 141.213.222.28 (talk) 13:44, 16 April 2007 (UTC).

Presumably the ones used for non-human animals would be safe for us, too. PetID is the big one in the USA. --TotoBaggins 14:04, 16 April 2007 (UTC)

I wouldn't assume that, for two reasons. One is the longer life span of humans than cats and dogs. Something entirely safe for 20 years may start to break down after 30 or 40 years and cause problems. The second reason is that a higher risk might be acceptable for pets, since it is more important for them, as a lost pet may be put to sleep if the owners can't be located, and we hardly ever do that with people. :-) StuRat 16:37, 16 April 2007 (UTC)

They are usually encased in a protective plastic tube. It protects you from the chip and the chip from you. Still, the body tends to build up scar tissue around the implant. --Kainaw ^(talk) 14:54, 16 April 2007 (UTC)

KDE Scripting

I have a directory of around 1,000 HTML documents. I want Konqueror to open and print each one. Is there such thing as KDE scripting where I can script: For each file in dir, konqueror file, tell konqueror to print, tell konqueror to close, end for? --Kainaw ^(talk) 16:18, 16 April 2007 (UTC)

I'm not sure why it would need to be specific to KDE, wouldn't a general shell script (bsh, csh, ksh, etc. ?) work ? StuRat 16:32, 16 April 2007 (UTC)

Konqueror is part of KDE. So, I'm trying to figure out how to control Konqueror from a shell script. --Kainaw ^(talk) 16:42, 16 April 2007 (UTC)

What if you just browsed to that directory with Konqueror, selected all the files, right-clicked the selection, and did "Actions => Print"? I've just done that with an ancient version of KDE, and it seemed to do the right thing. --TotoBaggins 17:17, 16 April 2007 (UTC)

It printed all the files - no problem. However, they are HTML files. So, it printed the source code of the HTML, not the rendered HTML. That is why I need it to open in Konqueror and then print. I also tried html2ps - it rendered the pages so badly that they were barely readable. --Kainaw ^(talk) 17:25, 16 April 2007 (UTC)

How about HTMLDoc? I have no experience with it, I'm just searching around. --LarryMac 18:36, 16 April 2007 (UTC)

I've used that before. It uses html2ps to convert HTML to postscript (or pdf). My solution... cat reports/* > allreports.html. Then, add <style>body{page-break-after:always;}</style> to the top of the conglomerate file. Finally, open the 5.1MB file in Konqueror and print it. --Kainaw ^(talk) 19:16, 16 April 2007 (UTC)

I've just been poking around with DCOP. It's pretty neat. Something like the bits below might do the job. I don't have a printer to test it on, but it does load each page and try to print it. --TotoBaggins 19:55, 16 April 2007 (UTC)

konqueror &
sleep 2
konq=$(dcop 'konqueror-*' | head -1)
cd reports/
for i in *.html; do
    echo "Doing $i"
    dcop $konq 'konqueror-mainwindow#1' openURL file://$PWD/$i
    sleep 1  # make this longer if the pages are slow to load
    dcop $konq html-widget1 print 1
done

I tried that (just for fun since I already solved the problem). It doesn't work because you have to manually select a printer. All you get is an error that the printer is null. --Kainaw ^(talk) 05:14, 17 April 2007 (UTC)

I set the printer to "print to PDF", and got it to work. It just prints to /root/print.pdf, and I move the PDF aside for each file. It's pretty slick, and making me change my previously dim view of KDE. --TotoBaggins 15:39, 17 April 2007 (UTC)

for i in *.html; do
   echo $i
   dcop $konq 'konqueror-mainwindow#1' openURL file://$PWD/$i
   sleep 2
   dcop $konq html-widget1 print 1
   sleep 5
   mv /root/print.pdf $i.pdf
done

Jesus, you run as root?? —The preceding unsigned comment was added by 149.135.28.177 (talk) 10:29, 18 April 2007 (UTC).

Jesus *would*, I imagine, but I only do so on the test machines at my office, which all get Ghosted weekly anyway. --TotoBaggins 13:23, 18 April 2007 (UTC)

gcc parse tree

GNU gcc with option -fdump-translation-unit puts out the parse tree to a file. Is there a propper description of the format of this file anywhere? 84.160.218.220 18:06, 16 April 2007 (UTC)

I poked around and didn't see one, but this Perl module parses it, so that might give you a good start. Also, the GCC source might have a pointer to a formal spec. --TotoBaggins 15:44, 17 April 2007 (UTC)

Thank you, you gave me the missing clue. I was just the man searching the horse he's riding. The perl module cited the reference: gcc-4.1.2/gcc/doc/gccint.texi 84.160.202.36 18:00, 17 April 2007 (UTC)