Computing desk
< March 3	<< Feb \| March \| Apr >>	March 5 >

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

March 4

using t-moible's web'n'walk under linux?

possible? I'd like to just try it in knoppix, otherwise I can't switch to linux...

I have a USB modem from t-mobile that gives me the web 'n' walk service, and the fact that it supports mac too leaves me hopeful... —Preceding unsigned comment added by 79.122.69.66 (talk) 00:01, 4 March 2008 (UTC)[reply]

I think your first problem would be having your computer detect and get your modem and other hardware to work. I am not sure what web'n'walk is but if the hardware works, you might have solved the problem. I see no reason why they would block access to the network itself. Kushal 05:17, 4 March 2008 (UTC)[reply]

page views on any wikipedia article?

Is there a way to get such info? 199.76.153.227 (talk) 03:46, 4 March 2008 (UTC)[reply]

Could you please clarify exactly what you mean by that? Useight (talk) 06:34, 4 March 2008 (UTC)[reply]

I think the OP means to ask if it is possible to find out the number of times a given Wikipedia article has been viewed. The answer is no, since the MediaWiki feature of page view counting is disabled for Wikipedia. -- Meni Rosenfeld (talk) 09:35, 4 March 2008 (UTC)[reply]

This site has some data, however. Algebraist 13:35, 4 March 2008 (UTC)[reply]

The MediaWiki counters being disabled is only part of the story. Anon visitors (the huge majority of page views) are typically served entirely from the squid caches, without troubling a web server at all. It is possible to collect statistics from the squid, but this is sometimes disabled for performance reasons (it used to be disabled most of the time; I don't know the current situation). We're perhaps unusual among high-traffic web organisations in that we largely don't care about how many hits a given article got, and we wouldn't do anything (editorially) if a given article was getting "too few" (whatever that might mean) hits. -- Finlay McWalter | Talk 13:46, 4 March 2008 (UTC)[reply]

Windows Vista idle mode

According to an article at New York Times, Windows Vista consumes "3 watts per hour" in idle mode. Am I missing something? Does it mean a minimum of three watts? Does the power consumption of a computer not depend on what hardware configuration you use? Kushal 05:11, 4 March 2008 (UTC)[reply]

I'm more concerned with the phrase "watts per hour", which clearly indicates that the writer has no idea what a Watt is. -- Meni Rosenfeld (talk) 08:39, 4 March 2008 (UTC)[reply]

In fairness to the journalist, he was probably just parroting what it said in the corporate press release. :) --Sean 14:03, 4 March 2008 (UTC)[reply]

Maybe it means that the power consumption increases by 3 watts every hour? :D\=< (talk) 16:07, 4 March 2008 (UTC)[reply]

I don't know what this "idle mode" is, but if it is anything like "hibernate" then the computer only needs enough power to keep a copy of the OS state in RAM, to my knowledge. Since only the RAM is "on", and there is no data going to and fro between it and the CPU, then not much power is needed. 206.252.74.48 (talk) 16:49, 4 March 2008 (UTC)[reply]

That's suspend, or "sleep". Hibernation writes RAM to disk and requires no power at all. --Tardis (talk) 19:37, 4 March 2008 (UTC)[reply]

Linux permissions: "Executable" MP3s?

In Linux, on an ext2 or ext3 file system, do permission flags -rwxrwxrwx for an MP3 file indicate that the whole file is corrupt?

On a related note, is there a way to search for files by permissions, but (unlike a grepped ls -l) get only their paths as output? Neon Merlin 07:07, 4 March 2008 (UTC)[reply]

Anything -rwxrwxrwx is a sign that the file permisisons were set by someone without a clue. The file contents are not necessarily corrupt, but shouldn't be trusted too much, being modifiable by every user on the system. For the second quesiton, use find. For example

find . -type f -perm -0002 -print

looks for regular files with the 002 (world writable) permission enabled. --tcsetattr (talk / contribs) 08:03, 4 March 2008 (UTC)[reply]

Copying or unzipping a file from a Windows system often makes it executable, presumably since having executability turned on is mostly harmless, while having it turned off will break things you want to execute. --Sean 14:00, 4 March 2008 (UTC)[reply]

Does tar preserve permissions? Anyone know? :D\=< (talk) 16:07, 4 March 2008 (UTC)[reply]

It does, but if the file was tarred on, say, a FAT filesystem, you can't expect tar to guess reasonable permissions to store. --Tardis (talk) 19:36, 4 March 2008 (UTC)[reply]

GNU tar has --same-permissions and --no-same-permissions so you can choose when you extract a tar file whether you want to trust the permissions inside it. --tcsetattr (talk / contribs) 21:39, 4 March 2008 (UTC)[reply]

So could you just download a tarred version of su that doesn't require any password, and that has the setuid flag set to root, and have easy privilege escalation? :D\=< (talk) 16:40, 5 March 2008 (UTC)[reply]

tar is running as you, and so can't very well cause root to own a file (for this very reason). (If it could, you wouldn't have to tar su to mount this attack, because you could just make all your programs run as root.) Setuid doesn't magically go to root, although that's its most common function. It goes to the owner of the file, which will forever be you. --Tardis (talk) 17:35, 5 March 2008 (UTC)[reply]

But if tar preserves permissions, can't I get an executable owned by Tardis if he tars it and sends it to me? Or by root? :D\=< (talk) 01:40, 7 March 2008 (UTC)[reply]

If you're not root when you untar it, then you don't have the ability to create files owned by anybody but you. When root untars things, he's gotta be careful (and should probably use --no-same-permissions if he didn't create the tarball himself). --tcsetattr (talk / contribs) 01:47, 7 March 2008 (UTC)[reply]

Automatically update a website

I'd like to be able to update my website automatically based on the current date and time so I don't have to be sitting behind my computer while I could be having holiday fun with my family. I'd be happy to learn some scripting language to do it, but I have yet to figure out what my best option would be. Can anyone help me out and point me to an example script that could do what I want. - 131.211.161.123 (talk) 15:32, 4 March 2008 (UTC)[reply]

PHP would be my personal preference, but almost any web scripting language could do what you want. I would set up everything either in a database, or file, then code into the file that people will be viewing, what should be displayed, depending on the date - there's a range of date functions that would be suitable. Good luck :). Ale_Jrb ^talk 16:22, 4 March 2008 (UTC)[reply]

But don't expect software to write blog posts for you. --grawity ^{talk / PGP} 17:38, 4 March 2008 (UTC)[reply]

Depending on what the website actually is, there might even be easy to use packages out there. WordPress, for example, is a popular blogging software written in PHP, and you can set up posts to "go live" at any time you want in the future. --98.217.18.109 (talk) 20:06, 4 March 2008 (UTC)[reply]

You are getting my drift with regard to what I want to achieve, but unfortunately, I'm not building a blog. Just a regular website with text and images. - 87.211.75.45 (talk) 23:17, 4 March 2008 (UTC)[reply]

I would personally use cron for this. Assuming your website is in /var/www/htdocs, I'd have a cron job look for /var/www/htdocs-YYYY-MM-DD each day. If that directory exists, replace /var/www/htdocs with it. Then, you won't have to learn a scripting language. Just copy /var/www/htdocs to something like /var/www/htdocs-2008-05-23 for the version that will appear on May 23rd. Edit the copy for that day. When the day comes, cron will move over the new content. -- kainaw ™ 04:00, 5 March 2008 (UTC)[reply]

browsing in TV using Wii

Can anyone give me more details on browsing Internet in TV using Wii and the opera browser pre installed in Wii? Anyone here used that? Will that work in my CRT TV? What resolution may I expect? Thanks in advance.

It'll work with any TV. I really can't give you a number for the resolution - it will look like anything else you have connected to your TV (the text will be a little hard to read, but you can zoom in so it isn't a problem). 206.252.74.48 (talk) 16:19, 4 March 2008 (UTC)[reply]

And you scroll up and down by pointing at arrows on the edge of the screen. Useight (talk) 18:04, 4 March 2008 (UTC)[reply]

"resolution" is a bit complicated when you're talking about analog TV, because a) it's interlaced and b) it's not entirely consistent about what is and what isn't displayed on the screen. NTSC nominally has 525 lines and PAL 625, but inconsistencies between equipment (and sometimes use of some lines for special purposes) mean that a bit less is really useful. TV encoder chips seem to run to 480 lines for NTSC and 576 or so for PAL, and about 700 "pixels" across (really 700 horizontal transitions, as there's no way to reliable correlate changes in the colour as the raster proceeds horizontally with the actual little blobs of phosphor on the screen). So (lets talk NTSC from now on, as it's the poorer performer) you're looking at 700x480, which is about what you get from an ASUS Eee PC. But it's not that good: the signal is vertically interlaced, with each line only being sent every other time (so only 30Hz for NTSC, 25Hz for PAL) - that's slow enough to be visible - a single pixel horizontal line will seem to flicker very noticably (and a two pixel line appears to shudder up and down); the same problem affects text too. Worse, analog TV components have some nasty reactiveness characteristics - change from black to white horizontally too quickly and the analog components ring, leaving little horizontal "echos". And (depending on the signal path) other nasty things like dot crawl can happen, for which a robust rendering solution has to cater. So, to avoid the text being horribly flickery, systems that render text for analog tvs (be that for broadcast or stuff like a WebTV or a Wii) use big fat fonts with smoothed (microgradient) edges (it's antialiasing, kinda, but adapted for the weird characteristics of an interlaced analog display). The end result of this is that you have to render in (what seems like, when compared with computer displays) very large fonts - glyphs have to be 20px or more to be reasonably readable. Given the very limited screen size you're going to be rendering text at around 35x25 characters per screen - you can fudge that a bit, at the expense of quality, and some systems do clever things with character design and lowered contrast, but that's the ballpark. That means (sorry for the huge technical exposition to get here) that you get very few letters on the screen, and browsing on a TV screen seems very cramped and unpleasant (requiring scrolling, zooming, or unreliable relayout to make things work at all). Things aren't quite so bad for graphics - they're still subject to interlace flicker, but some platforms run a gentle gaussian filter (with a vertically biased filter kernel) to reduce bad flickering. In super summary: it's pretty low resolution, and it'll feel even lower due to the big fonts. -- Finlay McWalter | Talk 18:44, 4 March 2008 (UTC)[reply]

Excellent answer! TL;DR: gj. :D\=< (talk) 16:36, 5 March 2008 (UTC)[reply]

Can you please explain to me what "TL;DR:" means? 206.252.74.48 (talk) 17:12, 7 March 2008 (UTC)[reply]

Type it into Google and see what you get! --70.167.58.6 (talk) 23:02, 9 March 2008 (UTC)[reply]

"Too long; didn't read." The Opera browser is no longer a free download for the Wii. It will now cost you 500 WiiPoints (approx $5USD) to purchase it. Once you do, though, it is an amazingly capable browser. It renders the Acid2 test perfectly, and has built-in SVG. It will render cell-phone web pages (WAP), and has a somewhat older Flash plugin. Youtube videos work, for instance, but some other sites demand that you upgrade your flash. TV viewing does leave quite a bit to be desired, but it really isn't too bad. --Mdwyer (talk) 21:48, 10 March 2008 (UTC)[reply]

Unix filesystem spanning multiple disks?

I've heard it's possible to make a Unix filesystem spanning multiple disks. How is this done? Does Fedora 8 Linux support such a filesystem natively or do I have to install some experimental third-party code? More importantly, how does the filesystem work? Once I make the filesystem, are the disks set in stone, or can I switch them on and off dynamically? What if I don't power up one of the disks? What decides which file gets written onto which disk? JIP | Talk 20:35, 4 March 2008 (UTC)[reply]

RAID and LVM have been part of the Linux kernel for a while now. I've never looked at Fedora, but I'd bet they're included as part of the base system. LVM makes it possible to expand an existing filesystem onto a new disk, so the size is not set in stone. RAID gives you redundancy, so everything keeps working after a disk failure. --tcsetattr (talk / contribs) 22:31, 4 March 2008 (UTC)[reply]

Fedora does use LVM by default. As such, you can add/remove drives from a volume. The volume of multiple drives appears to be a single logical volume to the user. I use this for database servers by putting 8 drives into a single volume. I've been working on getting LVM to spread the load evenly across all 8 drives to improve disk IO times. Right now, that is my bottleneck when running queries. -- kainaw ™ 03:56, 5 March 2008 (UTC)[reply]

To distribute io, you probably want to use RAID, because with 8 disks, redundancy doesn't hurt. LVM also listens to --alloc, which sets an allocation policy, but it says that it isn't implemented. You can use raid arrays as lvm physical volumes. 209.151.140.55 (talk) 23:15, 8 March 2008 (UTC)[reply]

microsoft

i was reading the microsft article and its endless antitrust law suits.Was wondering why dont they just comply once and for all.isnt it finacially suicidal for them to be paying millions and millions of fines?or is there a bigger to all this,coz i can read mischief.its like the money wasnt an issue,it was all about hurting their competition.?what really was on Bill Gate's mind?

2.Really,i need help to re-enable my sytem resore function.A virus has just infected my pc and am in trouble.its not on my computer icon when you right click.its been disabled by group policy.help.Or is there a software that can help me?

re antitrust 1. big companies are always in trouble somewhere... 2. EU antitrust case - one possible reason why MS has been 'tardy' to supply info to third parties is that it's not in their interests to do so - ie why bother - 'what's my motivation for doing this' or 'why should I do the work for someone else'. 2. Media player bundle problems - as a consumer I might appreciate these bundles - in the same way I appreciate getting a free set of lights when I buy a bike - does that make the bike shop guy the 'bad guy'? Also it is possible for a business to produce standalone products that are incompatable with all others - typically this is the case for the vast majority of products.. eg can you put a GM engine/steering wheel/gearbox in a Ford (car analogy) - not very easily - and why would you expect to? (rhetorical) Thus ends the pro-microsoft argument. A more cynical person might suggest you read Embrace, extend and extinguish I do too.

Also a class action would be far more harmful than fines from governing bodies etc.. just as an example of what would hurt - ie loss of faith from customers....87.102.44.156 (talk) 21:29, 4 March 2008 (UTC)[reply]

Virus problem - antivirus software looks like the answer here.. However if you already have the virus (seems likely) - it may attempt to block you trying to get rid of it??? You might need to re-install. Someone else may be able to help more with this hopefully.87.102.44.156 (talk) 21:16, 4 March 2008 (UTC)[reply]

Your bicycle analogy falls flat because your bike shop guy is not a monopoly bicycle supplier. If he were, then companies who make bike lights would have a legitimate complaint because your guy would be effectively putting them out of business. Similarly, neither Ford nor GM is a monopoly supplier of automobiles. The fact remains that there is a standing court decision specifically finding that Microsoft has a monopoly on operating systems sales, so they have to play by certain rules. --LarryMac | Talk 21:31, 4 March 2008 (UTC)[reply]

There's an interesting conflict here between 'being competitive' and 'achieving the competitive advantage' ie winning - then the goverment tries to push reset. Why?87.102.44.156 (talk) 22:16, 4 March 2008 (UTC) see also Dilemma 87.102.44.156 (talk) 22:19, 4 March 2008 (UTC)[reply]

Because it's bad for the consumer (the citizen). A nightmare really. Politicians want to be voted for so they try to avoid nightmares for their constituency. :D\=< (talk)

Once a monopoly sets into place then it makes it so that there is no more incentive for competition. See, for example, the long stagnation of Internet Explorer, a product which, once it effectively forced Netscape out of business, was not updated at all until new competitors finally appeared on the scene. --98.217.18.109 (talk) 01:31, 5 March 2008 (UTC)[reply]

The main reason anyone buys Microsoft's flagship products is their monopoly. So Microsoft has a choice of either continuing their monopolistic policies, paying a few fines and still ending up with an annual profit of $14 billion, or playing fair and having no profit at all. -- Meni Rosenfeld (talk) 10:24, 5 March 2008 (UTC)[reply]

Some say that there is big money to be made by controlling the market for the media player software (if you play it for the money). I think the belief is that people will do a lot of buying music, movies, and videos directly from the media player software. So they want to retain a place in the market. Am I right? Kushal 10:56, 6 March 2008 (UTC)[reply]

Wikipedia:Reference desk/Archives/Computing/2008 March 4

Contents