Computing desk
< January 21	<< Dec \| January \| Feb >>	January 23 >

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.

January 22

Ubuntu

How do I look at system information (CPU power, RAM, etc.) on Ubuntu? --T H F S W (T · C · E) 03:06, 22 January 2011 (UTC)[reply]

gnome-system-monitor, lshw-gtk, top, htop, cat /proc/cpuinfo, free, etc. ¦ Reisio (talk) 04:25, 22 January 2011 (UTC)[reply]

I'm afraid I don't quite understand what you mean. Are these online or in my desktop? One other thing, anyone know of a DVD driver for Ubuntu 10.10 ('Maverick Meerkat')? --T H F S W (T · C · E) 05:20, 22 January 2011 (UTC)[reply]

I'm guessing you type the command into a terminal window Nil Einne (talk) 05:48, 22 January 2011 (UTC)[reply]

thats right, you type that command in the command line terminal. unfortunately, with most linux distrobutions, you are still expected to do way too many simple things using the command line terminal, which most computer users nowadays have no knowledge or patience for whatsover. Roberto75780 (talk) 07:53, 22 January 2011 (UTC)[reply]

Wrong. It is merely simpler to tell someone what command to copy, paste, and run than telling them how to click around in menus for days as Windows users must. :p ¦ Reisio (talk) 19:30, 22 January 2011 (UTC)[reply]

I agree entirely. However, your answer wasn't "novice-friendly" because it did not specify that the user should use the terminal. The inclusion of three extra words ("Use the terminal:") would have made your otherwise flawless answer much more newbie-friendly. Keep in mind that not all users have extensive (or any) *nix experience). Nimur (talk) 19:41, 22 January 2011 (UTC)[reply]

I'm okay with people having to ask follow-up questions as he did. Many reasons to not go overboard on a first response. ¦ Reisio (talk) 20:07, 22 January 2011 (UTC)[reply]

That is also some of the beauty of the GNU/Linux world, IMHO. --Andreas Rejbrand (talk) 11:55, 22 January 2011 (UTC)[reply]

I must say, over the three years of using Linux right now, I have gotten to like and use the command line interface for things which I can also do from menus - playing music/videos, simple image conversion, software updates and so forth. IMHO it's far less hassle in a great majority of cases. --Ouro (blah blah) 21:02, 22 January 2011 (UTC)[reply]

Why not just use the 'System Monitor' application? On the top bar click System to get the drop down menu then 'Administration' then 'System Monitor' --Aspro (talk) 13:47, 22 January 2011 (UTC)[reply]

About your other question (about a DVD driver) - Ubuntu already ships with drivers to read and write CDs and DVDs. Beyond what, what do you want to do? VLC will play a DVD (as will Totem ("Movie Player"), although it will ask to install a special package to allow it to decrypt commercial DVDs). K3B and Brasero will burn a DVD. Bombono will let you author video DVDs. K9 copy will let you copy video DVDs and Thoggen will recode them for another destination like an iPod. All of these are either installed by default or can trivially be installed using Ubuntu Software Center or Synaptic. -- Finlay McWalter ☻ Talk 16:18, 22 January 2011 (UTC)[reply]

Easier method - install package hardinfo. You can do this by typing sudo apt-get install hardinfo into emulator, or by going to your system menu -> administration -> Update manager. Quick search for hardinfo. After you install this package, it will show up on your applications menu -> system tools -> System Profiler and Benchmark. Also, I further recommend you ask questions at http://www.ubuntuforums.org; they are much more knowledgeable about things Ubuntu. Magog the Ogre (talk) 19:41, 22 January 2011 (UTC)[reply]

I think you mean "package manager" not "update manager" - the latter only deals with updating existing packages with new ones for security/bug fixes. CS Miller (talk) 20:28, 22 January 2011 (UTC)[reply]

OK, I got VLC. It's not working though. I put the DVD in and hit play, and it just gives the filename of the DVD for a second and ten goes blank. --T H F S W (T · C · E) 23:42, 22 January 2011 (UTC)[reply]

hardinfo is really cool. Where has it been my whole life? Paul (Stansifer) 15:54, 24 January 2011 (UTC)[reply]

wget -- downloading to a directory and renaming a file

I can't quite figure out how to do something with wget in OS X. I would like to be able to download the index file of a site to a specific directory and rename it with the current date. For example:

wget http://wiki.riteme.site gives me the index.html file in my home directory, but I'd like this instead: /wiki-archive/2011-01-22-wikipedia-index.html.

Any help would be much appreciated. --CGPGrey (talk) 12:28, 22 January 2011 (UTC)[reply]

"-P" or "--directory-prefix" will allow you to set the directory, and "-O" or "--output-document" allows you to set the filename. There is more info in the wget manual. To set the filename to the current date you'll have to work out how to print the data in your OS. I have never used a Mac so I don't know how to do that, but if it helps you work a solution to do it in .bat on Windows would be "%date:~-4,4%-%date:~-7,2%-%date:~0,2%-%time:~-11,2%%time:~-8,2%.html" 82.44.55.25 (talk) 12:40, 22 January 2011 (UTC)[reply]

Thanks for pointing me in the right direction. I've found out that the answer is:

wget -O "/path/to/file/`date +"%Y-%m-%d --"` archive.html" http://wiki.riteme.site --CGPGrey (talk) 17:49, 22 January 2011 (UTC)[reply]

New components for an older PC

Hello. I'm about to buy some new components to revive a ~6 years old business PC for multimedia and gaming use. However, there are a few things that are confusing me.

First the PSU issue. I have no idea what kind of PSU is installed. The label on the side does say it's "ATE-OK". Is ATE the same as ATX? In other words, can I be as sure as usual that an ATX PSU will work? Also, most of the PSUs I find in the Internet only mention the amount of SATA connectors. Since some of the hardware is rather old, I'm not sure whether the new PSU will have the required connectors for them as well. I'm no expert on connectors, and I'm not going to query about the amounts of every different type (especially 'cause I don't know them all) but is there some specific connector that's fallen out of use and may be unavailable on newer PSUs?

And then, the graphics card. I'm thinking about Nvidia Geforce GTS 450, but on the technical specs on the site it says that it needs "two free card slots". I reckon it means that it needs space for air circulation, am I right about this?

Thank you all! 88.112.51.212 (talk) 13:14, 22 January 2011 (UTC)[reply]

I would be careful here. I believe one of the major drivers of motherboard changes have been the ever increasing demands of graphics cards. In my experience, the motherboard in an old PC places severe limitations on which new graphics cards are still compatible with it. And unless you intend replacing the CPU and clock, I would imagine the gaming performance would suck, even if you manage to get a graphics card that will work with the motherboard. Astronaut (talk) 13:56, 22 January 2011 (UTC)[reply]

Actually the Cpu is fine, surprisingly. I guess the computer was built for computing performance with average graphics back then. My Mobo has the required PCI-E slot, so it should fit in, but if I've understood correctly there are different types of PCI-E slots as well, right? 88.112.51.212 (talk) 14:25, 22 January 2011 (UTC)[reply]

(EC) Not really true on the motherboard point. If you the motherboard has a PCI Express x16 slot, then most graphics cards should work (in theory all should but I would never rule out compatibility problems). It's likely to only be 1.0e but unless it's an extremely high end card (and even then I'm not sure whether they've really reached a level that it's starting to become a bottleneck) it's not likely to make much of a difference. See for example this showing a GTX480 running on with 4 and 8 PCI express 2.0 lanes (8 lanes is equivalent to 16 lanes for PCI express 1.0) [1]. I would however be more concerned about the PSU and also concur the OP may want to consider whether the other new components will be enough to ensure the graphics card isn't overpowered compared to the rest of their system. ATE OK is just some sort of testing label. If that's the only label on your PSU (i.e. no power specs, not even a brand name), I wouldn't trust it with your current system let alone any upgrade. Oh and I forgot to mention it's correct if the card says it's two slots that mean it needs an additional slot for the HSF, this slot will have to be to the right of the card if the part of the card that plugs into the slot is on the ground and monitor etc connections on the card is facing towards you (look at an image of the card to get an idea). It obviously doesn't matter what sort of connection if any is present in the slot on your motherboard provided you don't need to use it. You may also want to check cleareances at the end opposite of the monitor connectors if it's a long card. Nil Einne (talk) 14:30, 22 January 2011 (UTC)[reply]

Thanks. The card isn't exactly high-end, and it won't be too overpowered, as my current Intel Integrated Graphics Card is the limiting factor of my gaming performance ATM~(I've already upgraded the RAM, the processor is adequate and performance is good on non-graphic benchmarks). The PSU does have some labels, mostly in Chinese though. I couldn't find anything of interest apart from the manufacturer (LITEON) and some power statistics (300W total output peak, it says). So, I was right about the cooling system. My current card layout is as follows: Network card on top (PCI I think), below it two empty slots that look similar to the PCI-E although very short, and then the PCI-E. I'm not planning to use these shorty slots so the card'll probably fit right in. Thanks for your help, and if you have further pointers, I'll be listening! 88.112.51.212 (talk) 16:27, 22 January 2011 (UTC)[reply]

I agree the graphics card isn't highend and the Intel integrated card is a very major bottleneck for gaming on the other hand the GTS450 is I think what I'd call a high mid so I don't know if I'd use it e.g. on a weak dual core with 2GB of RAM. (Not to say I wouldn't upgrade but I might not go that high.) However I appreciate this a complex equation since it depends e.g. whether you're likely to keep the card with any upgrades and also on pricing since even if the 450 is bottlenecked in some games by the rest of the system it may still be worth it price/performance wise. However this isn't something I've looked in to for a while and I don't game much nowadays so this is just a general explanation of what I was thinking, you appear to have looked in to it and I'm not trying to change your mind.

The short PCI-E slots are probably 1x (1 lane) PCIe slots. LiteOn is a known brand, so the PSU isn't as terrible as I thought. Also if you live and bought your PSU in Finland, I've heard before e.g. [2] that the Finnish regulators are fairly proactive in keeping the worst of the worst PSUs out of the market. (This compares to NZ where while we have decent consumer protection laws most computer stuff is still not really something the regulators ever do much about.) Still 300W is likely pushing it, may be doable depending on your system with a excellent PSU but this might not be yours (particularly not since it's 300W peak). From a quick search I think GTS450 cards need a 6 pin PCI express power connector. Some cards come with adapters for molex, these would be okay with some decently speced but old PSUs but I wouldn't recommend them for yours.

So yeah, I'd recommend a new PSU. At a random guess I'd say anything 380W and over should be more then enough, I'd concentrate more on a decent PSU then getting more. (I've spoken on this at length in the RD before, don't trust any claims that you need outrageously high power for a system, most of these are based on power usage figures that just add the maximum possible for every component despite the fact this never happens and also the assumption you never want to get close to the maximum for a PSU since with crap ones they tend to die. [3] is a good place to get an idea of real world PSU requirements.) It's difficult to recommend brands since these depend on where you live but something Enermax, Corsair, OCZ, Seasonic, Antec and a few brands I probably can't remember off the top of my head are usually good bets. I'd also consider something that's 80 PLUS certified since even if you don't care about the energy savings it's ensure some minimum quality (particularly nowadays when so many have it).

There hasn't been any major changes in connectors, your motherboard may only have 20 pin ATX but you'll normally have no problems plugging the 24 pin ATX in to the 20 pin motherboard slot (it's keyed so you can't do it wrong) and in any case many PSUs still come either with an adapter to convert the 24 pin to 20 pin or the 4 extra pins can be disconnected from the 20 pin to give a 20 pin ATX connector. Also many PSUs will now have an 8 pin +12V motherboard connector, your motherboard will probably only have 4 pin if any but again this shouldn't be an issue. A modern PSU will also likely have more SATA and less molex but unless you have a ton of drives or whatever with molex power connections this shouldn't be an issue (and there's always Y cables). There will probably be few (1 if any) FDD connectors but you can always use an adapter and I'm not sure you'll be using one anyway. Modern PSUs don't have -5V but this has been the case for a long time so I doubt your motherboard uses -5V. (On very old systems some particularly high speced modern PSUs sometimes have problems starting because of insufficient loading particularly 12V which wasn't used so much in the past but this much older then your system.)

As I hinted above PCIexpress is supposed to be backwards and forwards compatible so you should have no problem with it working in theory. I have heard of compatibility problems before although I think this was with forward compatibility (using a PCI express 1.0 card in a PCI express 2.0 slot) and was a while ago. I'd never rule them out since poorly designed systems or sometimes even flawed specs can allow problems. (I know of a case where a Maxis PATA HDD didn't work with a certain motherboard, not sure why.) Despite what you may read in some confused forums and articles PCI express 2.0 slots provide the same amount of maximum power PCI express 1.0 slots could, 75W so it definitely shouldn't happen. I have heard of motherboards that couldn't provide enough power for certain AMD CPUs even though they were only using what had already been in the spec but the PCI express thing is even more strange since there have been cards needing extra connectors since the beginning which I would have thought drawed close to 75W from the slot. Without knowing the details below, one possibility is it's the PSU's fault (which also should never happen even if you draw too much according to the ATX standards, it should just shut down). I believe Finland has good consumer protection laws so if you do live and bought the components there I'd expect you'd be entitled to have any damaged components replaced if it does happen, whatever is at fault and whenever you bought them since it would likely be a design defect (unrelated to wear and tear or whatever). However that could be more trouble then it's worth. (I'm pretty sure both are the case in NZ.)

Nil Einne (talk) 13:51, 23 January 2011 (UTC)[reply]

http://freewarehome.com Click on the sub-menus for systemutilities/systemtools/systeminformation. System Information For Windows and some other software on that site should tell you more about your computer. I would be very careful - I installed a graphics card that the internet told me was compatible, and it burnt out my motherboard and the card. 92.24.182.196 (talk) 21:00, 22 January 2011 (UTC)[reply]

Haven't got much to lose. The current hardware cost me less than 300€ and the new components I'm planning will cost even less. As for the system information software, thanks, but I've already got one for Linux. It's probably not the best available, but will do. 88.112.51.212 (talk) 11:21, 23 January 2011 (UTC)[reply]

I think 92 is suggesting he/she can provide better advice if you provide more info on your system, whatever tool you use. Nil Einne (talk) 13:55, 23 January 2011 (UTC)[reply]

Well, just for closure, I got the components and they work perfectly. Now, if only I could get Civ V work on the good ol' Nuxy... thanks everyone! 212.68.15.66 (talk) 05:59, 24 January 2011 (UTC)[reply]

Caps lock

Almost every day, either here on the ref desks or on the help desk, there is a question posted in all caps. The questioner is usually new to Wikipedia, or has a few other edits (where interestingly they didn't use all caps). It seems an odd choice to me, so why choose to write in all caps? Certainly, when told that it can be interpreted as shouting, there is no follow-up from the OP saying "sorry, I thought that was how questions are supposed to be asked", or "sorry, I got gum stuck on my keyboard", or some such. Astronaut (talk) 17:32, 22 January 2011 (UTC)[reply]

Well, I don't know exactly why individuals do this, but I will note that with typewriters, it was not uncommon to type in all caps, and was virtually unheard of to type in no caps. That's switched a bit with computers. --Mr.98 (talk) 17:51, 22 January 2011 (UTC)[reply]

Previous answers to ~~this question~~ you (see archive) have suggested it comes from the 8-bit era, when home computers typed capital letters by default and would fail to understand commands issued in lower case. (If I remember rightly, that's my previous answer to this question, but somebody else said something similar.) Capitals are also recommended for clarity on official forms: so one reason may be a general nervousness about being misunderstood or ignored. I also postulate that it's a cultural thing, and that there are social networks and forums somewhere out there where everybody types in CAPS all the time. 213.122.5.253 (talk) 18:05, 22 January 2011 (UTC)[reply]

8 bit era? Thay spel asif thay is still in seventh grade! I bet their moms' can't even remember the Monkeys.--Aspro (talk) 18:30, 22 January 2011 (UTC)[reply]

when i first started using computers often at home, i used to do that. there's two main reasons actually. first off, to many people, it looks neater. maybe not when contrasted with normal text, but when you use it exlusively, it can look neater. the other reason is out of laziness, so you dont have to use the shift key to capitalize, and still not have capitlization "mistakes" such a "i" and "england" etc. and then i realized that the missing capitalization mistakes do not annoy people nearly as much as the excessive emphasis/screaming perception when it is contrasted with other entries that are not all in caps. Roberto75780 (talk) 18:52, 22 January 2011 (UTC)[reply]

The equation of capital letters to shouting is not a universally accepted tenet, either. Sometimes, caps-letters do convey extra emphasis; sometimes this formatting conveys no information at all. Sometimes, it conveys metadata. Most often, I think, ALLCAPS is an indication that a human or a machine is designating some text that should be interpreted by both humans and machines. (Machine-consumption-only data is transferred in binary; human-consumption-only data is transferred in nicely formatted sentences). Here are some examples: print editions of many newspapers still use ALLCAPS for the first line (or for the date and location line) of news stories. Print and electronic versions of ALMARS messages are always printed in ALLCAPS, including things that are not being shouted (such as this UNITED STATES MARINE CORPS BIRTHDAY MESSAGE). (INSERT JOKE ABOUT HOW MARINES YELL EVERYTHING). Many NOAA data products and other messages are decoded to ALLCAPS text products like this CHICAGO AREA HRR. Nobody presumes that the HOURLY WEATHER ROUNDUP REPORT is being "yelled."

Using fewer characters eliminates a degree of freedom in text equivalence matching. This simplifies programs that must parse these data formats; so at a cost of "mild" inconvenience to human-readability, it greatly simplifies machine-readability. Even in this era of smart programmatic text-parsers that can ignore letter-case in multiple languages, there is still a great deal of additional complexity when crossing human-language boundaries (not to mention character encodings). Restricting input formats to only 26 ASCII characters, plus a few numerals and punctuation symbols, dramatically simplifies parsing and search. Nimur (talk) 19:00, 22 January 2011 (UTC)[reply]

I always take it as an attempt to state, "FORGET EVERYTHING THAT YOU MAY THINK IS OF ANY IMPORTANCE IN YOUR LIFE AND MAKE MY QUESTION THE MAIN PURPOSE OF YOUR ENTIRE EXISTENCE!!!" So, yes, it is basically shouting. -- kainaw ™ 01:12, 23 January 2011 (UTC)[reply]

I've noticed that a lot of the all-caps posters are from places that use other alphabets. Devanagari, for example, does not have two letter cases, so it's easy to understand posters unfamiliar with English conventions to not see much importance in which style to use. --Sean 14:38, 24 January 2011 (UTC)[reply]

Single-case people might prefer all caps to all lowercase for correctness reasons. We fill out crossword puzzles in all caps, for example, because "usa" or "tuvalu" are considered incorrect, but "BOOGIE" and "TEPEE" are not incorrect. Paul (Stansifer)

I'd always assumed it was old people not familiar with computer use and etiquette. Like someone else mentioned, if you're using a typewriter for an informal note, and you're not a good typist, it's pretty common to leave the caps lock on and write your note in block letters. (With modern fonts, it looks a lot better for informal notes to be all lower, but that's not as true with a typewriter font.) APL (talk) 06:31, 25 January 2011 (UTC)[reply]

Smart phone prices

Why do brand-name smartphones have disproportionately higher prices than similar brand-name media devices that lack phone functionality? A great example is the iPhone. The out right purchase price for the latest, basic capacity iPhone purchased outright with no contract, where available, and officially unlocked, where available, has always been about TRIPPLE the price of a similar iPod Touch. the main difference between the two is phone and 2G/3G data functionality, as well as GPS. sure it may also include a better processor and RAM or other components, but tripple the price? really? Now i know the case may be a little different with the iPhone's prestigue and incredibly high demand, but you can compare offerings from other manufacturers (even comparing models from different brands) and you will find high-end smartphones in the U.S. and canada generally cost $400-$650 to purchase outright, not subsidized by a contract. meanwhile, high-end media players (and in the past, PDAs) that include wifi, bluetooth, similar storage and connectivity, and sometimes even integrated gps, pretty much everything is the same except no cell phone network capability will normally cost about 1/3 to 1/2 the price. sometimes, even the operating system or software is the same (Android, windows, palm etc). especially with android now being a comprehinsive software package that cost nothing to license, the development cost for both these catagories can be significantly lowered, yet the phones remain 2 to 3 times more expensive that very very comperable devices without a phone radio. oh, and you really cant say that the phone radio itself contributes to this cost, because other cheap phones for well under $100 often have the same radio, even 3G data capable. Roberto75780 (talk) 19:08, 22 January 2011 (UTC)[reply]

Because in commodity consumer electronics, the "free market" has determined that technical specs are among the lowest priority factor during purchase decisions. Peruse this Google Book search result list. As a relevant comparison, peruse some commodity personal-computer websites - as of 2011, I had to click through three or four pages on each major-brand's website before I even got to see the CPU type. Consumers want colorful, fashionable, brand-identifiable products. Businesses cater to that market demand, and set prices accordingly. Arguably, this is the most intelligent way that major brand consumer electronics companies can fight back against commodification. Nimur (talk) 19:20, 22 January 2011 (UTC)[reply]

Eh, mentioning Apple products is starting off wrong, as they have consistently been twice the cost for ages now. You can get smart phones for $100, I just got one last week. The more popular the phone, the more units sold, the more money made, the happier they are to lower the cost. ¦ Reisio (talk) 22:19, 22 January 2011 (UTC)[reply]

Prestige, demand, cost to produce, cost to innovate (you're paying for R&D in some cases), cost of advertising... all of these things factor into the price points. --Mr.98 (talk) 23:49, 22 January 2011 (UTC)[reply]

WiMAX and competing 4G phone service

3G phones that work on UMTS (also known as WCMDA) are always capable of "falling back" to 2G or 2.5G GSM service when 3G networks are not available in the area. My question is, do the newest 4G phones "fall back" to GSM as well? can they first "fall back" to 3G signals if available in the area? I'm mainly interested in the WiMAX type, but any competing (truly 4G) standards are also of interest. Roberto75780 (talk) 19:22, 22 January 2011 (UTC)[reply]