Wikipedia:Reference desk/Archives/Computing/2007 December 18

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.

December 18

What specs are needed to run the game crysis with every single thing on max ???

What specs are needed to run crysis with every single thing (antialiasin, textures.......) on max, run the game on ultra high settings, playing with at normal speed?? —Preceding unsigned comment added by 201.79.210.75 (talk) 03:06, 18 December 2007 (UTC)

I would suggest taking a look at the crysis article if you havent already. BonesBrigade 03:58, 18 December 2007 (UTC) —Preceding unsigned comment added by BonesBrigade (talk • contribs)

I hear that CryEngine2 is horrifically bloated and slow, and that even $5000 perfect systems can't run it at max settings --ffroth 04:47, 18 December 2007 (UTC)

I wouldn't call their engine "bloated and slow" - the problem is that this is a game that has AMAZING amounts of high-end graphical content. They designed it (and the engine that drives it) to be something they can use for many years to come - so they pitched it at hardware that doesn't yet exist - and I'd agree that there is probably no hardware that money can buy that would run it smoothly at it's highest quality settings. It's actually pretty amazing what they've managed to do with current PC hardware. So rather than "bloated and slow", I'd say "fast - for what it does - and definitely ahead of it's time" (but "ahead of it's time" isn't necessarily a compliment in this context!). As a games programmer myself, I have to tip my hat to those guys for a job well done. SteveBaker (talk) 14:59, 19 December 2007 (UTC)

I believe System shock was designed in the same way. The CD version shipped with textures so large that no currently machine could possible run them in full-screen mode. No machine available at the time could run the game at 640x480. Today, of course, this limitation is laughable. I can run a couple different copies of it at the same time, without slowing down or seeing the dreaded "salt the fries" message. --Mdwyer (talk) 17:47, 19 December 2007 (UTC)

CON

I am using Windows Vista. Why can't I create a folder named CON? The error message says "The specified device name is invalid." Also, why is is that if I type "=rand (200, 99)" into word I get 500 pages of help files pasted to the document? 70.162.25.53 (talk) 04:54, 18 December 2007 (UTC)

Astonishingly, Windows still has a few filenames reserved for devices, inherited from MS-DOS. These names refer to I/O devices which you can, so the theory goes, access (i.e. read from and write to) just as if they were ordinary files. For example, CON refers to the "console" (keyboard + screen), LPT and/or PRN refers to the printer, etc. —Steve Summit (talk) 05:01, 18 December 2007 (UTC)

But if you know how to use the Command Prompt (if it still exists on Vista), simply prefix the full path with \\?\

example: \\?\C:\con

note: even if you create a file or folder with the same name as device (that includes con prn aux nul com1 com2 com3 com4) you won't be able to use it unless you like to mess with command line (I do).

note 2: If a file is named dev.ext Windows Explorer will see it as dev (here dev = any device, ext = any extension).

--grawity ^{talk / PGP} 17:11, 18 December 2007 (UTC)

The rand function is explained in this Microsoft document as a means of inserting sample text, similar to Lorem ipsum. =rand(200,99) is feeding 2 parameters to the function, which requests 200 paragraphs of 99 sentences each. The sample text is comprised, as you've discovered, of Word help files. --Kateshortforbob 10:34, 18 December 2007 (UTC)

.Mac Web Page

Can anyone please advise how I can add more photos to my existing web page which is http://homepage.mac.com/johnluckie/PhotoAlbum3.html. Many thanks.--88.110.51.242 (talk) 09:42, 18 December 2007 (UTC)

Well, that code is pretty dense, you clearly made it with some sort of page generator or HTML editing program. So the answer is probably to do it something similar to the way you did it before. --24.147.86.187 (talk) 15:14, 18 December 2007 (UTC)

Thanks for this input but I used Mac's iPhoto on System X Tiger which will generate a web page automatically, but so far as I can see does not allow one to add to it once it is in place--88.110.51.242 (talk) 18:10, 18 December 2007 (UTC).

Hmm, yeah, that would be tough. My suggestion would be to make a new set of photos in iPhoto and then generate a second page. You could either add it as a link to the first page or try and merge the pages together. The latter would take some knowledge of HTML and Javascript though and might produce problems if you didn't really know how iPhoto was generating the page. But anyway, the problem is that iPhoto generated a somewhat complicated page. It could be edited, but it would take a lot more knowledge than you have to do so, and would take a lot of time. It is easier to probably just generate another page in iPhoto rather than trying to add to that page. --24.147.86.187 (talk) 20:34, 18 December 2007 (UTC)

I'm sure that you are right, Thanks!--88.110.51.242 (talk) 07:31, 19 December 2007 (UTC)

Hard disk compatibility

I'm looking to buy a parallel ATA hard disk (that will live on the same cable as the an old disk or a CD drive) for a computer about 6 years old. I see that some disks are labeled ATA/100 and the rest lack any such label. I hope you can help me with these questions:

1. What ATA version should I assume disks have when they don't say "ATA/100"?
2. What happens if I get a too modern ATA version? Will it not work with my computer?
3. What happens if I get a too old ATA version? Will it work? Will it be slow?
4. How do I figure out the capabilities of my computer and my current disk? (The computer is currently running Windows XP.)

—Bromskloss (talk) 09:48, 18 December 2007 (UTC)

It doesn't really matter, since all PATA drives nowadays should come with ATA/133, the last of the PATA standard (AFAIK). They're all backward compatible with previous standards so it should be fine. --antilived^{T | C | G} 21:20, 18 December 2007 (UTC)

OK, good, that's what matters most. How about size restrictions? My computer has a 60 GB disk now. Can it handle, say, 300 GB or is there some limit between the two values that might be a problem? —Bromskloss (talk) 22:27, 18 December 2007 (UTC)

There is no hardware limit that I know of (other than what you can buy) but older operating systems may have a limit. XP's (with SP2) limit hasn't been reached yet though. TheGreatZorko (talk) 13:34, 19 December 2007 (UTC)

i'm sure there's a name for this, tell me it

In digital arithmetic (the kind computers do in binary and we do in denary (or whatever, it's a silly word)), not all incrementations are born equal. For example, 163 + 1 -> 164; only one digit changes. However, 169+1 -> 170; two digits change. In the pathological(?) or degenerate(?) case all digits change (e.g. 999,999+1 -> 1,000,000). Is there There must be a term for this case or phenomenon; tell me it, please. Also, ellaborations as to the philosophical, metaphysical, or algorithmic consequences of this inequality of arithmetical additions under digital "representation" of number would be assuredly welcome. 86.146.254.177 (talk) 15:04, 18 December 2007 (UTC)

This has nothing to do with addition. It is solely based on the base you are using. For example, in binary 1+1=10. In decimal, the same values 1+1=2. If "carrying the one" is an issue, just use a larger base. Another way to look at it is with vectors. Each value is the length of a vector from an origin. You are appending one vector to the end of the other and measuring the total distance from the origin to the tip of the second vector. How many digits are used is based on the base you use to measure the distance. -- kainaw™ 15:20, 18 December 2007 (UTC)

I think you are missing the point. It is clear that this phenomenon has nothing to do with numbers on their own. It comes from the connection between the numbers and some system used to represent them. Radix systems are mathematical constructions too. Your suggestion "use a different system and then it won't happen" is irrelevant - it is still a feature of the particular system, one that the OP is interested in exploring.

I don't think there is much to explore, though. The phenomenon of calculations for one decimal place effecting a higher place is called a carry, but I don't know of a special name for the "pathological" case. I also don't think it has any serious philosophical or metaphysical implications, and its only algorithmic consequences have to do with the design of adders. -- Meni Rosenfeld (talk) 15:34, 18 December 2007 (UTC)

You may be interested in reading the articles: Gray code and Hamming distance.—eric 16:13, 18 December 2007 (UTC)

You may also enjoy our Carry look-ahead adder article, where this phenomenon is exploited. Nearly all "serious" computer ALUs note this phenomenon and exploit it for faster operations.

Atlant (talk) 17:42, 18 December 2007 (UTC)

Thanks, Atlant and EricR 86.146.254.177 (talk) 18:29, 18 December 2007 (UTC)

That article is terrible, last year it took me 3 hours to go from "cool, neat idea" to "cool, it works" in a logic simulator. Anyway, you can also precondition the inputs to create a more efficient adder. Add the OR of the inputs with the AND of the inputs (essentially taking the positions where either input is 1 and adding the carries).. this is equivalent to the sum of the inputs. You'll still need carry logic but you'll never have 0+1, only 1+0, including in the carry logic, which lets you do some weird logic optimizations. --ffroth 18:53, 18 December 2007 (UTC)

Oh, and you can also precondition it so that the probability of a carry is only .25 instead of .5 by doing (X XOR Y) + 2(X AND Y), basically a clearer version of the last one in which it's obvious what's going on (places with no carry added to a left-shift-one of the places that have a carry). Not really helpful though in optimizing. And all of this is from that book recommended in the last digital logic discussion Hacker's Delight --ffroth 18:59, 18 December 2007 (UTC)

I might call it a "gang carry" or a "cascade". (But I don't know about anybody else, and those certainly aren't official terms or anything.) —Steve Summit (talk) 16:46, 18 December 2007 (UTC)

When one digit rolls over and the next one increments, that's a "Carry" - I'm not aware of a term for when one carry causes another, which causes another and so on. In the simplest adder circuits, the case when the number rolls over and ALL of the digits change state at once, that represents the slowest path through the circuit - and is typically the most power-hungry and the most cross-talk inducing - so from an electronics engineering perspective, it's the hardest case to manage. Hence there are all manner of designs for adders that attempt to deal with that. Aside from that, it's hard to think of any other deep significances. SteveBaker (talk) 17:44, 18 December 2007 (UTC)

You guys have no imagination... (OP)86.146.254.177 (talk) 18:16, 18 December 2007 (UTC)

You're welcome. -- Meni Rosenfeld (talk) 18:24, 18 December 2007 (UTC)

I'm sorry, I somehow got the impression you wanted an actual, truthful answer. If it's imagination you want, that's an ENTIRELY different matter. Did you know that every time your car odometer rolls over, a pixie dies? If you take out your calculator and add 0.001 to 99.999, Bill Gates will send you $100! It's a little known fact that the ancient Egyptians built their pyramids using the additional acoustic energy from a row of abacusses that were instructed to add 123456 to 987654. If Babbage had ever finished the analytical engine and added 4567 to 6543, the resulting ground shaking would have destroyed the earth! OK...I'm bored now. SteveBaker (talk) 20:32, 18 December 2007 (UTC)

Playstation 3 hardrive upgrade

If i bought this hardrive, Momentus 7200.2 160GB, would it be compatible with my playstation 3? Also would it allow me backward compatability with the playstation 2? --Hadseys (talk • contribs) 19:08, 18 December 2007 (UTC)

That's a 2.5" SATA drive, right? That is what the PS3 uses. What PS3 do you have? -- kainaw™ 19:36, 18 December 2007 (UTC)

• A forty gigabyte plastation 3 --Hadseys (talk • contribs) 20:29, 18 December 2007 (UTC)

Nothing that you can do to the Playstation hardware (unless you are a Sony Engineer I guess?) will give you PS2 and PS1 backwards compatability. That's built into the operating system and hardware. TheGreatZorko (talk) 13:38, 19 December 2007 (UTC)

Oh - I see, OK. This is a very messy situation!

The deal is that the version of the PS3 with the 40Gb hard drive does not come with the "Emotion chip" that enables it to emulate the PS2. That chip (which basically contains a PS2 which in turn contains a PS1!) is in the 20Gb and 60Gb(NTSC) versions of the playstation - but for some reason was not included in the 40Gb, 60Gb(PAL) or 80Gb versions. Changing the hard drive in your PS3 will increase the amount of space you have for downloaded games and other stuff - but it doesn't affect whether you have the chip or not - so you still won't be able to play PS2/PS1 games. There are PS3's (the 60Gb(PAL) and 80Gb versions) that have bigger hard drives and that don't have the "Emotion chip" but can run SOME PS2/PS1 games via software emulation. But for some reason known only to SONY, that software was not included on the 40Gb PS3. I can't imagine why not - except perhaps that SONY wanted a bigger differentiator between the cheaper 40Gb PS3 and the higher end versions. There is a really nice table in our PS3 article that neatly explains what's going on: PS3#Retail_configurations.

So, sadly, no. There is nothing that I'm aware of that will allow your PS3 to play PS2 and PS1 games. Sorry. SteveBaker (talk) 14:50, 19 December 2007 (UTC)

I continually see the PS3 80GB version criticized for only being able to play some PS2 games. I tested all of my PS2 games on it. Every one played without any trouble, lag, or messed up graphics. The only thing it doesn't do (and the emotion chip doesn't do either) is vibrate the controller. In my experience, it is proper to state that it plays most PS2 games since I haven't found any that it doesn't play. -- kainaw™ 15:37, 19 December 2007 (UTC)

I bow to your superior knowledge! I presume the reason the controller doesn't vibrate is that the PS3 controllers didn't have vibration gizmos in them until fairly recently - and I presume support for it didn't exist back when the emulators were being written. I've never tried running a PS2/PS1 game on my playstation. (But then it is about 18" wide by 3' long and 4" deep - and the only game I get to play on it is the one I'm writing!) SteveBaker (talk) 19:06, 19 December 2007 (UTC)

After googling, I found some lists of issues with PS2 emulation. For example, the videos in FFX-2 do not have subtitles on the PS3-80. I wouldn't have noticed since I never turned subtitles on. Some games did have problems that were fixed during upgrades. To upgrade, you need a network connection for your PS3. I have no clue how to get the updates on CD so you can install them without the network. The PS1 games (of which I have none) have a lot more issues. So, it is probably best to google for PS3 compatibility with your favorite PS2 game before considering getting rid of your PS2 console. It wasn't a choice for me. My PS2 just stopped reading DVD disks. It only played CDs. That pushed me to purchase a PS3 and I had to get the 80GB since the 40GB wouldn't play any of the PS2 games. Another issue is those cards. The PS3 doesn't have card slots. I had to get a converter ($12 online) to plug them into the PS3 so I wouldn't lose all my saved games. I'd hate to do that Chocobo run in FFX again. -- kainaw™ 19:16, 19 December 2007 (UTC)

Using a Wireless router as a network card...

I will be moving, and the house I am moving into has an existing wireless network, I have no problem connecting to the network(and actually have both my laptops and a desktop here in the same room with the wireless router all connected wirelessly, but I have now moved my regular desktop to the house here, and I moved it into the rear of the house(into the office, which I cannot figure out why the network doesn't originate there anyway). I was going to run cat5 to it, but I figured out that will be quite a bit of cable, and thresholds to cross, and don't really want to bother figuring the best way to do it, so I was wondering if I can use my existing wireless router(the one from my house) to connect just that one desktop to the existing wireless network?

Basically I'm wanting to use a router as a network card, is this possible? Dureo (talk) 19:14, 18 December 2007 (UTC)

The "client" router has to support wireless bridging/repeating/client mode, which isn't too common. Make sure you disable DHCP on it. Alternatively, just buy a wireless PCI card. --ffroth 19:49, 18 December 2007 (UTC)

Also I hear that this is a ridiculously difficult hack and almost impossible to pull off for most routers. You're going to have an internal network within the internal network, and God help you if you're trying to access another device in the house on the "outside" internal network. You really need enterprise networking equipment for this sort of thing --ffroth 19:52, 18 December 2007 (UTC)

Guess I'll just get a wireless card, too much work :), I kind of dislike the DSL, it seems inconsistant speed wise compared to the comcast I have at home, maybe I can talk my girl into changing to cable ;P. Dureo (talk) 05:40, 19 December 2007 (UTC)

On that note, I now have managed to collect 1 yahoo DSL wireless router/modem(hers), 1 netgear wireless router, 2 linksys wireless routers(one hers, one mine), and 1 linksys wired router... I cannot figure out how I managed this. Dureo (talk) 05:43, 19 December 2007 (UTC)

Unix question: * with spaces in it

How can I do * when I have filenames with spaces in them? Suppose I have the following files:

• donald duck
• mickey mouse

If I write "for i in *; do echo $i; done"; it will print the following:

• donald
• duck
• mickey
• mouse

How can I do it so that it considers the whole filenames instead? JIP | Talk 19:57, 18 December 2007 (UTC)

It may be your shell. I can't test it because I'm using bash and it returns the file names (with spaces) for me as you would expect it to. It does not break the file names up. -- kainaw™ 20:08, 18 December 2007 (UTC)

Both bash and plain-old sh on my machine (OS X) keep the filenames together, for what it's worth. Friday (talk) 20:15, 18 December 2007 (UTC)

So it seems. It works now. Thanks! JIP | Talk 20:21, 18 December 2007 (UTC)

Yeah - it should work OK. The parsing of the command line into tokens happens before * is expanded into more tokens - so this shouldn't be a problem. Which shell are you using - and are you sure this is the EXACT commands you are running? SteveBaker (talk) 20:25, 18 December 2007 (UTC)

Right. JIP, please have a look at the following:

for i in *; do ls $i ; done # ls tries to look for four files
for i in *; do ls "$i" ; done # ls lists two files

--Kjoonlee 20:28, 18 December 2007 (UTC)

Further Unix question: How to get rid of the extension

So I made this script to convert Ogg files to MP3 files:

#!/bin/bash
basename=`echo $1 | awk -F . '{print $(NF-1)}'`
basesongname=`echo $basename | awk 'BEGIN {FIELDWIDTHS="5 256"} {print $2}'`
albumname=`pwd | awk -F / '{print $(NF)}'`
artistname=`pwd | awk -F / '{print $(NF-1)}'`
oggdec "$basename.ogg" -o - | lame -b96 - "$basename.mp3" --tt "$basesongname" --ta "$artistname" --tl "$albumname"

It works if the filename of the Ogg file does not contain dots before the extension. But for example Deee-Lite has songs such as Try me on... I'm very you and E.S.P., which confuse the heck out the second line of the script. How can I make awk print out everything before the last dot? Or is there a ready-made command for it? Of course, I could always code one in C, but I'd rather avoid a homemade solution if a globally available one exists. JIP | Talk 20:39, 18 December 2007 (UTC)

echo foo.bar.baz | sed -e s\/.[a-zA-Z]*$\/\/

--Kjoonlee 21:41, 18 December 2007 (UTC)

basename foo.bar.ogg .ogg

--Kjoonlee 21:43, 18 December 2007 (UTC)

"When faced with a problem, some people say 'Let's use AWK.' Now they have two problems."

basename=${1%.*}

The ${variable%pattern} expansion gives the value of $variable with the minimal trailing pattern match removed. --tcsetattr (talk / contribs) 21:48, 18 December 2007 (UTC)

That stupid lame writes the song, album, and artist names incorrectly. It reads them as UTF-8 but writes them as ISO-8859-1, resulting in such gems as "Nahkatakkinen tyttö" from the album "Kerjäläisten valtakunta" by Dingo. And of course my media player shows the names as lame wrote them. All this makes me think the concept of an enumerated character set should have been invented anywhere else than in the USA. JIP | Talk 20:44, 19 December 2007 (UTC)

Just use something like EasyTag. --antilived^{T | C | G} 04:58, 21 December 2007 (UTC)

What is a "channel business"?

I first came across this term when I saw an IT business related magazine called "Channel Business", the online version of which is at http://www.channelbusiness.com. Flicking through the magazine I saw it repeatedly refer to 'channel businesses' without ever giving any clues as to what one is! here is an example of an article, from another source, using the term. It also talks about "pushing through the channel" leaving me even more confused. The channel article is no help. Any ideas? Tempus123 (talk) 22:07, 18 December 2007 (UTC)

That is because channel in this context is part of distribution (business). A channel business generally does not sell directly to customers, but uses a system of distributors and resellers. --— Gadget850 (Ed) ^talk - 22:15, 18 December 2007 (UTC)

Thanks for your reply. This is making some sense to me now. Tempus123 (talk) 23:05, 18 December 2007 (UTC)