Wikipedia:Reference desk/Archives/Computing/2009 May 21

Computing desk
< May 20	<< Apr | May | Jun >>	May 22 >

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.

May 21

Word 2007 - automatically populate document property using document contents

Hello,

I am a teacher & have to create a standard layout lesson plan for my management's occasional perusal.

I'd like the box (cell in a Word 2007 table) which they have called "Keywords" to automatically populate that document's tag (the metadata that is visible in Explorer and is made available to Vista's search via "tag:pythagoras")

Can anyone help? I assume I need to set that box to be a field, but then how to write the file's metadata, and how to capture multiple values into multiple tag values.

pythagoras; geometry ; proof

I'm in Word 2007 and about 5 years ago was a reasonable coder in VBA for Excel 2003 if that's any use at all!

Cheers Rob —Preceding unsigned comment added by 87.127.178.85 (talk) 00:49, 21 May 2009 (UTC)

Computer shuts down during high-load applications even after replacing power supply

Hello,

Recently, my computer started shutting down during high-load applications, such as 3D rendering and when playing video games that ran fine in the past. In these cases, the power supply was quite warm to the touch after the shutdown (the air being expelled by the power supply fan was also quite warm just before the shutdown event). As such, I replaced the power supply with one of a higher wattage (500W continuous for the new vs. 380W peak for the old) and roughly equivalent current allocation (e.g. 22A for +12V in the new power supply vs. 18A for +12V in the old power supply). However, the problem has persisted, even with the new power supply. Is it possible that there's some other problem unrelated to the power supply causing these issues, or is it possible that my new power supply is actually inferior to my old one despite its specifications being higher?

Any suggestions would be appreciated - I'm a tad annoyed at having purchased a new power supply and having the problem persist. —Preceding unsigned comment added by 67.68.50.215 (talk) 01:12, 21 May 2009 (UTC)

Check the temperature of air coming out of the CPU and graphic chipset fans. There may also some system status info you can look at, to check cpu temperature. Maybe those are causing the system to shut down. But yeah, power supply specs (especially on cheaper ones) are often exagerrated, like car stereos. 207.241.239.70 (talk) 03:04, 21 May 2009 (UTC)

And try to look for busted capacitors on your board or some other critical components in your system. Chances are that the reason why your rig gives up after heavy loads is because of a fried and/or damaged motherboard. Blake Gripling (talk) 05:29, 21 May 2009 (UTC)

I think overheating is your next most likely cause. Check the temperature inside the box when you aren't doing heavy CPU/3D work - then again when you are. Possibly you have a dead fan...probably the one on the GPU. It's possible that the power supply may be a contributory factor to overheating - when you run a power supply close to it's limits, it runs a lot hotter than a higher capacity power supply running well below it's limits. But still - I bet it's either a dead fan or a clogged filter or lots of fluff accumulating inside the case someplace. SteveBaker (talk) 13:32, 21 May 2009 (UTC)

Thanks for the suggestions, everyone. I've checked the aspects mentioned above, and:

• all the fans (CPU, power supply, and graphics chipset) are running
  • during routine operations, the fan on the CPU is the loudest
  • during high-load operations, the power supply fan and the CPU fan are both pretty loud, and the air coming from them is quite warm
• I cleaned out some dust from the top of the CPU heatsink and fan assembly
• the capacitors on the motherboard look fine; there's no evidence of leakage or bulging

So, it seems like either the new power supply I got had inflated specifications (and is actually inferior to my previous one, which didn't even have this shutting down issue until a month or so ago), or there's something wrong with my motherboard that's not immediately visible. For the record, my (new) power supply is an Antec EA 500, my graphics card is an ATI Radeon HD 2600 XT, and my motherboard is an ASUS P5K. Any further thoughts on what might be wrong (especially with the motherboard) would be welcome, but either way the assistance so far has been quite helpful. —Preceding unsigned comment added by 67.68.50.15 (talk) 16:59, 22 May 2009 (UTC)

EDIT: I was just going through the Windows event viewer (another shutdown happened today), and I've got a message from the application that monitors the graphics card saying that a "System shutdown due to graphics card overheating" occurred. I guess this mean the problem's due to the graphics card, which seems kind of strange given that I've never had any issues with it in the past. Well, somewhere to start, I suppose. Thanks to all who offered advice. —Preceding unsigned comment added by 67.68.50.15 (talk) 17:11, 22 May 2009 (UTC)

Then I'm definitely sticking with my claim that the GPU fan is the problem...you say that the fan was running - but perhaps not fast enough or something. Do you have any other boards plugged in next to the graphics card? If so, can you relocate them to a slot further away from the graphics card? You could try getting a can of compressed air from your local computer store and blast air into and around the fan to make sure there is no fluff or dust fouling up the bearings. Some of those fans have inlet and outlet holes that can get clogged - again, get in there with the canned air and blast it out. With some graphics cards, you can carefully unscrew the fan and pull it off - that would give you another way to make sure there is nothing fouling it up. Theoretically, you could replace the fan - but I'd be surprised if you could find a replacement. You may have to resort to a new graphics card. :-( SteveBaker (talk) 17:28, 23 May 2009 (UTC)

One way to be fairly sure that it's an overheating issue is to pull the side off of your PC and get a box-fan or a desk fan and point it into the PC and turn it on on full-blast. If your machine runs OK with that HUGE amount of airflow - then you know it's overheating...if it crashes still - then it could be something else. SteveBaker (talk) 17:32, 23 May 2009 (UTC)

Thanks for the tips, Steve. I pulled my graphics card out of the computer and it turns out the heat sink (which is under a plastic cover that can't be removed, something that strikes me as rather poor design) was covered with a thick layer of dust on the side nearest the fan. I was able to remove all this dust using a pair of electrical tweezers (not really their intended purpose, but it worked), and since doing so there haven't been any shutdown issues. I also installed ATI tray tools in case the fan speed wasn't scaling up with temperature increase on the card (which is apparently a big problem for a lot of 3000 series ATI cards), but to be honest I think it was just all the dust that had built up. —Preceding unsigned comment added by 67.68.48.175 (talk) 15:50, 24 May 2009 (UTC)

modern sorting algorithms

Anyone know a good reference for modern sorting algorithms, where Knuth vol 3 doesn't count as "modern"? I.e. I have a few hundred million items to sort on a multicore 64-bit x86 computer with several GB of ram (the internal sorting phase of a much larger external merge/sort). Where can I read up on algorithm selection and implementation, that utilizes the multiple cores, takes into account cache locality and possible NUMA memory setups, etc.? The current Wikipedia articles on sorting don't have much discussion of this stuff. Thanks. 207.241.239.70 (talk) 02:50, 21 May 2009 (UTC)

See Sorting algorithm and associated articles (if you haven't already). If this is a one time sort, I wouldn't spend too much time on finding the absolute best algorithms. I'd suggest you partition the data into manageable chunks small enough for in-memory sorts, and then merge the intermediate results into a final sorted list. A few hundred million records might take hours or maybe days, depending on the size of the key and data. I'm sure there are commercial and free software packages to do this. You might load the data into a database and sort it using a query. Any database worth its salt should use a reasonably efficient algorithm – possibly one that runs multiple threads and dynamically optimizes itself based on available memory. -- Tcncv (talk) 03:44, 21 May 2009 (UTC)

Thanks. I did look at that article but it mostly discussed older, single-processor algorithms. E.g. it didn't mention Burstsort (maybe I will add a mention). I have 100's of billions (maybe trillions) of records and will be doing this operation a lot, and anyway I'm interested in general in how the "big boys" approach sorting now that everything is going multicore. 100's of millions is just for the internal phase. The external phase will be controlled by something like Hadoop, probably. 207.241.239.70 (talk) 04:00, 21 May 2009 (UTC)

You mentioned Hadoop — did you see this announcement from 9 days ago where Yahoo says they broke records by using Hadoop to sort a petabyte in 16.25 hours? They appear to have used a very large cluster, and each machine has two quad-core Xeons. Sortbenchmark.org is probably going to be useful to you as well. Tempshill (talk) 04:54, 21 May 2009 (UTC)

(ec)Well, if you have that much data, I think you will find that you will be mostly I/O bound, meaning your biggest limitation will be how fast data can be transferred between the processor memory and external storage (disks). I suspect that all of the "big boys" will take the partition-sort-merge approach. The key is calculating the maximum amount of data that can be efficiently sorted in memory in one chunk. That will define your partition size. The overall process will be to partition the data, load and sort each partition, and write it back out to storage. Once all the partitions are sorted, the merge phase begins, There is even an art to the merge process. For maximum efficiency, you should write the sorted partition data out to as many different devices (physical disk drives) as are available, and then structure the merge process so that it reads a maximum of one partition per drive at any one time, which allows the drives to operate at maximum efficiency. This will likely require a multi-pass merge operation. (If you recall the 60's and '70's Hollywood view of computes as a bank of spinning tape drives, that was not too far from the truth. Sorting large amounts of data was typically done this way.)

For more information, a google search of "sorting large data sets" seems to yield promising results. My opinion is that your best bet will be to select a commercial sorting package. -- Tcncv (talk) 05:15, 21 May 2009 (UTC)

After looking at Tempshill's references, it appears I may have been off by a couple orders of magnitude on my initial guestimate. I guess I'm showing my years. I just ran a test on my two-year-old laptop and had SQL Server sort 100,000,000 random numbers in about 20 minutes. -- Tcncv (talk) 05:55, 21 May 2009 (UTC)

Tempshill: Yes I saw the slashdot story about the sorting contest. My dataset is large by the standards of typical desktop PC users but very small compared to "big boys" like Yahoo (they used a 3800 node cluster to sort that collection; I may be able to use around 5 nodes for my stuff). Tcncv, yes, I'm familiar with external sorting, I'm primarily wondering how to do the internal phase. Commercial software isn't an option. Re using multiple disks, hmm, I've never been able to keep more than one SATA channel full at a time (on linux), but I guess there are some tricks I can try. Realistically if I can sort a TB of data in 12 hours then I'm happy. I think Yahoo did it in under a minute. Thanks both of you for the url's, I'll see what they say about s/w and algorithms. 207.241.239.70 (talk) 07:32, 21 May 2009 (UTC)

iTunes movie archiving?

Hello, I'd like to free up some hard disk space by archiving some iTunes movies to another drive. Can I do that without losing one of the authorizations that iTunes' DRM allows? Thank you. LovesMacs (talk) 06:06, 21 May 2009 (UTC)

The authorization only applies to this computer, and if you only move the movies to a different disk iTunes doesn't care - as long as it's the same PC. --grawity 11:27, 21 May 2009 (UTC)

Cannot access google

Intel Core 2 Duo, Ubuntu 8.1.0, Firefox 3.0.10. When I try to access www.google.com, Firefox will stay at "Connecting to www.google.com...". I can ping Google fine. Konqueror works fine. IE/Firefox on Windows XP works fine. The weird thing is that www.l.google.com works fine in Firefox+Ubuntu. How do I fix this? F (talk) 07:18, 21 May 2009 (UTC)

Check your proxy setting, and possibly a firewall block. Graeme Bartlett (talk) 09:24, 21 May 2009 (UTC)

Make sure you've not got some website blocking software installed that might have accenditly been set to block google.com —Preceding unsigned comment added by 82.44.54.169 (talk) 11:30, 21 May 2009 (UTC)

Nope, if it is proxy/firefall, Konqueror wouldn't work as well. I don't have website blocking software. F (talk) 02:39, 22 May 2009 (UTC)

• Double check
  • Make sure Konqueror is actually accessing Google and not just getting it from its cache
  • Make sure Firefox is not in offline mode
  • Check your Firefox proxy settings
  • Check /etc/hosts
  • Try to telnet into www.google.com 80
• Install the Web Developer extension
  • Try disabling caching
  • Try disabling JavaScript
• Try disabling extensions or create a clean profile with no extensions
• Try HttpFox for more detailed diagnostics

chocolateboy (talk) 09:10, 28 May 2009 (UTC)

XP default file sort order

In Windows explorer, files sorted by name are listed in some "alphbetical" order I would like to find a character that I can prepend to a file name to cause it to get listed AFTER the alphbetical files. So we might have "@first.txt, alpha.txt, zeta.txt ?last.txt" where ? is what I'm looking for. I've tried (!£$%^&*_+=@'~#;`¬- and am running out of ideas. Does such a character exist? Is there a list anywhere of XP's sort order. (Please do not mention the XP numeric sorting thing - I understand that and it isn't relevant to this question). Thanks -- SGBailey (talk) 08:55, 21 May 2009 (UTC)

I don't know what algorithm they use but it might be this. Empirical testing indicates that nothing in Windows-1252 sorts after Z. If you widen your search to the whole of Unicode then you have lots of choices because all non-Latin alphabets and syllabaries sort after Latin. For example, Α (Greek capital letter alpha) sorts after Z if you want to be maximally confusing, as does circled katakana tu if you're in a whimsical mood. (Font support not guaranteed.) -- BenRG (talk) 11:25, 21 May 2009 (UTC)

I think that in Windows 2000 and older, the sorting was purely ASCIIbetical - based on the ASCII codes of the characters. --grawity 11:26, 21 May 2009 (UTC)

æ, ø or å would work with a Norwegian locale, but I do not know how that sorts if your location is set to an English-speaking place. Jørgen (talk) 23:39, 21 May 2009 (UTC)

If it's ASCIIbetical - then you want the highest numbered ASCII code - which for English/US-English is '~'. I have no clue whether that would work though - I'm more of a Linux guy. Failing that, '{', '}' and '|' are also higher than 'z' in the ASCII code sequence. If you are using some internationalised setup then you may have multibyte characters and all sorts of other horrors - and all bets are off! SteveBaker (talk) 01:01, 22 May 2009 (UTC)

I thought of the curly brackets when the question was posted, but no, they sort before "A". Characters that are not in the standard ascii range, such as "æ" "ø" "å", cause all sorts of trouble if you dual boot linux and windows, or try to use a portable disk that was written on a Spanish pc on a Norwegian PC, etc. So I've gotten into the habit of writing these as "ae", "oe", "aa" in filenames. I think the safest solution is to prepend "zzz_". --NorwegianBlue ^talk 13:35, 22 May 2009 (UTC)

Aren't file names always written in unicode in NTFS? Or where you not referring to a Windows Spanish PC and a Windows Norwegian PC? I've never had great problems with international characters (primarily CJK) in filenames on Windows (English version and locale) tranferring between PCs although it's not uncommon some programs will have display errors or even access errors if their support for unicode is poor (but this doesn't depend on locale) Nil Einne (talk) 19:30, 29 May 2009 (UTC)

Open Character Map (which should be under the Accessories menu, if it's installed), select the font "Trebuchet MS", scroll down to the bottom, you'll find several characters that work well, and look reasonable, including bullets. --Polysylabic Pseudonym (talk) 08:37, 27 May 2009 (UTC)

What is an API in simple terms?

I do not really understand the idea of an API. Could someone explain it in simple terms please? For example, a graphics API: many language implementations already have graphics commands built in to the language. So why do you need a graphics API? Or is the graphics API there but hidden, so that the ordinary user writing code is not aware of using it, even though its there? Or does an API only mean something that requires the user to write special code within their programs to access it? 78.146.67.27 (talk) 10:19, 21 May 2009 (UTC)

Hey. The problem with APIs is that the whole idea is kind of abstract. As the name suggests, an API - Application Programming Interface - is an interface (duh) that allows communication between stuff. Unlike, say, a GUI (Graphical User Interface), this communication is not between a user and a program - it's between a program and another program.

Here's an analogy. Imagine you are a program. You sit in your office, allocated to you by the manager (the operating system). Your whole point of existence is to perform calculations for customers (users). Of course, customers want you to tell them what your results are, but they want you to do this on the big screen on the side of the office building (the monitor). The problem is that you have no idea how to do that, and the control room (graphics card) is on a different floor that you don't have access to. So what do you do? You call up the switchboard, tell them you want the graphics people to do something, and they'll pass the message along.

In that complicated (and a bit rubbish) example, the switchboard would be an API. It allows one program to call functions and pass messages to a different program, even if they don;t have direct access to each other. Consider the Wikipedia web API ([1]). Programs don't have access to the Wikipedia database, but using the API, you can make calls, retrieve info, and even edit, just by sending requests to the API. You aren't directly calling a function as such, but it still happens.

On to graphics APIs. Put it this way: do you know how to make a graphics card do something? Not using a built in graphics function. Directly. Hm? Probably not. Programs do not directly interface with a graphics card - instead, they send a request to the graphics driver or operating system (using a graphics API) and the driver/OS does it. In-built graphics functions in high-level programming languages are simply pre-packaged and ready-to-use API calls, so you don't need to worry about opening requests to the driver, handling feedback, managing the graphics memory etc. so much. Lower level programming languages, like C, don't have this - so instead, you get a library of functions which does exactly the same thing.

DirectX and OpenGL are graphics APIs. High level programming languages hide their calls to DirectX/OpenGL under fancy function names. Low level programming languages have libraries that hide the calls. Of course, you can make calls directly if you want to , and know how - but there's no point.

Hope that massive essay clears it up - I got a little carried away :-). Ale_Jrb^talk 11:21, 21 May 2009 (UTC)

Thanks. So in summary if I'm using a recent variant of basic, for example, I can include one of the language's graphics commands in the program code, such as drawline(x1,y1,x2,y2), and it interfaces with the graphics API without me having to do anything more. But if I am using something like C which lacks graphics commands, it is much more complicated and I have to explicitly communicate with the graphics API. Is that correct? 78.146.67.27 (talk) 11:34, 21 May 2009 (UTC)

Essentially, yes. However, there are millions of libraries for graphics commands for C and other low-level languages, so if you get one of those, it will be little different to using in-built commands in BASIC. Ale_Jrb^talk 11:38, 21 May 2009 (UTC)

API = Application Programmer's Interface. It simply means that one group of programmers have written some code in which other groups of programmers will use. API is a reference to how these other groups of programmers will call the code. A Quest For Knowledge (talk) 11:59, 21 May 2009 (UTC)

There must be some language with "built in graphics" that you're thinking of - it might help us to help you if you told us what that was.

An "API" is just a term for the interface between the application and some kind of support library (in your case, it's a graphics library - but it could equally well be an MP3 player library or a library that loads JPEG images for you). Generally, the API consists of some constants and some functions - and perhaps some classes and such. The API is defined by some header file ("GL/gl.h" in the case of OpenGL, "d3d9.h" for Direct3D) - that lists what these functions, constants, classes, etc are...and the actual code that is run is in the library file - which (in Windows) is probably a ".DLL" file and in Linux/UNIX is a ".so" file. When you compile your program, the compiler looks into the header file to figure out what functions you're allowed to call - and when the program is 'linked' to the libraries, your program is simply calling functions inside the library file.

Some languages may have graphics commands built into the language itself - but these days that's largely frowned upon because it limits the language to use just that one way to do graphics (or sound or whatever). Most modern languages don't do that - they merely allow you to link your program to some external library and to call it via the API. So, in C or C++, you can link to the OpenGL library and call functions like glEnable(GL_ALPHA_BLEND); - or you can link to the Direct3D library and call SetRenderState(D3DRS_ALPHABLENDENABLE,true); Neither of those functions are built into C++, the compiler has no idea that they have anything to do with graphics. It just knows that you told it to call some function or other that's in a DLL/so file someplace.

That's different from some of the more old-fashioned languages such as the BASIC that came pre-loaded in your Amiga/Atari/Sinclair that had actual graphical commands that were a part of the language itself - and which gave you no choice. Removing those kinds of 'hard wired' commands from the language specification and making them library functions was a major step forward because it removed limitations of what the language designer thought was good for you. Neither OpenGL nor Direct3D existed when C and C++ were invented - so the designers of those languages had no concept of drawing polygons or applying textures. When C was invented, 3D graphics didn't exist at all - nobody had ever tried doing that!

If there are any languages that still have that stuff built into the syntax of the language itself, the compiler (or interpreter) is undoubtedly calling some graphics API or other 'under the hood' to do the actual work...but because that's a fixed choice, you have no flexibility to choose how you want your graphics done.

SteveBaker (talk) 13:25, 21 May 2009 (UTC)

Fantastic explanation Steve - but I think your Computer Graphics history is a bit wrong ("When C was invented, 3D graphics didn't exist at all - nobody had ever tried doing that!"! When C was invented, in the 1970s or so, guys like Paul de Casteljau at Citroen and Pierre Bézier at Renault and General Motors had been doing 3D computer graphics to model sheet-metal for automobiles for almost 20 years! You might be interested in The Origin of Computer Graphics within General Motors, (IEEE Annals of the History of Computing. Vol 16, No 3, 1994). Not only do some of these 3D graphics predate the C programming language - the earliest work predated the computer monitor by a few years! Take a look at DAC-1 - there's a nice History section in that article. Nimur (talk) 14:44, 21 May 2009 (UTC)

Well - just barely. The car designers couldn't interact with their designs in realtime - they mostly had plotters and techtronix storage tubes, and they really only had wireframe. It certainly wasn't 3D graphics as we'd recognise it today. SteveBaker (talk) 00:57, 22 May 2009 (UTC)

Don't be so hard on them! This video from the early 1960s shows some impressive graphics in real-time! I'm certainly not going to call it modern, and it clearly had some limitations, but it is recognizable 3D graphics! Nimur (talk) 02:07, 22 May 2009 (UTC)

The Whirlwind (computer) had real time 2D graphics in the 50s. Seemingly it had its origin near the end of WWII in an idea to build a flight simulator using a digital computer rather than an analogue one. Dmcq (talk) 09:24, 22 May 2009 (UTC)

The advantage of a graphics API in its simplest form, is to separate your program from the specific display device differences. So, for example, you don't have to work out how to draw a circle, you just call circle(center_x, center_y, radius) and a circle is drawn. For each physical display device, the internal workings of the circle function may be very different to support a normal screen, a pen plotter, or a character cell terminal; but it is always accessed the same way within your program.

Another example: I want a shaded 3D surface on the screen. I don't need to work out what colour to draw each pixel, the graphics API lets me specify all the details about the surface (its shape, its position, its colour, its reflectivity, etc.) and it works out which colour each pixel should be. Astronaut (talk) 15:25, 22 May 2009 (UTC)

Too many icons and shortcuts on desktop to see

I use Windows XP, in classic mode. Currently I have so many icons and in particular shortcuts on my desktop, that the desktop screen is over-full and they have disapeared off the right of my screen. Is there any way to see these icons, by for example getting the desktop screen to scroll left? 78.146.67.27 (talk) 11:48, 21 May 2009 (UTC)

Change to a higher resolution. Right-click on the desktop, select display properties, on the right tab you will see a slider. See if you can slide it further to the right. The higher the resolution, the more pixels you have on the screen, and the more icons you can show. They just get much smaller. -- kainaw™ 11:59, 21 May 2009 (UTC)

Well, if you use a LCD (as most do), there is only one appropriate resolution, so this will not help. You could try to install ant third-party software that lets you use a number of (often four) virtual desktops. --Andreas Rejbrand (talk) 12:32, 21 May 2009 (UTC)

(simultaneous edit:)

Depending on your display hardware, you may not be able to get a higher resolution. However, you should be able to get the desktop up as a regular folder: click Start, then "Run...", type "shell:Desktop", and click OK.

My recommendation would then be to delete some of the shortcuts (if you can't see them all, they're not exactly easy access, are they?) and re-arrange the remainder onto the screen - right click the desktop and look in the "Arrange Icons" sub-menu.

Also, I've never used it, but Windows XP has a "desktop cleanup wizard" - again, right-click the desktop, and it's under "Arrange Icons". - IMSoP (talk) 12:40, 21 May 2009 (UTC)

I use small icons on my 2K desktop by using this free ware. Try it! Oda Mari (talk) 14:26, 21 May 2009 (UTC)

Double-click "My Computer", then "Documents and Settings", then your username, then "Desktop". The window that opens will have all the contents of your desktop, but in a normal Windows XP window that you can view by icon (and hence scrollable), or view by List, or Details, or whatever - perhaps that will be useful. Tempshill (talk) 14:31, 21 May 2009 (UTC)

You don't even need third-party software for the multiple desktop management. Microsoft released a PowerToy for Windows XP, Virtual Desktop Manager, available here. Nimur (talk) 14:33, 21 May 2009 (UTC)

Just to put my 2 cents in; Just clear off some desktop icons, right-click on the desktop, highlight 'arrange' and click on 'auto arrange' I hope this helps. – Elliott(Talk|Cont)  16:35, 21 May 2009 (UTC)

now what i do is i have one folder on my desktop called "stuff" and everything goes in there. desktop stays clean and free and all my icons and other things are just a click away —Preceding unsigned comment added by 82.44.54.169 (talk) 18:37, 21 May 2009 (UTC)

Desktop is not a good place for icons, because you'd always have something open covering them up. A few things you need frequent and fast access to, eg email client or Office, you can put on the taskbar by dragging their shortcuts to it. One "stuff" folder or the Desktop folder is not very efficient. Make desktop folders to sort your icons by type, eg Office, Graphics, Security, Downloads etc. Drag shortcuts to these onto the taskbar as well.The three most important items will show with their own icons. The folders will show on a small pop-up menu,which you can sort by dragging the items around.These will open on TOP of whatever else you have open. The desktop itself need never show at all.KoolerStill (talk) 20:08, 21 May 2009 (UTC)

You could get a bigger screen. I've now got a WUXGA screen with 1920x1200 pixels and I haven't even filled a third of the screen with icons but I'm sure I'll soon be lusting after a WQUXGA screen with 3840×2400 pixels! The laptop will have to be a pretty big for me to see them all though :) It's a no brainer and you should always get the hardware to do the work rather than doing any thinking yourself (umm perhaps that is a bit over the top!) Dmcq (talk) 09:42, 22 May 2009 (UTC)

Thanks. After reviewing all the suggested options, I have put the shortcuts in a 'Stuff' folder as 82.44.... suggested. A simple and effective solution. 78.146.162.232 (talk) 13:41, 22 May 2009 (UTC)

PAL / NTSC

Hi

Not exactly a computer question but I know someone here should be able to help me out. I dug up and old ps-one (the rounded-edge smaller version) which happens to be PAL, and our current tv is NTSC. Whats the best way to get this setup working? (all my games show in black and white).

TIA PrinzPH (talk) 18:17, 21 May 2009 (UTC)

You need a PAL to NTSC converter. You won't get it working just by fiddling with the wires. Some are good, some are crappy. So, you'll get what you pay for. -- kainaw™ 18:27, 21 May 2009 (UTC)

Hm. Unfortunately, all PS1 games were, I believe, region-locked, so you can't simply buy an NTSC PS1 and play all your games on it. I don't believe it'll work, either, if you take a region-locked-for-Europe PS1 game and try to play it on an NTSC PS2 or PS3. You can do as Kainaw suggests, though that is expensive. Some televisions will take either PAL or NTSC video input. If you Google "multi system tv" then a bunch of them will pop up. If you have a fancy big screen LCD HDTV of some sort, it may take PAL input, too. Sorry, it looks like you're going to have to buy a new TV. Tempshill (talk) 22:25, 21 May 2009 (UTC)

I don't excactly know, but perhaps a TV tuner card with video-in would work? If it supports both PAL and NTSC you could then use the TV-out from your PC to connect to the TV again (if you think your computer monitor is too small). Jørgen (talk) 23:32, 21 May 2009 (UTC)

(this would be one example of such a card ($20 at Newegg).) Jørgen (talk) 23:34, 21 May 2009 (UTC)

chmod and App Runner "security risk"

Can anyone tell me if it is dangerous to do "chmod a+x" to a file instead of "chmod +x"? I know "a" stands for "all users". the problem is that someone said this was a security risk with App Runner: [2] Is this really a security problem? I think not, but I just want to be sure... This would just allow any user to run the file, right? where is the security problem in there? (the root password would still be required to run this as root) __ Hacktolive (talk) 23:09, 21 May 2009 (UTC)

The problem is not so much with '+x' but '+s'. That flag is the 'setuid' flag that causes the application to be run with the permission set of the file's owner. If the file is owned by 'root' then adding the 's' permission allows ordinary users to run this program as if they were 'root'. This is the cause of many exciting loopholes for the bad guys out there. Adding just 'x' only allows the user to run the program with his/her existing privilages - the program can't do anything that the user couldn't do some other way...so that's usually benign. However, for the security minded - turning off as many permissions as possible is clearly the smart move. SteveBaker (talk) 00:49, 22 May 2009 (UTC)

But... I don't use the "+s" flag... I was refering to the "+a" flag... does the "+a" has the problem you just mentioned? (Thanks!) Hacktolive (talk) 02:21, 22 May 2009 (UTC)

I don't think it will make problems. Any user could easily copy the file to their home directory and make it executable, so "+a" flag is not going to prevent anything.

Baeksu (talk) 08:13, 22 May 2009 (UTC)