Wikipedia:Reference desk/Archives/Computing/2010 July 15

= July 15 =

Double sided scanning on a Brother MFC-790CW
I have a Brother MFC-790CW and am trying to easily scan some double sided documents (to multipage .tiff or .pdf files would be great) so far I've had little luck. It has a document autofeeder so I definitely want to be able to use that. So far I have Microsoft Office Document Imaging, but when I use that document autofeeder acts strange. I would greatly appreciate any suggestions.  Xor  24  talk to me  00:30, 15 July 2010 (UTC)
 * This sounds more like a question for Brother's support line. I'd make sure you are using the latest drivers downloaded from Brother's website. The scanner itself should have come with scanning software from Brother. The Microsoft product may not know all the advanced features your scanner has. (Updated drivers might make it recognize the ability to double side scan). Then again, it might not. --70.167.58.6 (talk) 15:21, 15 July 2010 (UTC)

New softwares/programs automatically installed to computer
Whenever I turn on my computer, I find out that 3 or 4 softwares (antivirus programs) are installed to my computer. The softwares themselves seem viral because they don't do any virus removal for you but merely advertise their own softwares. I keep removing the programs but they are constantly reinstalled to my program. I scanned my computer and removed all the virus found by using a legit antivirus software, but it doesn't seem to be working. I'm tired of constantly having to remove the programs from my computer. Please help. —Preceding unsigned comment added by 70.68.120.162 (talk) 02:40, 15 July 2010 (UTC)


 * Can you let us know the names of the programs? Sorry, but without that we're shooting in the dark. ZX81  talk  04:28, 15 July 2010 (UTC)
 * It's impossible to list all the names of the programs because new software programs with different names are constantly installed. A couple of them were PC Cleaner and Antivirus.


 * The surest way to get the malware off your computer is to do a clean reinstallation of Windows and all other programs. If you choose to perform this rather time-consuming task ("reinstall") rather than attempt to "clean" the malware off your existing installation, you will need to have the original CDs for Windows, hardware drivers, other software such as Microsoft Office, etc. and must copy all files you want to keep to CDs or an external drive (the computer's hard drive is erased in the process). Some computers may not come with reinstallation discs; you might have to use a program provided on the computer to make a set.
 * Regardless of which approach you take to remove the malware from your computer, be sure to regularly update not only Windows but also web browsers and plugins (particularly Adobe Reader and Flash) to defend against browser exploits (Secunia PSI, free for personal use, can help with this), and be cautious when downloading and installing software (including the software you had previously installed). PleaseStand (talk) 04:37, 15 July 2010 (UTC)


 * I've seen Malwarebytes recommended here; but see the virus FAQ for more recommendations. Comet Tuttle (talk) 05:47, 15 July 2010 (UTC)
 * I think the "virus FAQ" ought to be deleted as it is only about reinstalling your operating system, and while backing-up you are likely to copy the viruses to the backup. It says nothing about other better procedures for getting rid of viruses or other malware. 92.15.9.213 (talk) 08:15, 15 July 2010 (UTC)
 * There was a deletion vote in May and the vote was Keep. The rationale for its existence is mainly that since this question comes up a lot, it was judged useful to keep around one rather comprehensive answer that we can point to, rather than having the answerers repeat themselves a few times a month and miss some items.  I totally agree that the page needs a lot of expansion on other procedures; please, do expand it; it's user-editable like anything else on Wikipedia.  Comet Tuttle (talk) 15:22, 15 July 2010 (UTC)
 * I suggest Malwarebytes, followed with a free PrevX scan. --mboverload @ 05:50, 15 July 2010 (UTC)


 * You could try doing a virus scan in safe mode, so that the malware is defenceless. I assume you have already checked that the malware is not listed in Add Or Remove Programs. Avast! can do a boot-time scan. I would try scans with other legit anti-malware such as Superantispyware or Spybot Search And Destroy, since they do not all cover the same malware. Clean your computer with Ccleaner before doing a scan. 92.15.9.213 (talk) 08:09, 15 July 2010 (UTC)
 * I check Add or Remove Programs often. In fact, that's where I delete the constantly reinstalled antiviral programs.
 * Malware is not "defenceless" in safe mode! An infected copy of Windows is still infected in safe mode.  Under Windows, "Safe mode" just means a bunch of components are disabled, in order to reduce the chance that they are contributing to system instability.  Safe mode doesn't mean you're booting from some bulletproof original copy of the operating system or anything.  Comet Tuttle (talk) 15:22, 15 July 2010 (UTC)
 * Agreed. Once a system is compromised, it is not safe to assume that any of its utilities are intact, including its bootloader, its "safe mode", its software-level antivirus / malware tools, etc.  While many "light-weight" malware infections can be removed using anti-malware software, the most insidious infections always require a full reboot.  And, it's possible that an insidious infection is pretending to have been totally removed.  The only guaranteed solution is to reinstall/restore an uncompromised system with a clean start.  Nimur (talk) 16:17, 15 July 2010 (UTC)
 * Unfortunately I'm not expert at computers at all so I don't know what choice I have. I don't want to get help from a computer repair store either because it's expensive.


 * First of all I would try doing scans with the free and well-thought of antimalwares Avast!, MalwareBytes, Superantispyware, and Spybot. Clean the computer before these scans with Ccleaner.


 * While the latter three are simply on-demand scanners and do not interfere with your existing anti-virus, you would need to remove your existing antivirus program to instal Avast! I think this is worth doing as a) Avast! can do a boot-time scan, search for rootkits etc, and b) your existing antivirus software has not managed to catch the malware. Download the Avast! instalation file while your existing antivirus is in place, save it to eg your desktop. Then turn off your modem if you have one, uninstall your old antivirus, then click on the Avast! instalation file to install it. Turn the modem on again when the computer restarts, and give it a five minutes to sort itself out. You could check to see if any updates are available although it probably will have checked that itself already. I would also download and instal Ccleaner, and run it (I would tick/check everything in the Advanced section also). Then scedule a boot-time scan with Avast! Then try scanning with the other scanners above. Each scan could take several hours, but you can leave them running unattended. If you still have problems then you move on to the next stage, which is submitting a Hijackthis! log file to a forum which specialises in studying them.


 * You could also instal the free Startup Control Panel by Mike Lin, which is an easy way to see what is starting up when the computer starts. 92.24.182.219 (talk) 10:10, 16 July 2010 (UTC)

Keeping or changing an old password
I'd like to reset my router to its factory settings, which of course means my current password for accessing my wireless network will be erased. Once I go through the usual steps of setting up my network, would it be safe to use the same password I had before? Or would I be better off with a new one? 24.189.87.160 (talk) 07:01, 15 July 2010 (UTC)
 * Unless you feel the security of the old password has been compromised, there isn't any need to change it. But you should always change your passwords now and then anyway. I used my neighbors wireless internet for 2 years even though it had a password because they are relatively easy to break. If they'd changed their password even just once every two months I probably wouldn't have bothered spending another week each time cracking the key. 82.43.90.93 (talk) 10:41, 15 July 2010 (UTC)
 * If you live somewhere where you have lots of other people living nearby, such as an apartment block or flats, there is an increased risk of your password being cracked so if you lived somewhere like that I would recommend changing it, otherwise if you feel the security of the old password isn't good enough anymore, you don't really have to bother. Chevy  monte  carlo  - alt 11:46, 15 July 2010 (UTC)

chain loading
Loadlin can chain load linux from DOS or win9x. Is there a program, which would allow chain loading some other OS from linux in similar way? (Or linux cannot be overwritten similarily to DOS). -Yyy (talk) 10:33, 15 July 2010 (UTC)
 * Never tried it with something else than another Linux kernel, but you might want to check out kexec. -- 78.43.71.155 (talk) 18:27, 15 July 2010 (UTC)

Trying to write a Windows batch file
I am trying to write a Windows batch file that will merge two similar folder trees, but I've never done this before and I am not a programmer. Same-name folders are merged, but same-name files are renamed before copying. It seems to be easiest to make a new copy of one of the folder-trees, and then merge the other folder-tree into it. This is the where I've got to so far:

C:\>FOR /R C:\mypath\ %%G IN (*.*) DO CALL :copyfile %%G


 * copyfile

IF NOT EXIST %destfile% copy %srcfile% %destfile% /V ELSE copy %srcfile% %destfile%.bak /V

1) What is the "%%G" for? What does it do? B) How do I get :copyfile to behave like a subroutine? 3) Most importantly, if it finds that there is a sub-folder in the source folder that is not in the destination folder, how do I get it to copy the subfolder (and all its files and subfolders) into the destination folder? Thanks 92.24.191.1 (talk) 12:47, 15 July 2010 (UTC)


 * On the command line, you can type "help for" (or "for /?") to get documentation about the syntax of the for command.  In this case, you are assigning %%G as the temporary result of each element of the directory.  The %% syntax is used in batch scripts, while %G would be suitable for a direct run on the command-line.  This % vs. %% idiosyncrasy is a historic artifact of the way the command interpreter expands variable-names.  As I've mentioned numerous times above, all of these requests can be done in DOS Batch, but the toolkit that it provides is both clunky and unintuitive.
 * DOS Batch does not have subroutines; you can use "GOTO" creatively or put features in separate script files and execute them as needed. In reference to your last question, you will need to construct some complicated logic to wrap your recursive search through the directory tree with a lot of "IF" and "GOTO" statements.  These are limitations of DOS Batch scripting.  Because you've chosen this toolkit, which is generally unsuitable for this task, you're going to end up with something that looks very complicated to do even a simple task.  I really recommend you investigate some alternative tools, like Perl, which have built-in platform-independent file management, directory navigation, and are much better suited to manipulating the text names of your files.  See File Copying, Find all files in a directory tree, and so on.  It will make your life easier, and your code more re-usable.  Nimur (talk) 14:10, 15 July 2010 (UTC)
 * Alternately, since ultimately I think you just want to synchronize directory trees (and not, presumably, to become the world expert on recursive programming with batch scripts), have you looked at the rsync tool? You should be able to get a Windows version, either using a command-line or a GUI, that can synchronize and merge your directory trees.  I'm not sure if this totally satisfies your needs.  You could also look at a full-blown content management system or version control system.  These programs would be at the edge of their intended-use-cases, but they can store and manage various versions of document directories, and assist you in integrating divergent branches back together. Nimur (talk) 14:24, 15 July 2010 (UTC)

FOR /R traverses the directiory tree, I believe. Once I have solved 3) above then I should have a quick solution for my problem, apart from tinkering with 2). Perhaps it will do what I asked about in 3) without needed anything else. 92.29.115.235 (talk) 14:28, 15 July 2010 (UTC)

I do not believe that rsync or any of the other similar tools will do what is described above - none of them will rename a file with a clashing name. I've tried them and searched for a solution for a long time - none found. Thanks 92.29.115.235 (talk) 14:49, 15 July 2010 (UTC)


 * rsync --backup will preserve pre-existing files&mdash;by default by adding a ~ to the end of the filename, but you can use a different suffix (with the --suffix option), or move them to a different directory (with --backup-dir). &mdash;Korath (Talk) 14:56, 15 July 2010 (UTC)

4) Is it possible to find out if a directory in the current source folder does not exist in the current destination folder? Thanks 92.29.115.235 (talk) 14:51, 15 July 2010 (UTC)
 * Well, if you've got the path you need to test for, "IF NOT EXIST %dest% GOTO DESTNOTEXIST" would work in batch scripting. Unfortunately, it won't work quite right, since EXIST can't distinguish between files and directories, but on a Windows system, files usually have extensions and directories don't so you may be safe in practice. &mdash;ShadowRanger (talk 15:22, 15 July 2010 (UTC)

Another helpful tool for copying subdirectories: xcopy /s will recursively copy an entire directory tree. Xcopy is the extended copy tool on Windows; type xcopy /? for documentation of its other features. It may be helpful inside your script at some point, providing features that regular Windows copy does not. Nimur (talk) 16:11, 15 July 2010 (UTC)


 * Here, you can use this. It will try "file.ext", "file (1).ext", "file (2).ext", etc. until it finds one that doesn't exist in the destination. Change the paths near the top to your source and destination directory names. Change "set op=echo" to "set op=" to make the changes live. If you don't change it, the script will just print the copy and mkdir operations without carrying them out. The script will fail if there's a source directory with the same name as a destination file. It's not clear what ought to be done in that case. The script always copies every file from the source folder. It won't test source files for equality with existing destination files. It wouldn't be too hard to add that capability.


 * By the way, this isn't DOS batch. No version of MS-DOS ever had half the features I'm using here. It's still a clunky language, but call it "Windows batch" or something. -- BenRG (talk) 19:52, 15 July 2010 (UTC)


 * Agreed, it's "a batch file for Command Prompt." Nimur (talk) 20:10, 15 July 2010 (UTC)

Thank you very much for the batch file. "The script will fail if there's a source directory with the same name as a destination file." Read that twice: "source directory", "destination file". Would there be any way of checking that before using the script please?

While rsync looks promising, according to the article it is for Unix systems and not Windows, but I also see that the article mentions some Windows versions DeltaCopy QtdSync and RsyncBackup which I will look at. Thanks 92.24.182.219 (talk) 10:15, 16 July 2010 (UTC)

All of DeltaCopy QtdSync and RsyncBackup only work over a network, and do not mention anything about not overwriting files, so appear to be of no use for this problem. 92.29.117.202 (talk) 16:25, 17 July 2010 (UTC)


 * This script will list any directories in the source that have the same name as files in the destination:




 * You might be better off with rsync than these scripts, though—I don't know. The chances are that the versions of rsync you found will work for local copies, though rsync's main raison d'être (on all platforms) is network backups. I would use Cygwin's version of rsync because Cygwin is popular and actively maintained. -- BenRG (talk) 23:03, 17 July 2010 (UTC)
 * Thanks very extremely for the extra code. Already mentioned that rsync is not for Windows a little further up the page. 92.24.178.184 (talk) 12:11, 18 July 2010 (UTC)

I'm not sure if I should run around shouting "Eureka!" but the utility here says it can rename things in its commands: http://syncdir.sourceforge.net/ On the other hand there is an implication that it overwites rather than merging the same directory names. There are seemingly dozens of things with a similar name. Thanks 92.24.178.184 (talk) 12:11, 18 July 2010 (UTC)

Using PCI Express 2.0 graphics card with 1.0 motherboard?
The graphics card to my first gen Mac Pro died. While the new retail options for Mac graphics card are pretty slim, they are all PCI Express 2.0. Wikipedia's entry on PCI Express says they are supposed to be backwards compatible with v1.1 and v1.0 motherboards (restricted bandwidth, of course). Has anyone tried this? --70.167.58.6 (talk) 15:14, 15 July 2010 (UTC)
 * Are you asking specifically about the Mac Pro? Because a quick search will show a very large number of people have used PCI Express 2.0 cards with PCI Express 1.1/1.0 motherboards. I myself have used a 9600GT in my PCI Express 1.1/1.0 (not sure) motherboard and my card is a 8600GT which I think is also PCI-e 2.0. Other then asking specifically about your combo, a perhaps better question is whether you could have problems. I wasn't aware of any but there are some random claims here sadly not really backed up by anything. While OT, interesting enough I'm also reading some more definitive reports of people having problems with PCI-e 1.0/1.1 cards in PCI-e 2.0 slots  Nil Einne (talk) 00:36, 16 July 2010 (UTC)

Computer speeds 2010 vs 1995
I was reading in Wired that each frame in Toy Story 3 took seven hours to render (on average). But 15 years earlier, Toy Story 1 took 5 hours per frame. How would you compare computer speed then to now? Low MHz unicore RISC Sparc CPUs vs today's GHz multi-core designs. 100x faster? --70.167.58.6 (talk) 19:32, 15 July 2010 (UTC)
 * I find it hard to believe that a single frame of a cgi animation would take more than an hour to render, even on my crappy computer. Could you link to the page you read that on? 82.43.90.93 (talk) 19:40, 15 July 2010 (UTC)


 * 1,084 days: How Toy Story 3 was made (04 June 2010), and The Toy Story Story (December 1995). Nimur (talk) 19:45, 15 July 2010 (UTC)


 * At face value, I would think that both articles use the term "render" in different ways. It is probable that in the 2010 edition, "rendering" includes a whole other host of 3D graphic processing, such as fabric physics, hair / fur, particle dynamics, fluid dynamics, character kinematics, and so on; whereas in the original article, "rendering" probably referred specifically to ray-tracing.  From the 2010 article: "FINAL RENDER: Surfaces -- walls, clothing and faces -- are fed through rendering software that simulates light and shadow. It also adds texture to Lotso’s fur, Barbie’s leggings and the carpet. An average frame takes more than seven hours of computing time to render. A more complex frame such as this one required 11 hours.  " Nimur (talk) 19:47, 15 July 2010 (UTC)
 * Another thing not being said is compositing and effects time. Traditionally, you never try to render everything in just one frame. It always needs color correcting, 2D effects compositing, lighting correction, etc --70.167.58.6 (talk) 16:24, 16 July 2010 (UTC)
 * Something isn't being said, clearly. If it took 7 linear hours the movie would take a few decades just to render, but according to the wired article it only took some 100-200 days.  Let's just say it was 200 days, everything worked around the clock, and they only had to render the some 130,000 frames (90 minutes of movie) *once*, which means a frame had to be done every 2.2 seconds in order to meet the timetable.  It may take 7 hours of "system time" which gets divided by each core working on it... but then you get incredibly fuzzy since you would need to know how many cores it was divided to, what kind they were, the interconnects used, etc. before you could compare the Toy Story 3 process to the Toy Story 1 process.  On top of that there's no mention of the frame resolution; it may have been suitable in 1995 to only render something like 1000 or 2000 lines per frame, but it's probably now desirable to render 4000 or more because projection equipment has advanced quite a bit (and they might want to put it on youtube at some point *grin*). Just something to think about; sorry for not offering an answer! --144.191.148.3 (talk) 14:06, 16 July 2010 (UTC)
 * One aspect I think you are missing is that Pixar didn't render the movie on one computer. They've always used massive render farms. So while one frame takes 7 hours, they have 200 computers rendering one frame each. So at the end of 7 hours, they have 200 frames done. I found this interesting tidbit about the Toy Story 1 stats  and  and IMDB has some more . From it, I gather Pixar used 117 dual and quad CPU SPARCstations as a render farm (294 CPUs total).  Comparatively, Finding Nemo (2003) used a 2000 processor render farm . I couldn't find TS3 stats, but I'm sure it's more.

Also this mentions that TS1 was rendered at 1526x922 pixels.

Toy Story aside, how much faster is a modern high end dual Quad-Core work station compared to a 1995 Sun SPARCstation? --70.167.58.6 (talk) 16:24, 16 July 2010 (UTC)


 * Consider reading our articles on Benchmark (computing) and, for example, the SPECint test suite before looking at raw comparisons. A direct comparison of two computers from different eras, with different processor architectures, with different operating systems, with different peripheral hardware, is of dubious usefulness.  Even the names of the architectures reveal a lot - "dual quad core" - so, unless your benchmark/comparison has some meaningful way to compare a single-core computer to a multi-core computer, the raw numeric result is not useful for comparison.  However, you can directly run, for example, the single-CPU benchmark provided by the SPECint numerical benchmark on a 1995 SPARC and a 2010 dual quad-core.  This archived press release, Sun discloses UltraSPARC-II ...  (October 13, 1995), says that "UltraSPARC-II will deliver an estimated 350 to 420 SPEC92int and 550 to 660 SPEC92fp" (UltraSPARC II was released in 1997 and was awesome - but only if you ran Solaris!).  Now, if you want, you can run SPEC92 (from 1992) on modern hardware; but nowadays it is more likely to run SPEC2006: SPEC.org lists results for modern computers, including a 3000 MHz Intel Dual-Core Xeon, 1333 MHz Bus Apple system; they quote 2929 on SPECint2000.  So, "by dividing these values", you can say that the 2006 Dual Quad-Core is about 7 times "faster" than the 1995 UltraSPARC-II.  As I repeatedly caveat: these tests have limitations - most notably, that SPEC2000 is not the same test as SPEC1992, and that both tests are designed for single-cores, and fail account for the fact that one modern dual quad-core PC contains 8 cores and can run 16 threads simultaneously.  Nimur (talk) 19:50, 16 July 2010 (UTC)


 * I should note, though - 7x "direct speedup" is a reasonable estimate: from around 1995, there were about 3 cycles of 18-month periods during which "Moore's law" applied directly to double the processor frequency. That brings us to estimate around 8x speedup between 1995 and year 2000.  Approximately this time, computer architectures began to focus more on multicore and less on clock-frequency; and since 2000, the maximum clock-rates have hovered around the 2.5 to 3.0 GHz mark.  Additional performance boosts have come through bigger CPU caches and multi-core processor systems (not to mention the mainstream implementation of bizarre and wacky things like out of order execution, speculative execution, and Pentium 4's extraordinarily complex and deep instruction pipeline - released Year 2000).  Nimur (talk) 19:57, 16 July 2010 (UTC)

PHP Code
A number of times over the years, while browsing various sites (including Wikipedia) I've been presented with the actual php code of the page I requested. This has in at least one occasion, in this case a small forum, given me the moderators password which was in the php code. After another refresh, the site was back to normal displaying the correct page. My question is, why does this happen, and what can be done to prevent it if one were running their own server with php pages? 82.43.90.93 (talk) 19:38, 15 July 2010 (UTC)


 * This will occur when apache has not been properly configured to process PHP as php, but rather serve it directly as any other file. Apache must be set to send the php code through the interpreter first and then return the interpreter's output --- if not you will get the raw code as a file.  Three things must be in place on the web server:(1) PHP must be installed, (2) The PHP module must be enabled and loaded by Apache, and (3) the .php extension must be routed through the interpeter with a configuration line (usually /////AddType application/x-httpd-php .php////).  While I cannot speak for the small forums, I can say that it is highly unlikely that Wikipedia would ever temporarily experience a disabled php interpreter.  Is it possible that maybe you were seeing something else?  Can you replicate? Also, FYI on an interesting note, any file extension can be set to interpret as php by adding additional lines to the apache configuration (usually in local .htaccess files) .pphhpp, .ryan .tt, anything--rocketrye12talk/contribs 19:54, 15 July 2010 (UTC)


 * Sometimes, certain error messages will dump debug information, including the line of PHP that caused an error. This is not the same as "delivering" the PHP source - it is printing an error that the PHP script encountered.  In the strict sense, the PHP interpreter and its attached web-server module is functioning correctly - it is the script that is failing, and possibly dumping fragments of its own code to the end-user.  Most websites prefer never to let the user see such technical error-messages.  Wikipedia (MediaWiki) will occasionally spew some PHP error-messages, especially if there are problems connecting to the Settings files or the databases.  See MediaWiki PHP Errors for common configuration issues and symptoms.  Nimur (talk) 19:59, 15 July 2010 (UTC)

Thank you for the responses :) @Rocketrye12 I'm certain it was the php code; started with "<?php" etc, contained everything you'd expect to see in php code, including a config section with moderator passwords, board settings etc, and the url of the page was something like http://example.com/index.php?page=4 When it happened on Wikipedia I think I was trying to edit a page, but again it was clearly php code with bits of garbled html. After a refresh it was back to normal. I can't replicate it - in however many years I've been browsing the internet I've only seen this happen around 3 times 82.43.90.93 (talk) 00:17, 16 July 2010 (UTC)


 * well, if you get just the right errors occurring so that the server doesn't recognize or process the page as php code (could be something as random as a momentary fault that reads a single quote as a double quote), the server will toss back unprocessed instead of processed code. these things happen one time in a billion, but there are billions of pages served every day, so these things happen.  -- Ludwigs 2  00:42, 16 July 2010 (UTC)

international laptop warranty: Which manufacturers offer it? Alternatives
Which laptop manufacturers offer international warranties (= even hardware failure can be fixed at no cost in countries other than the country where it's bought)? Alternatively, are there other companies offering the same (offering to fix a laptop, cover the costs of fixing a laptop, etc.) for a reasonable price (<100, mayyybe <150)? Or where else do I find people who know the answer to this question (other than calling every single manufacturer)? Thanks a lot, Thanks for answering (talk) 23:20, 15 July 2010 (UTC)
 * I've had excellent experience with AppleCare. They exchanged the power brick of my old G4 Powerbook (bought in the US, bricked in Scotland) and the motherboard of my Aluminium MacBook Pro (bought in the US, broke in Germany) fast, and without any fuzz, although I never even had the proof of purchase on me. --Stephan Schulz (talk) 14:15, 16 July 2010 (UTC)
 * When I bought my Dell, I asked this specific question. They said the standard warranty was in your home country only.  Astronaut (talk) 10:20, 17 July 2010 (UTC)