Wikipedia:Reference desk/Archives/Computing/2010 February 26

Computing desk
< February 25 << Jan | February | Mar >> February 27 >
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.


February 26 edit

Clearing the history of Internet explorer without starting the program edit

I'm trying to help my father-in-law with getting internet explorer working. Symptoms: the program starts, tries to open ~10 tabs and crashes. Chrome works fine. I suspect the problem is that something on one of the pages that IE tries to load causes the crash. So it's a catch-22 situation. To get IE working, i need to clear the history and cache. To clear the history and cache, I need to start IE. How do I get around this? Thanks, 85.200.133.168 (talk) 08:52, 26 February 2010 (UTC) (NorwegianBlue, not logged in)[reply]

You could try http://www.nirsoft.net/web_browser_tools.html - these tools will delete a variety of caches - this might fix it.87.102.67.84 (talk) 10:09, 26 February 2010 (UTC)[reply]
Because it's opening the last browsing settings you might want to turn that off (temporarily) http://www.mydigitallife.info/2009/07/21/how-to-disable-reopen-last-browsing-session-in-ie8-internet-explorer-8/ (last resort)
The location of the files you want to delete are alledgedly here (I haven't checked for correctness) If this works I think it will be your first choice. http://www.mydigitallife.info/2009/07/21/ie8-last-browsing-session-history-storage-location-for-backup-restore-deletion-or-cleanup/ 87.102.67.84 (talk) 10:16, 26 February 2010 (UTC)[reply]
Thanks. Meanwhile, I found out about the iexplore -extoff option (starting with extensions disabled), and it turns out than internet explorer starts when extensions are disabled. There was nothing that stuck out as being the probable culprit, so I'll disable all, and reactivate them one by one. --85.200.133.168 (talk) 10:35, 26 February 2010 (UTC)[reply]
...and after re-activating everything, everything works (in the process, there was both an avg update and a vista update). So, whatever caused the problem, appears to be gone now. --85.200.133.168 (talk) 11:20, 26 February 2010 (UTC)[reply]
For the record, you can get to "Internet Options" via the Control Panel, which then gives you the ability to delete history, cookies, etc. --LarryMac | Talk 13:19, 26 February 2010 (UTC)[reply]
Install and run Ccleaner. Will do eaxactly what you want very easily. 78.146.242.196 (talk) 13:57, 26 February 2010 (UTC)[reply]
Simplest solution: Pull the cable or disable WiFi, start the browser, wait for timeouts, delete the tabs and/or history. --Stephan Schulz (talk) 18:06, 26 February 2010 (UTC)[reply]
Thanks, everyone! --85.200.133.168 (talk) 20:29, 26 February 2010 (UTC)[reply]

Is there anything I need to take into consideration when choosing an MPI? I have set up a Beowulf cluster and I will be doing some work on the 2D Ising model. Is there benefits to different types of MPI like LAMMPI or OpenMPI, thanks. Mo ainm (talk) 13:23, 26 February 2010 (UTC)[reply]

Angles in computers trivia edit

For storing and computation on angles (0 to 2pi) on a computer I've found it useful to map the range of values onto 0 to 1 ie (000000000000000 to 1111111111111 etc) (with 1 not included since 2pi=0) . ie using a single register. (I'm using a combination of lookup for the most significant n bits plus euler's formula for the rest of the 32-n bits dependent on accuracy required and formula for sin(a+b) cos(a+b) as well etc to join the two results) . I imagine the same mapping could be used for other trig. algs. such as CORDIC.

Enough about me, my question is: is this mapping of 0 to 360 onto 0 to 1 commonly used in computers, if at all; if so do we have an article on it.??87.102.67.84 (talk) 13:47, 26 February 2010 (UTC)[reply]

This article about the source code for Doom says it used fixed-precision integer math (16.16 bits), a scheme called Binary Angle Measurement. This excerpt has more details about BAMs. This article at MSDN also discusses an integer angle where 0xFFFF+1 is a full rotation, which seems to be the scheme you're talking about. These integer schemes made sense on the 80486 microprocessors for which Doom was targetted; later iD games use floating point (mostly). -- Finlay McWalterTalk 14:39, 26 February 2010 (UTC)[reply]
Thanks (again :) ) , I almost started writing a new article based on the links you gave me, luckily found in time that there was a short mention at Binary scaling#Binary angles so I expanded that and created redirects from places I would have expected to find an answer.
I'm familiar with software components that use milliarcseconds in their internal representation of angles. A circle = 1,296,000,000 milliseconds, so angles of 0 to +/- 2pi can be neatly represented in a 32-bit int. decltype (talk) 21:18, 27 February 2010 (UTC)[reply]

Follow up "brads" edit

Brads and Angle#Units mention "brads" - (google isn't helping) - one section states that 1brad=1/256 degree - but I can't find any examples of use. Anyone got a good reference. Thanks.Shortfatlad (talk) 16:25, 26 February 2010 (UTC)[reply]

Also List_of_unusual_units_of_measurement#Furman mentions "furman" being 1/65536 of a circle - anyone got a ref for that too.Shortfatlad (talk) 16:32, 26 February 2010 (UTC)[reply]
Searching for "binary radians" (the expansion of brad) finds quite a lot, including this. -- Finlay McWalterTalk 16:37, 26 February 2010 (UTC)[reply]
Thanks again, (removed cite tag) I think google must be censoring my search results - I did look..87.102.67.84 (talk) 16:50, 26 February 2010 (UTC)[reply]

Pocket device to watch old films on? edit

I like watching old b&w films, and many of them are out of copyright and available to download. I have some long train journeys coming up. What device would be the best or cheapest way of watching them? A laptop would be too big. I'm unfamiliar with recent gadgets. Thanks. 78.146.242.196 (talk) 15:09, 26 February 2010 (UTC)[reply]

I recently got a 3rd generation iPod Touch. Its screen is large enough to see the picture pretty clearly, but the device is small enough to go into a pocket or (as in my case) to be worn on an armband (when used as a music player while exercising). I use HandBrake to reencode movies to the correct size and format. There are a number of players, from various manufacturers, that compete directly with the iPod Touch (and are likely a bit cheaper). -- Finlay McWalterTalk 15:30, 26 February 2010 (UTC)[reply]
If a iPod is too small a netbook might be big enough - they are typically 1/2 the size of a laptop - ie half size as A5 (paper) is to A4. Also bigger than a iPod and (and cheaper than either) is a portable DVD player - nowadays these are relatively cheap with a good viewing size - all you need is to convert and burn the film to a blank recordable DVD.87.102.67.84 (talk) 15:41, 26 February 2010 (UTC)[reply]
Oh and I think the recent (and still cheap) portable DVD players will take SD cards - which are a lot more compact than DVDs - they'll probably play other media formats too (eg wmv)- but you'd have to check the instructions.Shortfatlad (talk) 16:38, 26 February 2010 (UTC)[reply]
If you want to go down the portable DVD route one thing to watch out for is the screen resolution which in some cases can be quite poor - best to see one first, or check the specifications before buying.87.102.67.84 (talk) 18:01, 26 February 2010 (UTC)[reply]
An older and possibly cheaper solution might be a PSP, which uses Memory Stick Duo cards and can play .mp4 movies. Comet Tuttle (talk) 17:25, 26 February 2010 (UTC)[reply]
Yeah, and the PSP would be a good device to use for films because of it's widescreen. Chevymontecarlo. 08:49, 27 February 2010 (UTC)[reply]

Perhaps a netbook would be small enough? —Preceding unsigned comment added by 82.44.55.204 (talk) 00:50, 28 February 2010 (UTC)[reply]

There's also 9" netbooks, smaller than the current standard 10". I watch my videos on it. A word of advice- don't get a Celeron model, get an Atom. They tend to give better battery life (my ASUS 900HA is rated for 5 hours. I can get 4 hours watching videos, with the screen brightness on low.) Mxvxnyxvxn (talk) 20:26, 1 March 2010 (UTC)[reply]

mcods.exe edit

  Resolved

I am frustrated an average of 3 or 4 times a week with slowness and lagginess on my Windows XP machine. When I check out the currently running processes with Windows Task Manager, I notice that mcods.exe is generally running and taking up around 40% of my dual-core CPU, meaning almost a whole core. This is apparently a McAfee virus scan. How do I tell it to get lost without just killing the process? I wanted to do this nicely via the McAfee GUI but I don't see any indication there that the virus scan is even running. This link suggests I use services.msc to set it to run only manually, but, oddly, McAfee Scanner is already set to run only manually. Comet Tuttle (talk) 20:52, 26 February 2010 (UTC)[reply]

I may have answered my own question in record time. Like any security minded fellow, I normally use an account with no administrator rights. I switched over to my admin account and the McAfee icon was right there on the Start bar, in the lower right, animating to show that a scan was ongoing. I launched McAfee from that icon and there was a nice Cancel button which halted the scan. Comet Tuttle (talk) 21:02, 26 February 2010 (UTC)[reply]
For the future, you might try checking McAfee's configuration to see if you can run the scan at Below Normal priority; it will still take CPU, but should relinquish it to user programs. XP doesn't support prioritized I/O so you might still have some slowdowns due to the file scan, but it shouldn't be quite as bad. If McAfee doesn't have a setting to configure, you could follow the steps given in the very last post in this thread on service priority. If you were on Vista or later, you could prevent it from interfering with other startup processes by changing the setting from "Automatic" to "Automatic (Delayed Start)", but that only deprioritizes the task's CPU and I/O during initialization, and wouldn't help once the service was up and running.
Alternatively, if you're comfortable mucking about with Windows settings, you could create a Task Scheduler task that runs the same program as the service, and the Task Scheduler will default to Below Normal priority (and with some mucking about with exporting, editing, and reimporting the config file would let you set arbitrary priorities). Then you can either set the services.msc entry to disabled, or if you're feeling daring, change the definition of the service so that it triggers your scheduled task.
Of course, this is all open to problems if McAfee does startup fixes to its services (not unreasonable for a virus scanner) or relies on communicating with the running service in a way that only works when the service manager is in control (quite likely). If that's the case, I'd consider looking into replacing McAfee with another scanner; in my experience most of the reputable virus scanners are roughly equally effective (they'll miss anything really new, but they'll catch a lot of the older stuff) and the primary differentiation is in their resource footprint. I've had decent experiences with AVG and Trend, bad experiences with McAfee and Norton. Your mileage may vary.
Of course, if you're as security conscious as I think you are, upgrading to a 64 bit OS (whether XP x64, Vista or Win7) would remove a large part of your security concerns, and if you've got a hardware firewall (e.g. most routers act as such), you might be okay with an on-demand scanner for specific files and skip the real time and scheduled scans unless you really want them. The 64 bit upgrade vastly improves your security profile: most buffer overruns (the source of worms, which anyone is vulnerable to, as opposed to trojans, which a security conscious user can avoid with on-demand scans) rely on filling a null-terminated string with executable code, but any memory address on current 64 bit machines contains two nulls, making it much harder to do anything useful with the overrun. —ShadowRanger (talk|stalk) 21:28, 26 February 2010 (UTC)[reply]

Japanese and similar characters in Firefox edit

When I look at pages with what should be Japanese characters, instead of each character I get a small box with four very small letters or numbers in it. Is there something I can download to show the Japanese characters? I do not understand any of them, but they look nicer. 92.29.32.229 (talk) 21:20, 26 February 2010 (UTC)[reply]

It is attempting to show Japanese (unicode) characters, but you don't have the font installed. For starters, install a basic CJK font (Chinese-Japanese-Korean). How you do that depends heavily on your operating system. -- kainaw 21:26, 26 February 2010 (UTC)[reply]
Help:Multilingual support (East Asian) is our help page on this subject. It has instructions for several different operating systems. Comet Tuttle (talk) 21:30, 26 February 2010 (UTC)[reply]

RAR vs. 7z edit

  Resolved

I need to store many thousands of .html files. They don't need to be accessed often, but occasionally they do or I might like to search them. Which is the better archive format for this task; RAR or 7z? I know from my own tests that 7z has slightly better compression results than RAR, but RAR supports searching archives without extracting. I realize this is a more of opinion than absolute facts, but any input, insight into other solutions etc would be appreciated. Thanks —Preceding unsigned comment added by 82.44.55.204 (talk) 21:38, 26 February 2010 (UTC)[reply]

All depends on how often you intend to extract/search them, vs. how much the compression really saves. I'd suggest compressing a representative sample of your files to compare real world compression ratios; the compression ratio for general files will differ quite a bit for HTML, given that HTML is much more structured than "typical" data. You might also try stuff like tar combined with bzip2 or gzip to see if they get better compression or faster searching. Or you skip tar entirely, and just bzip2 or gzip compress the individual files, so you can still index them by name efficiently, and decompress individually without the complexity of a full on archive+compression solution. —ShadowRanger (talk|stalk) 22:04, 26 February 2010 (UTC)[reply]
Addendum: HTML compresses well enough that even a solution like NTFS compression will get some decent results on it, and you can continue to use them transparently without the need to deal with manual decompression at all. —ShadowRanger (talk|stalk) 22:05, 26 February 2010 (UTC)[reply]
You don't say what OS you are using, but if it's a recent Windows, you might want to just tell Windows to compress that folder. On Vista, you'd right-click the folder, choose Properties, then Advanced, then click the appropriate check box. Since hard disks are US$110 for 1.5TB these days, I suspect that the number of bytes saved is probably not the actual priority here, so just having Windows compress the folder may be the winning solution for simplicity and search capability. Comet Tuttle (talk) 22:11, 26 February 2010 (UTC)[reply]
FYI, Windows's built-in compression *is* NTFS compression. It's not an OS feature, it's part of the file system spec. So you can't copy it to a non-NTFS drive and keep the compression, but at the same time, a non-Windows OS with a good NTFS driver could still read it. —ShadowRanger (talk|stalk) 23:02, 26 February 2010 (UTC)[reply]


Thank you everyone for all the helpful answers. I just did another test with 3gb of files; 7z compressed them down to 350mb (with solid compression) whereas WinRAR compressed to 530mb. So it looks like I'm going to go with 7z. I did consider NTFS compression but it doesn't compare at all to the space savings achieved with the archivers. Again, thanks for the help :)