Wikipedia:Reference desk/Archives/Computing/2009 December 17

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 17

Legal strength of corporate email signatures?

Couldn't find anything on Wikipedia addressing this. Corporate email signatures tend to claim all manner of silly things, for example If you are not the intended recipient, you are hereby notified that any use, forwarding, dissemination, or reproduction of this message is strictly prohibited and may be unlawful. Is there any legal precedent for this nonsense? Has it ever been litigated? 218.25.32.210 (talk) 02:13, 17 December 2009 (UTC)

Very interesting question. I don't have the resources to look up what you're asking exactly right now, but if you're interested in the Legal ethics question (for instance, when one law firms mistakenly sends confidential information to their adversary) then the ABA has specifically addressed the issue (and their rules have been adopted in most, but not all, states) then look at Model Rule 4.4 [1]. The short answer, in that context is the only obligation is to notify the other person they sent it and you got it. The ABA considered other options, like prohibiting the lawyers from reading it, but that was ultimately rejected.

Note that none of this speaks to other possible laws, including insider trading, trade secrets, and other general fraud statutes. I don't think anyone here can give you an answer on those, but I'd be curious to hear if anyone knows of any current (i.e. post ABA rules litigation) on that issue, or of any states that deviate from rule 4.4(b). Shadowjams (talk) 03:13, 17 December 2009 (UTC)

Consensus from my web browsing is that they do not bind the recipient, but may ameliorate the sender's liability under certain circumstances, outlined here. See also google, metafilter. --Tagishsimon (talk) 03:27, 17 December 2009 (UTC)

Under traditional [but maybe not some statutory] contract principles they won't bind a recipient, but they might be relevant for knowledge requirements in other laws, most notably fraud and insider trading statutes. Shadowjams (talk) 03:44, 17 December 2009 (UTC)

I kinda see what you're driving at, but is it not possession of the content of the email, rather than the legal rubric at the bottom of it, that is relevant w.r.t. other laws such as insider trading? Is there a case where absent the rubric there is no insider trading issue, whilst with the rubric there is? --Tagishsimon (talk) 04:25, 17 December 2009 (UTC)

Note - You might want to transfer this question to the Humanities reference desk. Shadowjams (talk) 03:46, 17 December 2009 (UTC)

Auto Connection to the internet

I've been able to make my PC wake up from standby and run specific tasks with the help of Windows Scheduler, but I haven't yet been able to make it connect to the internet automatically. Is there any way to do this, after the PC's woken up from hibernation?? 117.194.234.200 (talk) 07:42, 17 December 2009 (UTC)

Try creating a basic task that runs a program. The program is C:\WINDOWS\system32\ipconfig.exe. Browse to that and add /release to the parameters. Then make it run ipconfig again with the /renew switch.--Drknkn (talk) 07:49, 17 December 2009 (UTC)

Could you say a) how you connect to the internet and b) what it is you want to do when you wake the machine up and connect? --Phil Holmes (talk) 09:56, 17 December 2009 (UTC)

I download stuff from the net at night, when downloading's free under my broadband scheme. And I'm afraid I didn't entirely get what Drknkn was trying to tell. There's a lot of jargon there that I didn't understand. Could you please rephrase what you said into something easier for me to understand? For starters, how do I "creat a basic task that runs a program"? DO I have to use DOS or something? Thanks. 117.194.228.76 (talk) 11:51, 17 December 2009 (UTC)

What version of Windows are you using (XP, Vista, etc.)? In Vista, it's one of the options on the right of the Scheduled Tasks dialog. In XP, you'd just click "Add a Scheduled Task." You said you're using Windows Scheduler, so I assume you went into the Control Panel and selected "Scheduled Tasks" (in XP) or went to Administrative Tools and then Scheduled Tasks (Windows Vista)?--Drknkn (talk) 16:47, 17 December 2009 (UTC)

He wants you to use Scheduled Tasks to run "ipconfig.exe /release" and then "ipconfig.exe /renew". Personally I would use a batch file, writing up these two commands, and then making the batch file the Scheduled Task. Comet Tuttle (talk) 17:46, 17 December 2009 (UTC)

Though by the sound of it OP's problem is not connecting to the internet, but getting the computer to schedule the downloading of stuff in some way during the night? 81.131.7.99 (talk) 20:49, 17 December 2009 (UTC)

I use XP, and I went to Control Panel--> Scheduled Tasks earlier. Can you tell me how to write a batch file to run programs? I don't know much about these kinds of things... @81.131 - I can get the torrent client to open all right at night, but my problem is connecting to the internet, not vice versa. 117.194.226.134 (talk) 05:03, 18 December 2009 (UTC)

OK. I went to Command Prompt and typed "Start C:\WINDOWS\system32\ipconfig.exe /release" and then "Start C:\WINDOWS\system32\ipconfig.exe /renew", but the computer didn't get connected to the net. All that happened was that my LAN connection got repaired. Please help... 117.194.227.22 (talk) 05:14, 18 December 2009 (UTC)

If you're on a LAN, you almost certainly are "connected to the internet" whenever the PC is on. What makes you think that you're not? --Phil Holmes (talk) 15:34, 18 December 2009 (UTC)

No, no!! There's a default LAN connection in my PC, even though there's no Local Network.. I was talking about my broadband connection.... 117.194.227.59 (talk) 19:19, 18 December 2009 (UTC)

Is your broadband connection not always on? That's unusual, why not? 213.122.6.175 (talk) 12:44, 19 December 2009 (UTC)

A matter of a school IP I could be accused of using

I also need a few adminstrators to look this over as well, back when I first began on this website I had a few accounts, and I narrowed them down to one account.

Which is the one I'm using to post from, Theme Parker. The problem that I have is that of the accounts was registered from this IP that I'm using now to send in this question. There have been notices and a warning posted on it's talkpage.

My worry is, I don't want to be accused of Sockpuppetry on this website when I have nothing to do with the notices and edits involved with the IP.

So, what can I do?216.11.96.2 (talk) 13:21, 17 December 2009 (UTC)

Also, I add my proper signature to prove that I use ThemeParker, not the IP to post from above. ThemeParker 13:23, 17 December 2009 (UTC)

The reference desk is not the right place for this question. You should ask at WP:Help Desk, WP:AN or Wikipedia_talk:Sockpuppet_investigations. Friendly advice though from someone who's been in a similar situation; checkusers do not care about anything you say if the accounts are on the same ip and come from the same version of browser (which if they're from a school or public computer they most likely will), which is fucking ridiculous because how many people do you think are running Firefox 3.5.5 on windows xp? Loads that how many, millions and millions of people and it's NOT conclusive evidence that two people on the same ip rage use the same OS and browser. But try telling that to the checkusers. tl;dr if there is evidence against you, regardless of the actual facts you'll most likely be blocked and they won't even listen to your pleas that you're innocent.

That's a rather ill-considered and hostile piece of bad advice from the anonymous user above. A whois to 216.11.96.2 reveals it is Oakland Schools. Checkusers would tend to spot that.

The advice I'll give (even if it is in the wrong place) is, mark the user page of any alternative account - as opposed to any IP address - with {{Alternative account}} and, ideally, stop using the alternate account. Do not worry about edits coming through from IP addresses you have used, or through which you use your account. As noted, checkusers understand that school IPs, in particular, are compromised, and will take that into account should there be further suspicions of sockpuppetry. --Tagishsimon (talk) 13:55, 17 December 2009 (UTC)

What is this "swarming' effect called?

I'm searching for papers on this topic, but I don't know what it is called. I can only describe it - so Googling is not being very effective. Assume you are keeping a top 20 list of something (perhaps the top 20 songs you've played on your mp3 player). After a while, each item in the top 20 will have a large count - say over 1,000 each. Now, you add a new song. It is your favorite. But, teh only way to get it into the top 20 list is to listen to it repeatedly over 1,000 times. I see this as a swarming effect because the list swarms around a limited subset of the entire set. Getting it to swam around something new is hard. I want to find (and read) papers that discuss methods for getting the new item into the "swarm", but I can't find anything because I don't know what keywords to use to get Google to find anything useful. -- kainaw™ 13:51, 17 December 2009 (UTC)

I think, depending on how you look at it, this is Sampling bias or Selection bias and may well also be covered in Bias (statistics). The bias in this case might be inappropriate sample period - say of 1 year when your new fave rave is only a fortnight old, or (which is much the same thing) an inappropriate algorithm, such as counting average frequency over the year rather than average frequency per shorter time period. Swarming does not really convey anything to me, but you;re welcome to your own similes ;) --Tagishsimon (talk) 13:59, 17 December 2009 (UTC)

I think this is the same situation: If there is a "top 10" list of products online, people tend to click the items in the top 10, which self-reinforces the top ten products remaining in the top 10 forever. (Apple's App Store has a similar issue.) Is there a name for this? Comet Tuttle (talk) 17:49, 17 December 2009 (UTC)

It reminds me of the Matthew effect. --Mr.98 (talk) 18:14, 17 December 2009 (UTC)

Comet Tuttle's description is exactly what I'm referring to. The Matthew Effect article links to Positive feedback, which is another name for the same thing. Now, I'm going to use those to see if I can find some technical papers discussing causes and fixes for positive feedback. I want to learn more about how to make something like a Top 10 list that will not reinforce itself into being a dominant Top 10 excluding all newcomers. -- kainaw™ 19:39, 17 December 2009 (UTC)

It's an interesting topic, even more so because the portals (such as the App Store, iTunes, Amazon...) don't see this as an immediate negative; if the Top 10 list drives sales of anything then they benefit. They will probably eventually care that their list eventually looks stale. This is OR and I don't have a reference, but some solutions include having separate "Top 10 of All Time" and "Top 10 Of The Week" lists, and the latter starts out zeroed every week; or keeping it on a single list, but having old sales decay in their importance to the rankings as time goes on. More OR: The former is probably the better choice if the objectives include accuracy. Comet Tuttle (talk) 20:29, 17 December 2009 (UTC)

The Matthew Effect article also links to preferential attachment, which sounds more precisely like your top 20 example. 81.131.7.99 (talk) 21:08, 17 December 2009 (UTC)

In something like this I also have the top fastest rising ones. It's a bit complicated, I use the rise divided by the square root of the count and then just give the first ones in popularity order that are amongst the top third when sorted by this. Rather ad-hoc. Dmcq (talk) 13:21, 18 December 2009 (UTC)

By the way you might like the Collaborative filtering article. Dmcq (talk) 12:42, 21 December 2009 (UTC)

2 Questions... help with excel and the taskbar

Excel

First off, I'm trying to print an excel spreadsheet onto a single paper, and the problem is that excel automatically re-sizes my spreadsheet to fit the paper. This is fine, except my spreadsheet has more columns than rows, and the result is that the information it contains ends up being very small (almost unreadable), taking up only 1/3 of the paper's space. Changing the fonts to make it larger or smaller makes no difference, as in the printing page setup, excel automatically scales the spreadsheet to fit onto a paper. The problem, basically, is that it is keeping both width and height scaled proportionally, while I just want to scale the height, so that it fits the entire paper (it's in landscape format), and not just a strip across the center of the page. I've tried everything. I even tried to just save it into paint and scale it there, but I can't copy it all because the data file is too large. Any help?--72.178.133.37 (talk) 16:22, 17 December 2009 (UTC)

What version of Microsoft Office are you using? —Akrabbim^talk 17:00, 17 December 2009 (UTC)

In Excel 2000 (Yes, I know I'm out of date!), one just goes into File --> Print preview --> Setup. Then there is a choice of scaling to any percentage, or fitting to any number of pages. Click the former for flexibility. The sytem may be different in recent versions of Excel. Dbfirs 17:22, 17 December 2009 (UTC)

I have the latest Excel. I tried that feature, and like I said, it doesn't allow you to scale width or height individually, which is what I want to do. And I want the spreadsheet on just one paper. Like I said, my spreadsheet has more width than height, and I want to scale the height so that it fills the entire page, and not just a strip. In other words, I want to stretch it vertically, so that it becomes readable without having to strain your eyes. --72.178.133.37 (talk) 18:12, 17 December 2009 (UTC)

I toyed with Excel 2007 for a bit just now, and can't find how to do this from within Excel, either. Have you tried printing to a PDF file and perhaps some PDF reader has the option to print-to-fill-the-page? (I'm skeptical, by the way, that the results will look good with text that is so stretched vertically.) Comet Tuttle (talk) 18:31, 17 December 2009 (UTC)

Well, I'm sure if I scaled it to completely fit the page, it wouldn't look great. But I want to at least be able to manually stretch it as I see fit. Does OpenOffice have this problem too?--72.178.133.37 (talk) 18:57, 17 December 2009 (UTC)

I just tried with OpenOffice 2.0.4 (note: latest version appears to be 3.1.1) and in 2.0.4, at least, the options appear to be the same as in Excel — I can fit to page, but not independently stretch the vertical axis. Any luck with a PDF reader? Comet Tuttle (talk) 20:14, 17 December 2009 (UTC)

Increasing the row height will allow the layout to fill the page, but only by leaving bigger gaps between the rows. To stretch the characters vertically, you will have to save the output as an image (by printing and re-scanning if necessary), then you can stretch it as you wish as an image in Word (or various other programs). sorry I didn't read your question properly at first Dbfirs 21:33, 17 December 2009 (UTC)

Yeah, what he said. If you don't mind, could you tell us why you want to do this? Maybe we can come up with an alternate solution altogether. You seem concerned that your text is horizontally scrunched, but stretching the vertical won't change that, you'll just have text that's stretched (or gapped) and scrunched. If you want to force the gaps, you can manually alter the size of the rows, but you'll still have tiny text. Matt Deres (talk) 12:22, 18 December 2009 (UTC)

Why not print to two pages, then just put one half on top of the other to to get the full (split) sheet on one page? You can do this in Word without having to scan or manually cut-and-paste. Dbfirs 12:53, 18 December 2009 (UTC)

Taskbar

One thing I have noticed is that sometimes, when I have many windows open on my Windows Vista Home Edition, Windows sometimes decides to merge tabs of the several instances of the same program on the taskbar. For stuff like Firefox, that's no problem. But for stuff like instant messengers, it's a pain, especially since when I right-click on it and have to click TWICE (instead of once) every time I want to open a particular window when it is hidden by firefox being maximized. there is no option to "undo" the merging. Even closing the other programs doesn't undo it. It just stays merged. The only way that I know of is closing the group and then restarting the program. Is there any other way? Thanks.--72.178.133.37 (talk) 16:22, 17 December 2009 (UTC)

You can turn off grouping by right-clicking the Taskbar, selecting Properties, then unchecking "Group similar taskbar buttons" and hitting Apply. —ShadowRanger ^(talk|stalk) 16:49, 17 December 2009 (UTC)

Thanks.--72.178.133.37 (talk) 18:58, 17 December 2009 (UTC)

But I still would like to know if there is a way to just "undo" the grouping, because sometimes the grouping is useful (as with firefox instances).--72.178.133.37 (talk) 18:58, 17 December 2009 (UTC)

I know of no way of doing so. However, I have found that in some cases, disabling and then reenabling grouping will "teach" Vista to keep them next to each other, as opposed to grouping them. As for Firefox, why do you have more than one window open? Tabbed browsing should mostly obviate the need for more than one or two Firefox windows. —ShadowRanger ^(talk|stalk) 19:31, 17 December 2009 (UTC)

Because I'm the kind of person that has many tabs open, and I want tabs that are related to be grouped together, and I do that by having different instances of firefox running.--72.178.133.37 (talk) 19:57, 17 December 2009 (UTC)

For many years now I've been using an enormous, auto-hiding taskbar on the left edge of the screen. Lots of items fit onto it because they stack vertically, and you can read their full titles. This was inspired by a desire to open new windows in firefox instead of use tabs (why should I use tabs when the system already provides a taskbar?) ... but I've gone over to the dark side now and started using tabs because huge numbers of new windows in firefox seemed to cause crashes, or at least take up more memory than tabs. They needn't do, but that seems to be how it's designed now, with the expectation that users will use tabs. 81.131.7.99 (talk) 21:21, 17 December 2009 (UTC)

system variable

I found a variable in a batch file which is %cd:~0,2% .How can it be interpreted and what does it mean? —Preceding unsigned comment added by 113.199.186.108 (talk) 18:48, 17 December 2009 (UTC)

To find out what a system variable is, echo it out. In this case, you'd type echo %cd:~0,2% at the command prompt, which gives the value of the current drive. In my case, the output was C:.

Whenever you type cd at the command line, it prints out the current directory. The same is true when you type echo %cd%. The ~0,2 is shorthand for the first two letters of output. In other words, it shortens the output of cd to the first two letters, which ends up being the drive letter.--Drknkn (talk) 19:09, 17 December 2009 (UTC)

Thanks a lot, wonderful answers.Thanks again! —Preceding unsigned comment added by 113.199.179.126 (talk) 02:01, 18 December 2009 (UTC)

Lucida Grande question

I have a Mac using OS X 10.4. I'd like to use the characters "barred small capital i" (U+ID7B, ᵻ) and "barred Latin upsilon" (U+1D7E, ᵾ), which I've heard are available on newer versions of Lucida Grande (they even showed up fine on a school computer using Windows Vista), but on my version they just show up as boxes. Is there any way I can get these characters? --Lazar Taxon (talk) 19:33, 17 December 2009 (UTC)

Using one application on two computers

How to connect two desktop computers and use the same application program on one of them? —Preceding unsigned comment added by WoodRon (talk • contribs) 20:26, 17 December 2009 (UTC)

What OS are you running? On a Windows machine, for some simple applications that are currently on Computer A, you could turn on file sharing on Computer A, and walk over to Computer B, and mount Computer A's hard disk as a network hard disk on Computer B. Then open that mounted hard disk, double-click the application, and it may work. However, most Windows applications add data to the Windows Registry that is required in order to run the app at all. In these cases, you can't really run the application remotely in this way. (If you were using two GNU/Linux machines, this would not usually be a problem, AFAIK.) What may still work for you, though, is using Windows Remote Desktop or VNC so that you use Computer B to control the screen, mouse, and keyboard of Computer A; Computer A is always running the application, but you're using Computer B to control and view it; and you can usually copy and paste stuff onto Computer B's clipboard while you are working. That may do what you want. (I added a section header for you — next time please click the "new section" tab at the top of the page so when you ask a new question you can just type in a new header.) Comet Tuttle (talk) 20:40, 17 December 2009 (UTC)

Mencoder problem

Resolved

Hello! I'd post this on a Mencoder forum, but most that I found had a lot of questions unanswered, so I figured there were some Wikipedians familiar enough with mencoder to be able to help me. All I want to do is take a bunch of jpegs and make a single video out of them. I basically copied the instructions for how to do this from the mencoder website, but it doesn't work. I'm using a mencoder build for Windows from November this year, and for the most part, it's been working okay. (I haven't been able to get the build of the most recent sources.) Here's my command-line output. I tried to turn the jpegs into an mpeg and a mjpeg, but neither option worked. I tried flv before, which basically got me the same error message as a mpeg. I don't really care about the format, as long as it compresses out all the redundant frame-by-frame data of the jpgs (so mjpeg probably wouldn't be a good format for that, but I tried it because I figured it had the most chance of working). Where's my problem? Thank you very much for any feedback!

Microsoft Windows [Version 6.0.6002]
Copyright (c) 2006 Microsoft Corporation.  All rights reserved.

C:\Users\<my name>>cd downloads

C:\Users\<my name>\Downloads>cd MPlayer-rtm-svn-29851

C:\Users\<my name>\Downloads\MPlayer-rtm-svn-29851>mencoder.exe mf://C:\users\<my name>
\jpegs\*.jpg -mf w=800:h=600:fps=5:type=jpg -ovc lavc \ -lavcopts vcodec=mp
eg4:mbd=2:trell -oac copy -o C:\users\<my name>\desktop\output.avi
MEncoder Sherpya-SVN-r29851-4.2.5 (C) 2000-2009 MPlayer Team
success: format: 16  data: 0x0 - 0x0
MF file format detected.
[mf] search expr: C:\users\<my name>\jpegs\*.jpg
[mf] number of files: 100 (400)
VIDEO:  [IJPG]  800x600  24bpp  5.000 fps    0.0 kbps ( 0.0 kbyte/s)
[V] filefmt:16  fourcc:0x47504A49  size:800x600  fps:5.000  ftime:=0.2000
Opening video filter: [expand osd=1]
Expand: -1 x -1, -1 ; -1, osd: 1, aspect: 0.000000, round: 0
==========================================================================
Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family
Unsupported PixelFormat -1
Selected video codec: [ffmjpeg] vfm: ffmpeg (FFmpeg MJPEG)
==========================================================================
File not found: '\'
Failed to open \.
Cannot open file/device.

Exiting...

C:\Users\<my name>\Downloads\MPlayer-rtm-svn-29851>mencoder.exe mf://C:\users\<my name>
\jpegs\*.jpg -mf w=800:h=600:fps=5:type=jpg -ovc copy  -oac copy -o C:\user
s\<my name>\desktop\output.avi
MEncoder Sherpya-SVN-r29851-4.2.5 (C) 2000-2009 MPlayer Team
success: format: 16  data: 0x0 - 0x0
MF file format detected.
[mf] search expr: C:\users\<my name>\jpegs\*.jpg
[mf] number of files: 100 (400)
VIDEO:  [IJPG]  800x600  24bpp  5.000 fps    0.0 kbps ( 0.0 kbyte/s)
[V] filefmt:16  fourcc:0x47504A49  size:800x600  fps:5.000  ftime:=0.2000
videocodec: framecopy (800x600 24bpp fourcc=47504a49)
Writing index...
Writing header...
ODML: Aspect information not (yet?) available or unspecified, not writing vprp h
eader.

Video stream:      nan kbit/s  (-2147483648 B/s)  size: 0 bytes  0.000 secs  0 f
rames

--el Aprel (^facta-_facienda) 21:26, 17 December 2009 (UTC)

In your first attempt, you took an example command that was split over 2 lines with a backslash at the end of the first line, and joined them together, and kept the backslash. That broke it. It was written that way in the first place to allow it to be pasted directly into a unix shell, where a backslash at the end of the line means "command not finished yet, continued on next line". I presume that doesn't work with that DOS-looking shell you're using, so you need to join the lines manually and get rid of the line-joining backslash. That's why it said it had tried to open a file called '\', because it couldn't make sense of the backslash in any other way than interpreting it as an input filename.

I'm not sure what went wrong with your second attempt. Try it again and add the -v option to get more information. Adding it twice (-v -v) is even better. Also, is 800x600 the correct size for your jpegs? If not, fix the w=800:h=600 part. 98.226.122.10 (talk) 10:20, 18 December 2009 (UTC)

Thank you for your response! I removed the \ as you suggested, but it still didn't work. All that happens is I get an avi file on my desktop that's only 4.01 KB large (just a header, I would imagine), and it won't play on any video software, not even MPlayer. I ran it with the -v -v option, and here's what the command line outputs:

Command-line output

C:\Users\<my name>\Downloads\MPlayer-rtm-svn-29851>mencoder.exe mf://C:\users\<my name> \jpgs\*.jpg -mf w=800:h=600:fps=5:type=jpg -ovc lavc -lavcopts vcodec=mp eg4:mbd=2:trell -oac copy -o C:\users\<my name>\desktop\output.avi -v -v MEncoder Sherpya-SVN-r29851-4.2.5 (C) 2000-2009 MPlayer Team Configuration: --extra-cflags=-I/c/Work/mplayer/live --prefix=/mingw --enable-st atic --enable-md5sum --enable-menu --disable-vdpau --disable-vidix --disable-sdl --enable-caca --enable-gl --enable-matrixview --enable-freetype --enable-png -- enable-mng --enable-jpeg --enable-gif --enable-tga --enable-mad --enable-tv --en able-theora --disable-dvb --disable-dvbhead --enable-runtime-cpudetection init_freetype Using MMX (with tiny bit MMX2) Optimized OnScreenDisplay Config pushed level is now 2 Config pushed level is now 3 Setting mf=w=800:h=600:fps=5:type=jpg get_path('fonts') -> 'C:/Users/<my name>/Downloads/MPlayer-rtm-svn-29851/mplayer/f onts' WINSOCK2 init: 0 STREAM: [mf] mf://C:\users\<my name>\jpgs\*.jpg STREAM: Description: Multiple files input STREAM: Author: Benjamin Zores, Albeu STREAM: Comment: success: format: 16 data: 0x0 - 0x0 s->pos=0 newpos=0 new_bufpos=0 buflen=0 MF file format detected. [mf] search expr: C:\users\<my name>\jpgs\*.jpg [mf] number of files: 100 (400) ==> Found video stream: 0 VIDEO: [IJPG] 800x600 24bpp 5.000 fps 0.0 kbps ( 0.0 kbyte/s) [V] filefmt:16 fourcc:0x47504A49 size:800x600 fps:5.000 ftime:=0.2000 SPU: Display only forced subs now disabled WINSOCK2 init: 0 WINSOCK2 init: 0 [file] File size is 0 bytes STREAM: [file] C:\users\<my name>\desktop\output.avi STREAM: Description: File STREAM: Author: Albeu STREAM: Comment: based on the code from ??? (probably Arpi) Opening video filter: [expand osd=1] Expand: -1 x -1, -1 ; -1, osd: 1, aspect: 0.000000, round: 0 ========================================================================== Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family Unsupported PixelFormat -1 INFO: libavcodec init OK! Selected video codec: [ffmjpeg] vfm: ffmpeg (FFmpeg MJPEG) ========================================================================== ds_fill_buffer: EOF reached (stream: video) Flushing video frames. Filters have not been configured! Empty file? Writing index... Writing header... ODML: Aspect information not (yet?) available or unspecified, not writing vprp h eader. Video stream: nan kbit/s (-2147483648 B/s) size: 0 bytes 0.000 secs 0 f rames Uninit video: ffmpeg DEMUXER: freeing demuxer at 0275EB18 DEMUXER: freeing sh_video at 0275FDC8 WINSOCK2 uninit

And yes, all of my jpg images are 800x600.--el Aprel (^facta-_facienda) 22:28, 18 December 2009 (UTC)

I don't know what went wrong there either. The "EOF reached" appears where it should be reading the first jpeg. Earlier, it found "number of files: 100" (correct number of jpegs I hope) but then when the time comes to start copying them, it doesn't find them. Maybe there's some trouble with the file globbing. Try it this way: cd into the directory containing the jpegs and run it with mf://*.jpg 98.226.122.10 (talk) 23:54, 18 December 2009 (UTC)

The more I look at it, the more I think this will be the answer. mplayer/osdep/glob-win.c contains a really sad-looking imitation of a glob function which uses FindFirstFile/FindNextFile to do the matching. It's quite possible that it forgets what directory it found the jpegs in. The only missing piece is some official documentation stating that FindFirstFile/FindNextFile strips the directory components off the results. 98.226.122.10 (talk) 00:24, 19 December 2009 (UTC)

Thank you so much, 98.226.122.10! Your last suggestion got it to work fine. Just goes to show once more how smart and resourceful the people who frequent Wikipedia are. Thanks again!--el Aprel (^facta-_facienda) 04:23, 19 December 2009 (UTC)

If someone reported this to mplayer's bug database, it would probably get fixed to make mf://somedir/*.jpg work in a future version. 98.226.122.10 (talk) 06:30, 19 December 2009 (UTC)

Remote desktops

Why does remote dekstop software work by sending an image of the rendered screen? Wouldn't it be more efficient (higher frame rate at the same bandwidth) to send the data and instructions being sent to the video driver, at least when the client's video driver provides the same API? NeonMerlin 22:06, 17 December 2009 (UTC)

I'm going to guess that the instructions being sent to the video driver are in fact more data than compressed jpg images. Think about it, the video driver needs to accurately make every single pixel whereas sending compressed images takes much less data. —Preceding unsigned comment added by 82.43.88.54 (talk) 22:15, 17 December 2009 (UTC)

What I think you're proposing is to intercept calls to the display subsystem API (commands like "draw_rectangle, draw_text, draw_line). If that's the case, many remote desktop protocols do just that. The X protocol (the network layer of the X Window System deals in display commands, as did NeWS (which used display postscript as its description language). This extension for RDP intercepts GDI/GDI+ commands and emits them over the network. The downside is that when the existing graphics infrastructure isn't built this way (as it mostly isn't in Windows) it's a bigger job to splice into the graphics subsystem to intercept those calls. If there are a lot of bitmaps getting shoved around, it's probably more efficient to just do a remote framebuffer (which is command agnostic); that's what VNC, SunRay, and mostly RDP do (SunRay did it to keep the load on the diminutive CPU to a minimum). Note that the bitmap ones send changes (not the full screen each time) and use various kinds of delta elision and bitmap compression to keep both the CPU and network usage to a manageable level. -- Finlay McWalter • Talk 22:39, 17 December 2009 (UTC)

Another possible concern is that the rendering will appear differently on the client if, say, its video driver differs from the server's, so font rendering might differ. Granted, a minor concern in most cases. Comet Tuttle (talk) 23:32, 17 December 2009 (UTC)

Fonts raise an interesting problem for the networked-GDI/Xlib style remote desktop. It's very common for client libraries to enquire about how big a given string will be when rendered using a given font (using something like getTextExtent or the like). For calls like that, the client (the program you're running) has to enquire about this to the server (the display) and await the response, and the program (and the toolkit that it uses) almost always processes these calls synchronously (and thus serially), with each taking a few ms of back and forth network time. When rendering a complex control like a menu or a tree widget, there might be dozens of such calls. So while the bandwidth associated with this traffic is trivial, the latency is unpleasant. Stuff like this can be mitigated by client side caching (in the library or the remoter), but when the cache fails this can amount to quite an appreciable delay (even though the communications like is mostly unused during the transaction). -- Finlay McWalter • Talk 00:13, 18 December 2009 (UTC)

File formats for Geographical information systems or mapping software

I'm writing a program that outputs some data about noise intensity for a grid of x,y coordinates, and the coordinates of roads. It would be nice if I could use some existing no-pay (freeware, open source etc) software to display this visually. Is there any such software that has a simple file format? Or is there an existing standard for GIS files? If so, then I would modify my program to output its results in that file format. Thanks 78.146.167.109 (talk) 22:16, 17 December 2009 (UTC)

We have an article, GIS file formats that lists many of them. A variety of software also exists to display the common ones, such as DTED, but I am unsure exactly which one would best suit your purposes. decltype (talk) 22:21, 17 December 2009 (UTC)

It may be best if I could find out what free GIS or mapping software there is, and then work backwards to find what file formats they use. 78.146.167.109 (talk) 22:30, 17 December 2009 (UTC)

Perhaps USGS DEM, which is essentially a heightfield format. Or you could just emit it as a simple format (like Netpbm format) and render it as a heightfield using a renderer like POV-ray. -- Finlay McWalter • Talk 22:44, 17 December 2009 (UTC)

I'm looking for something which will show the data in two dimensions only, like a map. Generating a 3d image would cause confusion. 89.242.147.247 (talk) 00:16, 18 December 2009 (UTC)

Then you might as well just emit a simple 2D bitmap format like netpbm and view it with an ordinary image viewer. Beyond that, I don't see what additional display behaviour or file characteristics you need that a GIS format would have that a plain old bitmap wouldn't. -- Finlay McWalter • Talk 00:20, 18 December 2009 (UTC)

Thanks, that is a raster-like format. Are there any vector formats please? Netpbm is still interesting though and I will study it further when I have more time. 89.241.43.33 (talk) 11:58, 18 December 2009 (UTC)

It's pretty easy to generate SVG documents even without an SVG library, or you may wish to do the work in gnuplot. -- Finlay McWalter • Talk 17:10, 18 December 2009 (UTC)

How easy please? I'm just an amateur using BASIC. I do not understand C or anything that is not very basic. 89.243.91.31 (talk) 21:46, 18 December 2009 (UTC)

SVG files are very simple text files (with some rather inscrutable boilerplate text at the start). This tutorial gives an excellent introduction to the various simple primitives SVG has. For an application like yours, you'd probably just emit a bunch of line or circle (using small circles for points) commands. -- Finlay McWalter • Talk 22:41, 18 December 2009 (UTC)

Performance of average home computer of today compared to computers in the past

In terms of speed and memory, how does the average new home desktop computer compare to computers of previous decades? I dimly remember computers called workstations I think - how do they compare? And how far back do you need to go to reach a time when the best computers in the world would only be equal to the average new home computer of today? 78.146.167.109 (talk) 22:36, 17 December 2009 (UTC)

There is a problem in referring to "the best computers in the world". It depends on the task. Home computers are multi-functional computers. They are designed to handle many different tasks from simply mathematics to database management to video display to gaming. Any specific one of those tasks is handled better by customizing a computer to the job. For example, ray tracing involves performing the same task over and over on a large set of numbers. So, a scalar computer, like a home computer, has to load each value one by one and do the calculation. A vector computer can load all the values at once and perform the operation on all of them (nearly) at once. So, for say 100 values, a vector computer that is 100 times slower than a modern scalar computer would get the job done in about the same time. -- kainaw™ 23:24, 17 December 2009 (UTC)

This entry covers historical supercomputers in terms of processing power: Supercomputer#Timeline of supercomputers. The fastest processor today for home use is the Intel Core i7, which processes 69 billion floating-point poperations per second (GFLOPS). That equates to the processing power of a supercomputer from the early 1990s. A fast Pentium IV from a few years ago did about 10 GFLOPS, which equates to a super-computer from the late 1980s.

Still, the average computer today has a faster hard drive than those mainframes. Hard drives in those days transferred data at about 4 megabytes per second (MBps) whereas the average drive today transfers data at about 200 MBps. RAIDs weren't invented until the late 1980s. It also has a faster bus. In those days, computers used ISA buses, which slows communication between the graphics card, printer, network card, etc. Later came PCI and AGP, and now we're all using PCI-Express. Likewise, memory was also dramatically slower. So, although the processing power is equivalent to some computers back then, the real-world performance (e.g., opening programs, copying files) was vastly inferior to even ordinary computers today.--Drknkn (talk) 23:45, 17 December 2009 (UTC)

200 MBps is a little high. Take a look at this review of the Seagate Barracuda XT [2]. The maximum speed (i.e. at the outer edge of the platter) is under 160 MBps. The average is 110 MBps. The same here for example [3]. While this is a 2TB drive, that's actually close to the highest you can expect for a 7200RPM drive because these things need dense platters [4]. See [5] and [6] for example. (You may get a higher average on a short stoked drive but obviously the maximum is still going to be about the same.) Laptop magnetic/mechanical drives would generally be slower. Obviously 15k RPM drives and the like will do better [7] but they aren't in average computers and even then may not achieve an average of 200 MBps. Fancy SSD drives would do a lot better some even saturating a SATA 3 Gbps link (see some of the earlier links for example) but while they're becoming more common particularly in laptops they're still not likely to be in the average computer. And the cheaper ones may not reach 200 MBps [8]. Some won't even beat desktop magnetic drives (not sure about laptop drives) in max speed, maybe even the average or lowest sequential speed (obviously random seek times will always be incomparable) [9] Nil Einne (talk) 17:28, 18 December 2009 (UTC)

Mainframes used all sorts of proprietary interfaces. The idea of using more than one device at a time to improve I/O didn't start with RAID and there were special devices with more than one head to improve I/O - but yes it was still fairly slow. Even so the central processors nowadays have improved speed faster than the peripherals and the reason the peripherals haven't become the limiting factor is that memory is so cheap one can keep the data in memory; the peripherals need only be used at the start and end. Even quite large databases can now be kept in memory and the disks just written to but not read. Some people use only part of their disks and have more disks so they can do more I/O in parallel, extra disk capacity is of no use to them. Dmcq (talk) 11:45, 18 December 2009 (UTC)

I'm surprised that a better home computer of today would be as good as the world's best computers from 1990 - I expected the date to be much further back. All that computer power going to waste on computer games rather than solving world peace, hunger, etc. 89.241.43.33 (talk) 12:02, 18 December 2009 (UTC)

Bit off the original topic here, but they're "human" problems, it doesn't how much computer power you have, it's not going to solve those issues. ZX81 ^talk 12:22, 18 December 2009 (UTC)

A feature filled operating system and lots of free goodies when the system is sold can give all that power a more human scale ;-) Dmcq (talk) 13:13, 18 December 2009 (UTC)

Request Entity Too Large

My main browser is Google Chrome and while attempting to access the New York Times website - and specifically the TimesWire feature - I get an error message, "Request Entity Too Large" followed by "A request entity is longer than the server can handle." TimesWire works fine in my other browser, Internet Explorer 8. I've sent a message to the Times who reported no problems and reported the bug using the Chrome bug report feature. (Here's a screenshot.) What should I do? --Blue387 (talk) 23:01, 17 December 2009 (UTC)

The same error has been reported by NY Times users of Firefox and Safari at various times, as well as other Chrome users. Some people suggest clearing out cookies (surely only the nytimes.com ones) and if that fails, the cache too. Right now, on Chrome-x86-linux, TimesWire works fine for me. -- Finlay McWalter • Talk 23:13, 17 December 2009 (UTC)

I have cleaned out the cache and cookies, the page is currently working fine. Thanks! --Blue387 (talk) 05:46, 18 December 2009 (UTC)

In which case, contrary to previous discussions linked, I would say it definitely is a bug in the NYT site. Looking at the cookies set by that page (and there's a hell of a lot of them!), I see several that appear to be comma- or pipe-delimited lists. My guess is that at least one of these is being appended to every time you view that page (or every time you view it and it's changed?), and never cleared - why there are so many massive cookies instead of one keying into a server-side session I can only guess.

Eventually, this ever-growing cookie reaches a critical length whereby, along with all the other cookies and miscellaneous headers, attempting to send it back exceeds the limit on request length of their web server - Sun ONE, according to the headers. The message is presumably produced by the browser because it received HTTP Status 413 from the server. Since neither the server nor the client realises that this particular cookie is to blame, the error will persist every time you visit that page until the cookie is manually deleted. If before deleting them, you viewed all the cookies set, you would presumably see the mammoth that was causing the problem.

Now, anyone know a developer at NYT (or whoever runs that part of their site) who can get this fixed? ;) - IMSoP (talk) 19:10, 20 December 2009 (UTC)