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

Computing desk
< May 17 << Apr | May | Jun >> May 19 >
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 18 edit

Microsoft Visual C++ Runtime Library edit

Hello. When I insert an Encyclopaedia Britannica 99 CD into my CD writer, I can search for definitions in the dictionary. If I leave my Pentium 4 computer idle for a few minutes and I search for another word, my computer initiates a search. A window appears and says:

Microsoft Visual C++ Runtime Library
Runtime Error!
Program: C:\PROGRA~1\BRITAN~1\BCD\Program\g.exe
abnormal program termination

The only buttons I can click are OK and the X button on the upper-right corner of the window. After I click on one of those buttons (it does not matter which button I click), the dictionary closes. To make the dictionary work again, I must put the CD back into the DVD-ROM. However, this problem does not occur in newer model computers. Why is this so? Thanks in advance for your help. --Mayfare 01:06, 18 May 2007 (UTC)[reply]

There is an error in the program you use to view your encyclopedia (the one that is on the cd together with the data). There is a plethora of reasons why it could happen only on some computers (for exapmle a race condition). The only thing you can do about it is to see if there is an update for the encyclopedia software from the publisher that fixes the bug.

Apache, .htaccess edit

Hello all,

I have created an .htaccess file to protect a subdirectory in an Apache server. To test it, I took the first file at hand, in this case 'pdftk.1.txt', and uploaded it to the subdirectory. Strangely I was able to download it. I tested the configuration with many more files, but just this was not protected. Why does Apache do not protect files like *.1.*? Mr.K. (talk) 03:44, 18 May 2007 (UTC)[reply]

It's really hard to diagnose something without knowing what it is in the first place. Maybe you should put your .htaccess file up here? --antilivedT | C | G 09:24, 18 May 2007 (UTC)[reply]
Also, we need to know the httpd.conf or apache.conf settings. Is your Apache configured to allow user .htaccess files? --Kainaw (talk) 12:19, 18 May 2007 (UTC)[reply]
Check the <directoy> stuff in httpd.conf is set up to check for .htaccess. I forget the exact command - check apache.org --h2g2bob (talk) 20:08, 18 May 2007 (UTC)[reply]

My .htaccess file is:

 AuthUserFile /home/content/.../.htpasswd
 AuthGroupFile /dev/null
 AuthName "Members Only Area"
 AuthType Basic
 <Files pdftk.1.txt>
 require valid-user
 </Files>
 require valid-user

With or without <Files... </Files> the file like *.1.* are not being protected...This is in a shared server and I don't have access of httpd.conf or apache.conf. Mr.K. (talk) 02:28, 19 May 2007 (UTC)[reply]

If Apache is not configured to allow users to override base settings with a .htaccess file, it doesn't what you put in your .htaccess file. It will be ignored. So, before going crazy trying to get your .htaccess file working, ensure that you are allowed to have an .htaccess file first. --Kainaw (talk) 03:32, 19 May 2007 (UTC)[reply]
Users do are allowed to override the settings. The problems is that files like *.pdf or *.doc are being protected and files like *.1.* not. It is not lifethreatening, since I could let a couple of files unprotected, but I still want to understand why this is happening. Mr.K. (talk) 13:14, 19 May 2007 (UTC)[reply]

Half Life 2 List of Weapons edit

Where has the page which portrays all the weapons in half-life 2 gone?! I've tried searching and it comes up with a list of combine technology.

It was probably deleted for being a guide or cruft -- Phoeba WrightOBJECTION! 08:40, 18 May 2007 (UTC)[reply]
See it's AfD discussion. It lives on at the Internet Archive's Cached version from May 2006. --h2g2bob (talk) 20:01, 18 May 2007 (UTC)[reply]
Also appears to have been transwikied to stratergywiki. --h2g2bob (talk) 20:05, 18 May 2007 (UTC)[reply]

Uploading SVG Images edit

Hi I recently created

File:ChaserWar.svg
this

vetor image, it looked great in inkscape but when I uploaded it in firefox it turned out all munted. Any Ideas what when wrong? --Lwarf 10:15, 18 May 2007 (UTC)[reply]

It looks broken in Firefox's SVG renderer too. There's something weird about the three text objects - when you select them (with the text tool) in Inkscape, their blue boxes are inexplicably huge. If I were you I'd recreate the three text objects again. -- Finlay McWalter | Talk 10:28, 18 May 2007 (UTC)[reply]

The Text is ment to be huge, itś suppost to be the chasers war on everything title page for a userbox. A smimlar thing happend with

 
this

which is ment to look like

 
this

--Lwarf 10:41, 18 May 2007 (UTC)[reply]

You misunderstand me. If I make a new text object in Inkscape, its blue text-clipping box is the same as its normal select box. But your text objects have blue boxes that are vastly bigger than their select boxes - much much bigger than the SVG's cliprect. It looks to me like you made the text objects by dragging with the text tool (to make a blue text box), typed text into them, and then scaled the text to fill the space you wanted. It seems Inkscape's renderer supports that properly but libsvg doesn't. If you make new text objects just by clicking with the text tool and then typing (which makes an auto-sizing blue text clip) and then resizing with the normal select tool, then that makes text objects which precisely fill their blue text clips, which libsvg can handle okay. -- Finlay McWalter | Talk 10:56, 18 May 2007 (UTC)[reply]
Incidentally, it looks like you're hoping to reproduce a copyrighted logo (from The Chaser's War on Everything). Even though you've drawn the SVG yourself, that SVG would still be a derivative-work of the original, so its use on Wikipedia would have to be under the fair-use doctrine - it's not GFDL. And that, unfortunately, would mean it couldn't be used in a userbox ;( Perhaps just the text "WAR" would be sufficient, and shouldn't cause a copyright concern. -- Finlay McWalter | Talk 19:07, 18 May 2007 (UTC)[reply]

You may want to convert your text objects to paths. You can't edit them as text after that, but it helps ensure they will display the same anywhere, even on systems that don't have the same fonts. rspeer / ɹəədsɹ 17:52, 18 May 2007 (UTC)[reply]

Try selecting the text and clicking Unflow from the Text menu. --h2g2bob (talk) 19:56, 18 May 2007 (UTC)[reply]

Well this is the least of my problems now, it kinda blanked my hard drive when trying to dual boot SUSE and Ubuntu. I've got a backup of my important file on my thumb drive, but this image was wiped. --Lwarf 10:23, 19 May 2007 (UTC)[reply]

cd/dvd piracy edit

Hi. Is it possible for a computer/cd/dvd drive to detect a difference between a pressed original of software and a digital copy of that data onto a recordable cd/dvd? (without using some extra method of authentication eg online registration). ie are there features on a pressed optical disc that can not be replicated on a recordable disc? Thanks87.102.47.231 16:47, 18 May 2007 (UTC)[reply]

Sort of. There are several tricks which can be used, like marking sections of the disk as "bad sections" and seeing whether they still exist - this would catch some types of copying. The disk could also check items like the disk serial number. Also, you can't (normally) burn dual-layered disks, so the disk capacity of rewritable DVDs is effectively half (4GB single layer / 8GB dual layer).
The normal way to get around such restrictions is to have the DVD data on your hard drive and use virtualization to pretend that data is in a DVD drive (see also virtual machine). Content providers (ie the publisher) can try to detect this by running copy-protection software on your PC, like starforce does. My knowledge on this is limited, but I think it modifies the device drivers for the CD/DVD drive, effectively rooting the machine. --h2g2bob (talk) 19:50, 18 May 2007 (UTC)[reply]
There are many ways to protect the disc from copying, but pirates always win :) Some of the copy protection schemes are Starforce, SecuROM, SafeDisc and so on, with their opponents being disk image emulators like Alcohol 120% and Daemon Tools among others. After some time copy protection programs started checking for the presence of the disk emulation software which was countered by the programs (like sd4hide) that hide the presence of above mentioned, disk emulation software. It is worth mentioning that some copy protection programs can seriously damage or infect with viruses your computer. Shinhan 14:53, 19 May 2007 (UTC)[reply]
link fixed --cesarb 16:03, 19 May 2007 (UTC)[reply]
So much talking, but the answer is simply: yes! See Absolute Time in Pregroove.

How to improve a download/install process edit

There's a program called NVDA (Non-Visual Desktop Access) which is a screen reader to enable the blind to use a computer. It's freeware, so could be quite useful. However, it's incredibly difficult to install. First off, it comes as a ZIP file which is not self-extracting. Then you have to change which directory it stores into, at least with my ZIP extraction software, or it will put it in a temporary directory. Then it doesn't have an install wizard, but rather you must navigate to NVDA.exe and run it to do the install. Then it doesn't provide a desktop icon/sytsem tray icon or submenu under Start + Programs, so I had to create a shortcut icon, move it to the Desktop and rename it. I've written up a 22 step install process here: User:StuRat/NVDA. My question is, how can I convert this freeware to run with a single install wizard ? I'd like a blind, elderly person, who has never used a computer before, to be able to do the download and install it, themself, if possible. I realize that's rather ambitious, though, so would settle for something a sighted family member can install. Another option might be to put it on CD, already installed, and distribute it that way. StuRat 16:51, 18 May 2007 (UTC)[reply]

I guess you could unzip the stuff yourself and then package it up with NSIS. It looks like all the NSIS installer has to do is unzip and create a shortcut, which is well within its capabilities. -- Finlay McWalter | Talk 17:09, 18 May 2007 (UTC)[reply]
I think there's a step between the unzip and shortcut. It has to "install", which requires navigating to NVDA.exe, clicking on it, and letting it run. I'm not quite sure what it's doing at that point, perhaps changing the system registry and/or linking. Can NSIS handle that part, too ? StuRat 18:33, 18 May 2007 (UTC)[reply]
This page describes how - I think you'd just need to call ExecWait in particular. -- Finlay McWalter | Talk 18:37, 18 May 2007 (UTC)[reply]
And if that inner-installer needs user input, you can (apparently, I've not done it myself) use FindWindow and SendMessage to click buttons and type stuff automagically into that window. I have seen installers that do this, and is a fairly uncanny poltergeist-like experience. -- Finlay McWalter | Talk 18:45, 18 May 2007 (UTC)[reply]
One complication is that the process may be slightly different on different Windows versions, using different browsers, different download managers, and different unzip utilities. How are all these differences addressed by NSIS ? StuRat 22:20, 18 May 2007 (UTC)[reply]
A few things are solved by doing stuff relative to environment variables like %systemroot% (rather than C:\WINDOWS). But beyond that you'll be writing a more sophisticated install script (in NSIS's scripting language) that sniffs around for things and does the appropriate thing (generally you'll figure out what is installed, and where, by looking for their respective entries in the windows registry). You don't need to worry about unzip utilities because NSIS has its own compress/uncompress stuff - it'll compress the files when you build the installer, and will decompress them (to destinations you specify) when the uninstaller runs. -- Finlay McWalter | Talk 22:48, 18 May 2007 (UTC)[reply]
Thanks. StuRat 17:05, 19 May 2007 (UTC)[reply]

If this is to be done only once, the best way will be that you'll go there and do it by hand, especially to see any unexpected behaviour. If this should be done repeatadly, much work is to be done. My approach would be to write a command line script (a batch file) to do this. As I'm not into windos, I'd install cygwin (including perl) and do it with a unix-like shell script. At least, "navigating to NVDA.exe and clicking on it" would simply be sh -c $(find . -name 'NVDA.exe'). 84.160.252.186 19:32, 18 May 2007 (UTC)[reply]

Testing versions do have an installer so try them. If you want to get the main version wroking with NSIS, you can use their NSIS scripts. --h2g2bob (talk) 00:11, 19 May 2007 (UTC)[reply]
How bizarre that they have installers for test versions but not for the production version. StuRat 17:05, 19 May 2007 (UTC)[reply]

I assume this is dealing with Windows? You could do something like this with a Package Manager, so you'd only have to add a repository and update your system every so often, but that would only work on (some, I assume) linux distros (such as ubuntu). I don't know where to start on that though. -- Phoeba WrightOBJECTION! 06:20, 19 May 2007 (UTC)[reply]

This is only intended for Windows, and it could be limited to Windows XP, if necessary. StuRat 17:00, 19 May 2007 (UTC)[reply]

Historical currency exchange rates? edit

Now this is not a pure computing question but, since i can only make do with data already in computer readable format, I'm askin that question here.

Is there a source for historical currency exchange rates somewhere on the net, free, at least for private use? I'm thinking of daily exchange rates among US$, EUR, YEN and possibly more, of the recent years, going back into the past as long as possible. 84.160.252.186 20:58, 18 May 2007 (UTC)[reply]

There are several sources, though the amount of detail and the number of currencies covered tends to diminish as you go back in time. (Also, it used to be more common that currencies exchanged at more or less fixed rates, in which case daily information would be unnecessary.) Just google on "historic*" and "exchange rates". --Anonymous, May 18, 2007, 23:25 (UTC).

Mac OS X program to monitor socket connections edit

Is there a free (hopefully open source) and preferably command line that will listen to a socket connection that has already bean made by another program, and display it on screen in real time (the data that is passing through). Thanks!--Ryan 23:22, 18 May 2007 (UTC)[reply]

Wireshark? It's probably able to run from command-line. Splintercellguy 23:44, 18 May 2007 (UTC)[reply]
It has some command line stuff,[1] and it has an OSX port. And I heartily recommend it. --h2g2bob (talk) 00:04, 19 May 2007 (UTC)[reply]
tcpdump is the old-reliable for UNIX-like systems, so presumably it works on OS X. --TotoBaggins 00:42, 19 May 2007 (UTC)[reply]
Haven't yet tried wireshark, but tcpdump works well. Thanks everyone. You know what's funny? I can tell what websites my friends are on now ;)--Ryan 01:24, 19 May 2007 (UTC)[reply]
You must use this power wisely. (HHOS. Sniffing network connections, or doing other things that only root can do, is a sysadmin-like thing, and sysadmins -- good ones, anyway -- take their users' privacy seriously.) —Steve Summit (talk) 04:07, 19 May 2007 (UTC)[reply]
Don't worry. I'm not going to do anything malicious, just mess with my friends :)--Ryan 04:26, 19 May 2007 (UTC)[reply]