Wikipedia:Reference desk/Archives/Computing/2014 March 28

Computing desk
< March 27 << Feb | March | Apr >> March 29 >
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.


March 28 edit

CrystalDiskMark - what is QD? edit

CrystalDiskMark (http://crystalmark.info/?lang=en) is a program to benchmark drive speed. It tests:

  • sequential
  • 512K
  • 4K (QD=1)
  • 4K QD32 (QD=32)

What is "QD"? Bubba73 You talkin' to me? 02:52, 28 March 2014 (UTC)[reply]

QD is queue depth, the number of 4K requests in the queue. Because of the way an OS might cluster IO requests, QD=1 may give different throughput than QD=32. In particular, if there are multiple requests in the queue, the OS may try to optimize head seeks to speed things up, see, for example, elevator algorithm. --Mark viking (talk) 03:19, 28 March 2014 (UTC)[reply]
Thanks. I was testing on an SSD and it made a big difference:

          Sequential Read :   450.777 MB/s
         Sequential Write :   417.344 MB/s
        Random Read 512KB :   418.733 MB/s
       Random Write 512KB :   384.495 MB/s
   Random Read 4KB (QD=1) :    25.948 MB/s [  6335.1 IOPS]
  Random Write 4KB (QD=1) :    89.090 MB/s [ 21750.4 IOPS]
  Random Read 4KB (QD=32) :   246.661 MB/s [ 60219.9 IOPS]
 Random Write 4KB (QD=32) :   228.627 MB/s [ 55817.2 IOPS]

Bubba73 You talkin' to me? 03:33, 28 March 2014 (UTC)[reply]

gateways edit

Hello,

I was surprised today when my motorola changed to an Arris device... I read about this and it seems like ISPs have the power to upgrade your firmware. My question is, how much power does an ISP have over my gateway. Note that I do not rent this gateway, I bought it thinking that I would be able to control stuff like this... If I am concerned about stuff like this, would buying a separate modem and dedicated router do the trick? Thanks! Brusegadi (talk) 03:58, 28 March 2014 (UTC)[reply]

Ubuntu + SSH edit

I recently decided to switch to key-based authentication for SSH on my network. After dedicating a few hours to figure out how it worked, it seemed simple enough. However I hit a snag when I tried running ssh-add on my Ubuntu 13.10 install. While I could get it to work, I could never get it to retain my key/passphrase information across reboots.

After combing through some internet forums, I found various "solutions" but none of them really seemed to address the problem. Adding identity files in my ~/.ssh/config is not a solution in my mind (it makes key management messy and does not scale well). No shell scripts I found will work either as all my keys are protected by passphrases. I suppose if I must, I will remake keys with no passphrases, but I would prefer not to.

What is causing this issue and how can I get the ssh-agent to work properly and retain my keys across reboots?

I have some clues that it might have something to do with Unity's keyring implementation, but I do not know how that works. I will happily disable it if possible though.

Thanks in advance. -Amordea (talk) 11:43, 28 March 2014 (UTC)[reply]

Nevermind. After a few more complications with Unity, I've just decided the Unity interface is not worth mucking with. I will just use a different distribution for this project. -Amordea (talk) 15:29, 28 March 2014 (UTC)[reply]
It's not supposed to retain your passphrase across reboots. It keeps the secret key in RAM for security (don't want it on the hard drive where someone could recover it). So when you reboot, you have to type the passphrase again. That is intentional. Maybe what you really want is a hardware security token, if you never want to re-enter your passphrase. You should probably use a passphrase anyway though, per the principle of two-factor authentication. 70.36.142.114 (talk) 22:27, 28 March 2014 (UTC)[reply]
It can if you place it in your key-ring. This causes the passphrase to be encrypted and stored on disk. It gets decrypted upon login. I have been able to do this in Ubuntu using Unity, but I have not sshd anywhere in ~ 6 months... Brusegadi (talk) 23:27, 28 March 2014 (UTC)[reply]
On my Lubuntu install, it does. Which I actually thought rather strange since ssh-add doesn't work if I don't input the passphrase. I don't actually mind typing the passphrase each time I log in. What I want to avoid is having to type ssh -i path/keyname user@host every time, when I should only have to type ssh user@host. A few extra keyclicks, I know, my poor fingers! But I don't really see what the point of the ssh-agent is if it only works during a single login session. Or maybe there is a thing I'm still not understanding about the ssh-agent? I'd welcome some more insight on SSH best practices if anyone's wanting to share them. -Amordea (talk) 11:05, 29 March 2014 (UTC)[reply]

real time scatterplot utility edit

Dear Computing desk, I'm in the process of writing a fitting program that dumps its output into a CSV every X number of iterations. Does anyone have any recommendations for a program that could replot the output in real time? Right now I'm just looking at the RMSD calculated after every step, but I'd much rather have a graph that would automatically refresh after every refinement step. Preferably something X11 friendly, but I'm not really picky. It's not at all a crucial feature, more decorative than anything. Thanks, (+)H3N-Protein\Chemist-CO2(-) 13:52, 28 March 2014 (UTC)[reply]

I can think of 3 ways to run the plot program periodically:
1) The program which saves the data can also kick off a run of the plotting program, after each file save. This is likely the best option. A system call from the main program can probably accomplish this.
2) The plotting program remains running constantly, and can monitor the file, and plot more data as it arrives. This is a bit trickier, as you now have two programs trying to access the same file at the same time, and might be more CPU intensive, as you have to keep checking for updates to the file. It might work better to add each iteration of data as a new file with a timestamp on it, then the plotting program can detect any new file in that directory and plot it.
3) The plotting program can just have a time delay in it and replot at those intervals. You might still run into file contention problems, though, and you might sometimes get hiccups where the same data is plotted twice or you skip plotting one iteration, making the results look rather haphazard. Weather maps in motion often have this jumpy look. StuRat (talk) 14:32, 28 March 2014 (UTC)[reply]
Yes, number 2 is how I was thinking of doing it. Possibly number 3. I was mainly looking for recommendations for plotting programs. (+)H3N-Protein\Chemist-CO2(-) 20:11, 28 March 2014 (UTC)[reply]
@Protein Chemist: Depending on your scripting skillz and time available, I'm pretty sure Gnuplot would work. Definitely free/X11 friendly, not necessarily user friendly :) This would work with Stu's option 1) and 3) I think, not sure about 2), but probably... SemanticMantis (talk) 20:12, 28 March 2014 (UTC)[reply]
Or if you're already familiar with Python (or MATLAB), check out Matplotlib. SemanticMantis (talk) 20:15, 28 March 2014 (UTC)[reply]
Thanks. Gnuplot was the first option I looked into, but I didn't really have the time to play around with it. Guess it's probably worth giving it another go, nothing to lose. (+)H3N-Protein\Chemist-CO2(-) 18:01, 31 March 2014 (UTC)[reply]

Hello. I'm a student and I have dyslexia, so I'm using finnish audiobook library Celia. They have audiobooks downloadable from Internet using Daisy format. The problem is that they are recommending only expensive Daisy CD-players and programs for Windows. I have a Mac computer and not money for Daisy CD-player. I found free program named Olearia for Mac, but I think this program is quite poor, because it could not save bookmarks or search beginning of a chapter. The audio of the books are divided in 45 minutes mp3 files, so the program thinks that chapters begin in the beginning of the file. Could someone recommend some better audiobook listening program or maybe a mp3-player, which supports Daisy format?

Yours, 178.55.172.60 (talk) 14:12, 28 March 2014 (UTC)[reply]

Pagination appearing only on Homepage edit

Hi, I am facing a new problem in my blog. The problem is pagination is appearing only on Homepage..! While going through the pages it's not appearing..! Here is the HTML code of pagination I'm currently using:

/* Page Navigation styles */
.showpageNum a, .showpage a {
  box-shadow: 3px 3px 2px #888888;
  background: #ffffff;
  padding:0 8px; 
  margin:0 4px; 
  text-decoration: none;
  color:#000000; 
  border:1px solid #000000;
  border-radius:3px;
  -webkit-border-radius:3px;
  -moz-border-radius:3px;
}
 
.showpageOf {
  margin:0 8px 0 0;
}
.showpageNum a:hover, .showpage a:hover {
  box-shadow: 3px 3px 2px #888888;
  color:#ffffff;
  background: #000000;
  border:1px outset #000000;
  border-radius:3px;
  -webkit-border-radius:3px;
  -moz-border-radius:3px;
  }
 
.showpagePoint, .showpage a {
  box-shadow: 3px 3px 2px #888888;
  color: #000000;
  background: #000000;
  padding: 0 8px;
  margin: 0 4px; 
  text-decoration: none;
  color:#ffffff;
  border:1px solid #000000;
  border-radius:3px;
  -webkit-border-radius:3px;
  -moz-border-radius:3px;
}

/*...........................................................*/

<!--Page Navigation Starts-->
<script type='text/javascript'>
var home_page="/";
var urlactivepage=location.href;
var postperpage=6;
var numshowpage=6;
var upPageWord =&#39;Previous&#39;;
var downPageWord =&#39;Next&#39;;
</script><script src='http://mybloggertricksorg.googlecode.com/files/colorizetemplates-pagenav.js' type='text/javascript'/>
<!--Page Navigation Ends -->

Can anyone please tell, what change should I need to make in the above code for appearing pagination in all pages, including homepage?--Joseph 14:50, 28 March 2014 (UTC)[reply]

consolidating mutliple pdfs into one document by name edit

Hey all. (Mac user). I have Adobe Acrobat Professional. I was wondering if there is any way to compile a pdf in one shot of multiple items in a folder that are already, by name, in the order I would want them to be consolidated. I already know how to do it the slow way (document, insert pages and attach each one by one) but it's not very efficient. Here's an example of the contents of a folder:

Page 1.pdf
Page 2.pdf
Page 3.pdf
Page 4.pdf
Page 5.pdf
Page 6.pdf
Page 7.pdf
Page 8.pdf
Page 9.pdf
Page 10.pdf
Page 11.pdf
Page 12.pdf

And I want to create a single pdf of these documents in page order. Anyone know of a way? Thanks.--71.167.166.18 (talk) 14:50, 28 March 2014 (UTC)[reply]

I've done this with Ghostscript,an free open-source tool. this article describes a number of techniques, starting with the Ghostscript method. -- Q Chris (talk) 16:05, 28 March 2014 (UTC)[reply]
Hi, you can combine PDF files, if you've Adobe Photoshop installed in your system. Here is how to combine PDF files in Photoshop. Hope this will help.--Joseph 16:45, 28 March 2014 (UTC)[reply]
You can do it with pdftk by running this bash script:
IFS='                                                                                                               
'
pdftk $(ls Page* | sort	-g -k2) cat output out.pdf
-- Finlay McWalterTalk 17:48, 28 March 2014 (UTC)[reply]
Yes, Acrobat Pro can do this: on the splash screen that comes up when you open Acrobat, the last item under Select a Task should be "Combine Files into PDF". --Canley (talk) 09:17, 29 March 2014 (UTC)[reply]

Does Windows use X deep down in its display code? edit

20.137.2.50 (talk) 17:18, 28 March 2014 (UTC)[reply]

No, I don't think so. See X_Window_System and Windows_operating_system. If MSWin contained any code from X project, they would have to include a copy of the source, as well as a copy of the MIT_License. I haven't used MSWin in years, but I'm pretty sure they've never done any such thing, that's not really how MS does business. I did a quick check for references, but it's hard to find people saying e.g. "Win95 does not use any code from X window system", because it's kind of obvious. A related example: even though OS_X is built on Unix, Apple still developed Quartz_(graphics_layer) for native windowing, though there is a version of X11 that Apple distributed with later versions of OSX, specifically to make it easier for certain Unix programs to be ported to OSX. SemanticMantis (talk) 20:20, 28 March 2014 (UTC)[reply]
No, not at all. An outline of Windows display architecture is in this diagram (see the Direct3D System Integration section). Our articles in this space include Graphics Device Interface, Direct X, and Direct 3D. -- Finlay McWalterTalk 20:12, 28 March 2014 (UTC)[reply]
I assume the implementation of an X server (call it A) or whatever Microsoft's display controller module is (call it B) is at the layer above the hardware level and that a command to make the pixel at position (x,y) have such-and-such rgb values is accomplished by A or B by writing 1s and 0s to the graphic card's memory accordingly. Where can I read about that specification, the graphics card standard that lets A and B implementations know what bytes to put at what memory locations in order to make a certain pixel a certain color? 75.75.42.89 (talk) 20:51, 28 March 2014 (UTC)[reply]
Some (e.g. WiredX) were simply GDI programs which drew Xprotocol calls into a GDI window. Fancier ones, which aim to support X extensions for things like transparency, and OpenGL, will have to be built on Direct X. A decent modern example is Xming, which is open source, so you can examine the docs that come with its source. -- Finlay McWalterTalk 21:01, 28 March 2014 (UTC)[reply]