Try this:

&lt;pre&gt;Xvfb :21 -screen 0 1024x768x24 +extension RANDR &amp;&lt;/pre&gt;

&lt;pre&gt;
Xvfb --help

+extension name        Enable extension
-extension name        Disable extension
&lt;/pre&gt;


Jeue&#39;s answer didn&#39;t help me. After hours I finally found the solution for my system and I think this will help other people too. I had to set the LD_LIBRARY_PATH like this: 

       export LD_LIBRARY_PATH=/usr/lib/x86_64-linux-gnu/

After that, everything worked very well, even without any &quot;-extension RANDR&quot; switch.

It seems that when this error appears it is an indication that the selenium-java plugin for maven is out-of-date. 

Changing the version in the pom.xml should fix the problem

Given that TimeUUID handily allows you to use `now()` in CQL, are there any reasons you wouldn&#39;t just go ahead and always use TimeUUID instead of plain old UUID?

Cassandra UUID vs TimeUUID benefits and disadvantages

If I convert a  character to `byte` and then back to `char`, that character mysteriously disappears and becomes something else. How is this possible?

This is the code:

	char a = &#39;&#200;&#39;;       // line 1		
 	byte b = (byte)a;   // line 2 		
  	char c = (char)b;   // line 3
  	System.out.println((char)c + &quot; &quot; + (int)c);

Until line 2 everything is fine:

- In line 1 I could print &quot;a&quot; in the console and it would show &quot;&#200;&quot;.

- In line 2 I could print &quot;b&quot; in the console and it would show -56, that is 200 because byte is signed. And 200 is &quot;&#200;&quot;. So it&#39;s still fine.

But what&#39;s wrong in line 3? &quot;c&quot; becomes something else and the program prints `?  65480`. That&#39;s something completely different.

What I should write in line 3 in order to get the correct result?

 
	
	

Byte and char conversion in Java

I&#39;m trying to run some Webdriver Maven tests on a headless server via xvfb using Chrome (Firefox also didn&#39;t work) but I&#39;m getting this error message:

    Xlib:  extension &quot;RANDR&quot; missing on display &quot;:21&quot;.

What more information should I provide and what could the problem/solution be?

Xlib: extension &quot;RANDR&quot; missing on display &quot;:21&quot;. - Trying to run headless Google Chrome

I'm trying to run some Webdriver Maven tests on a headless server via xvfb using Chrome (Firefox also didn't work) but I'm getting this error message:
<pre><code class="hljs language-vbnet">Xlib: extension "RANDR" missing on display ":21".
</code></pre>
What more information should I provide and what could the problem/solution be?

I spent quite a lot of time trying to figure out why video embedded like here:

    &lt;video height=&quot;256&quot; loop autoplay muted controls id=&quot;vid&quot;&gt;
        	 &lt;source type=&quot;video/mp4&quot; src=&quot;video_file.mp4&quot;&gt;&lt;/source&gt;
             &lt;source type=&quot;video/ogg&quot; src=&quot;video_file.ogg&quot;&gt;&lt;/source&gt;
    &lt;/video&gt;

starts playing automatically once the page is loaded in FireFox but cannot do autoplay in Webkit based browsers. This only happened on some random pages. So far I was unable to find the cause. I suspect some unclosed tags or extensive JS created by CMS editors.

Video auto play is not working in Safari and Chrome desktop browser

I am trying to disable the **User Agent Stylesheet** in chrome. How can I turn this off/disable this?

![enter image description here][1]


  [1]: http://i.stack.imgur.com/C23HS.jpg

chrome : how to turn off user agent stylesheet settings?

**Updated** for 2020, I&#39;ve [linked my article](https://dev.to/noseratio/comparing-process-working-sets-of-webview-based-windows-desktop-apps-5dkk) where I compare the memory footprints of different approaches to hosting HTML WebView in a basic Windows Desktop application:

- [MSEdge WebView2 WPF](https://github.com/noseratio/CompareWebViews/tree/main/WebView2App)
- [Chromely/CefSharp](https://github.com/noseratio/CompareWebViews/tree/main/ChromelyApp) 
- [Electron v10](https://github.com/noseratio/CompareWebViews/tree/main/ElectronApp)
- [Legacy IE11 WebBrowser WPF](https://github.com/noseratio/CompareWebViews/tree/main/IE11WebBrowserApp).

&lt;hr&gt;

Internet Explorer-based WPF WebBrowser control suffers from some [keyboard and focus issues][1] and [memory leak issues][2]. As an alternative solution to these problems, we&#39;re considering available options for hosting Chromium instead of WebBrowser control in our WPF/C# project based around HTML editing. Similar questions have been [asked][3] here previously. I&#39;ve read the answers and done my own research, **but I hope to obtain some more feedback from people who have actually used any of the following options in production-quality projects**:

[Awesomium][4] and [Awesomium.NET][5]

It looks very appropriate, but I don&#39;t like the fact the project is not open-source and the full source is not easily available. Also, it might be an overkill for our project, as off-screen rendering is not something we really depend on.

[Chromium Embedded Framework (CEF)][6] and [.NET bindings for CEF][7]

This is probably the best option currently available. The project seems to be alive and active, being currently in sync with Chrome v27. CEF3 uses Chrome multi-process architecture. It also looks like Adobe is giving it [some endorsement][8]. 

[Google&#39;s Chrome Frame][9]

While the original purpose of it was to be an HTML5 plugin for IE and Firefox, it actually works as standalone ActiveX control too, so I could wrap it for use with WPF. It exposes a [sufficient API][10] for interaction with the inner web page (`onmessage, addEventListener/removeEventListener, postMessage`).
I&#39;m aware Google is to [discontinue][11] Chrome Frame, but I assume the sources will remain in Chromium repository. It should not be difficult to update it with the latest Chromium code as we go, and we would have full control over this.

[WebKit .NET wrapper][12]

Not exactly Chromium-based and doesn&#39;t use V8 engine, so it is not really an option.

Is there any other option I might have overlooked? 

**I would greatly appreciate if someone shared her/his experience with any of the above options for a real-life, production-quality WPF project. Did you have any integration, licensing, or deployment implications?**


  [1]: https://stackoverflow.com/questions/18256886/webbrowser-control-keyboard-and-focus-behavior
  [2]: https://stackoverflow.com/questions/8302933/how-to-get-around-the-memory-leak-in-the-net-webbrowser-control
  [3]: https://stackoverflow.com/questions/790542/replacing-net-webbrowser-control-with-a-better-browser-like-chrome
  [4]: http://awesomium.com/
  [5]: http://awesomium.codeplex.com/
  [6]: https://bitbucket.org/chromiumembedded/cef
  [7]: https://github.com/cefsharp/CefSharp
  [8]: http://www.google.com/search?q=%22Chromium%20Embedded%20Framework%22%20site:adobe.com
  [9]: http://www.google.com/chromeframe
  [10]: http://src.chromium.org/svn/trunk/src/chrome_frame/chrome_tab.idl
  [11]: http://blog.chromium.org/2013/06/retiring-chrome-frame.html
  [12]: http://webkitdotnet.sourceforge.net/

Options for embedding Chromium instead of IE WebBrowser control with WPF/C#

I can get the value of a storage key with the Chrome Extension API:

&lt;!-- language: lang-js --&gt;

    chrome.storage.sync.get(&quot;someKey&quot;, function() {});

How can I get all key names that exist in the Chrome storage?

Get all keys from Chrome Storage

I&#39;ve run into a problem while attempting to use SVG marker elements in an SVG based visualization. I&#39;m adding my changes to a web application which happens to include a base tag on every page, so that any references to CSS files, javascript files, etc can be relative. 

I have some example code below which reproduces the issue. There is a line element, and a marker element defined. The marker element is referenced by the line in its &#39;marker-end&#39; attribute, via uri and id of marker. Without the base tag, the arrow displays fine. With the base tag, it is not shown. The reason is because the base tag changes the way the browser resolves urls.. even for the simple id based url specified in the marker-end attribute of the line. 

Is there any way I can get around this problem without having to remove the base tag? 

I can&#39;t really remove it because the use of it is fairly ingrained in the product I&#39;m working on. I need to support Firefox, Chrome and IE9+ for my webapp. Firefox and chrome both produce this problem. IE works fine (ie. arrow marker displays).

    &lt;html&gt;
    	&lt;head&gt;
		&lt;base href=&quot;.&quot;&gt;
		&lt;style&gt;
		.link { stroke: #999; stroke-opacity: .6; }
		marker#arrow { fill: black; }
		&lt;/style&gt;
	&lt;/head&gt;
	&lt;body&gt;
		&lt;svg width=&quot;100%&quot; height=&quot;100%&quot;&gt;
			&lt;defs&gt;
				&lt;marker id=&quot;arrow&quot; viewBox=&quot;0 -5 10 10&quot; refX=&quot;0&quot; refY=&quot;0&quot; markerWidth=&quot;20&quot; markerHeight=&quot;20&quot; orient=&quot;auto&quot;&gt;
					&lt;path d=&quot;M0,-5L10,0L0,5&quot;&gt;&lt;/path&gt;
				&lt;/marker&gt;
			&lt;/defs&gt;
			&lt;line x1=&quot;100&quot; y1=&quot;100&quot; x2=&quot;333&quot; y2=&quot;333&quot; marker-start=&quot;url(#arrow)&quot; class=&quot;link&quot;&gt;&lt;/line&gt;
		&lt;/svg&gt;
	&lt;/body&gt;
    &lt;/html&gt;

Using base tag on a page that contains SVG marker elements fails to render marker

I am running a python 2.7 script on a p2.xlarge [AWS](https://aws.amazon.com/) server through Jupyter (Ubuntu 14.04). I would like to be able to render my simulations.

Minimal working example

    import gym
    env = gym.make(&#39;CartPole-v0&#39;)
    env.reset()
    env.render()

`env.render()` makes (among other things) the following errors:

    ...
    HINT: make sure you have OpenGL install. On Ubuntu, you can run 
    &#39;apt-get install python-opengl&#39;. If you&#39;re running on a server, 
    you may need a virtual frame buffer; something like this should work: 
    &#39;xvfb-run -s \&quot;-screen 0 1400x900x24\&quot; python &lt;your_script.py&gt;&#39;&quot;)
    ...
    NoSuchDisplayException: Cannot connect to &quot;None&quot;

I would like to some how be able to see the simulations. It would be ideal if I could get it inline, but any display method would be nice.

**Edit**: This is only an issue with some environments, like classic control.

___
**Update I**

Inspired by [this](https://stackoverflow.com/questions/16726227/xvfb-failed-start-error) I tried the following, instead of the `xvfb-run -s \&quot;-screen 0 1400x900x24\&quot; python &lt;your_script.py&gt;` (which I couldn&#39;t get to work).

    xvfb-run -a jupyter notebook

Running the original script I now get instead

    GLXInfoException: pyglet requires an X server with GLX

___
**Update II**

Issue [#154](https://github.com/openai/gym/issues/152) seems relevant. I tried disabling the pop-up, and directly creating the RGB colors

    import gym
    env = gym.make(&#39;CartPole-v0&#39;)
    env.reset()

    img = env.render(mode=&#39;rgb_array&#39;, close=True)  
    print(type(img)) # &lt;--- &lt;type &#39;NoneType&#39;&gt;

    img = env.render(mode=&#39;rgb_array&#39;, close=False) # &lt;--- ERROR
    print(type(img)) 

I get `ImportError: cannot import name gl_info`.

___
**Update III**

With inspiration from @[Torxed](https://stackoverflow.com/users/929999/torxed) I tried creating a video file, and then rendering it (a fully satisfying solution).

Using the code from &#39;[Recording and uploading results](https://gym.openai.com/docs)&#39; 

    import gym
    
    env = gym.make(&#39;CartPole-v0&#39;)
    env.monitor.start(&#39;/tmp/cartpole-experiment-1&#39;, force=True)
    observation = env.reset()
    for t in range(100):
    #    env.render()
        print(observation)
        action = env.action_space.sample()
        observation, reward, done, info = env.step(action)
        if done:
            print(&quot;Episode finished after {} timesteps&quot;.format(t+1))
            break

    env.monitor.close()


I tried following your suggestions, but got `ImportError: cannot import name gl_info` from when running `env.monitor.start(...`. 

From my understanding the problem is that OpenAI uses `pyglet`, and `pyglet` &#39;needs&#39; a screen in order to compute the RGB colors of the image that is to be rendered. _It is therefore necessary to trick python to think that there is a monitor connected_

___
**Update IV**

FYI there are solutions online using bumblebee that seem to work. This should work if you have control over the server, but since AWS run in a VM I don&#39;t think you can use this.

___

**Update V**

Just if you have this problem, and don&#39;t know what to do (like me) the state of most environments are simple enough that you can create your own rendering mechanism. Not very satisfying, but.. you know.

Content Type	Original Author	Original Content on Stackoverflow
Question	Kaloyan Roussev	View Question on Stackoverflow
Solution 1 - Google Chrome	jeue	View Answer on Stackoverflow
Solution 2 - Google Chrome	KIC	View Answer on Stackoverflow
Solution 3 - Google Chrome	cw24	View Answer on Stackoverflow

Xlib: extension "RANDR" missing on display ":21". - Trying to run headless Google Chrome

Google Chrome Problem Overview

Google Chrome Solutions

Solution 1 - Google Chrome

Solution 2 - Google Chrome

Solution 3 - Google Chrome

Byte and char conversion in Java

Cassandra UUID vs TimeUUID benefits and disadvantages

Attributions