Opera 9.62 vs Opera 10.0 alpha 1
On December 4, 2008, the first alpha version of Opera 10.0 was released on the Presto 2.2 engine. You can read more about this joyful event and the engine's innovations in an article written by fidelich . And what gives the user a new engine in numerical terms, I will try to find out under the cut.
The testbed will be my laptop. Configuration:
Processor: Intel Pentium M 1.73 GHz
RAM: 1.00 GB
Operating system: Windows XP Professional sp2
Experiments will be, you guessed it, Opera 9.62 and Opera 10.0. I will conduct testing with a set of synthetic tests taken from the site http://nontroppo.org/Before each test I’ll try to give a brief description of what the test is. Well, let's get started:
On the graph, the full page load time with images:
And now "full":
In the "full" test, the tenth opera demonstrates a performance increase of about ten percent. I think this is a good result.
Very clearly visible work Presto 2.2
In general, in my opinion, Opera Software specialists did a good job on the new browser. And the first alpha version is already a good product. I will look forward to the release ...
PS Many thanks to DYPA , without this article there could have been no.
The testbed will be my laptop. Configuration:
Processor: Intel Pentium M 1.73 GHz
RAM: 1.00 GB
Operating system: Windows XP Professional sp2
Experiments will be, you guessed it, Opera 9.62 and Opera 10.0. I will conduct testing with a set of synthetic tests taken from the site http://nontroppo.org/Before each test I’ll try to give a brief description of what the test is. Well, let's get started:
General Browser Load-Time Test
This test uses several timers to check page rendering speed. First Access - returns the time when the browser was able to access the DOM object at the top of the page. Doc Load - returns the time when JavaScript tells the browser that the document is loaded, but images and other things are still loading. Doc + Images Load - returns the standard OnLoad.On the graph, the full page load time with images:
JavaScript Progressive Raytracer
The essence of the test is that JavaScript draws and paints the image pixel by pixel, each pixel is a div . The test has 2 options, “basic” (using a DIV of three pixels) and “full” (using a DIV of one pixel). First, the results of the "basic" test:And now "full":
In the "full" test, the tenth opera demonstrates a performance increase of about ten percent. I think this is a good result.
Mesh transform
This test is taken from the Webkit Wiki, it tests the speed of the ECMAScript engine.3D cube
Another test of the ECMAScript engine. The test has 2 options with a large and a small cube. In the graph, the test results with a large cube (average time of one turn):Celtic Kane Mixed JS Test
The next JavaScript test produces a whole complex of various manipulations and gives the total execution time. Which is displayed in the graph:Core dom performance
The test shows the speed of the basic operations with the DOM:Very clearly visible work Presto 2.2
DOM Animation Tests
During the test, the browser renders the animated image. In this case, not real images are used, but pixel-by-pixel image information stored by JavaScript array and DHTML to dynamically create animations. The test was performed in four versions: using tables , canvases , and two tests using divs ( first , second ):In general, in my opinion, Opera Software specialists did a good job on the new browser. And the first alpha version is already a good product. I will look forward to the release ...
PS Many thanks to DYPA , without this article there could have been no.