Search Results for: webxprt

Mystery solved

on July 2, 2015

As we mentioned a few weeks ago, the WebXPRT Local Notes test would not complete on recent builds of Windows 10 when using the Edge browser. Other browsers complete WebXPRT in recent Windows 10 builds without any problems.

We now know what is causing this behavior. The Local Notes test stores encrypted content in LocalStorage as UTF-16 character encoded Unicode strings. The encrypted content included values that are not considered valid characters in certain use cases. The current Edge implementation treats these characters as undefined and cannot store them. Other browsers may not have had an issue with the characters because of differences in the way they implement LocalStorage.

We’ve been able to work around this by using escape sequences for unsupported Unicode code points. Testing so far has not shown any perceptible change in results, so we believe that we will able to make this change to WebXPRT without compromising the comparability of the results.

Because this issue affects both WebXPRT 2013 and WebXPRT 2015, we’re planning to update both versions. We’ll let you know as soon as they are available.

If you’d like more details about this issue and the fix, please let us know.

Eric

Posted in Performance benchmarking, WebXPRT, WebXPRT 2015, What makes a good benchmark?, Windows 10 |

Bit by bit

By Eric Hale

on June 11, 2015

We’ve been working to internationalize the XPRTS. Our initial attempts have focused on China.

Both BatteryXPRT and WebXPRT have Chinese UI options. We expect to have a version of MobileXPRT with a Chinese UI option available in a couple of weeks.

We’ve also been working to make the benchmarks more accessible in China. WebXPRT has a mirror host site in Singapore. We’re also getting the XPRTs into Chinese app stores. MobileXPRT is available in two Chinese app stores: Xiaomi (http://app.mi.com/detail/90862) and Zhushou 360 (http://zhushou.360.cn/detail/index/soft_id/2984653). We aim to have BatteryXPRT and the Chinese version of MobileXPRT available in those stores as quickly as possible.

Obviously, we will continue to work to improve our localization. This is an area where we can use the help of the community. If you have the translation skills and want to contribute the strings for a UI for your language, let us know.

Eric

Posted in BenchmarkXPRT, MobileXPRT, WebXPRT |

Rolling with the changes

By Eric Hale

on June 4, 2015

While WebXPRT 2015 has been running fine on earlier beta versions of Windows 10, we have found a problem on some recent versions. Starting with build 10122, the Local Notes test hangs when using the Microsoft Edge browser. (Note: This browser still identifies itself as Spartan in the builds we have seen.) Chrome and Firefox on Windows 10 have successfully run WebXPRT 2015, so the problem appears to be restricted to Edge/Spartan.

Because WebXPRT ran successfully on earlier builds of Windows 10, we are hoping that upcoming builds will resolve the problem. However, we have been investigating the issue in case there is something we can address. The problem is that the encrypted strings that the test is trying to write to LocalStorage are not being written correctly. Non-encrypted strings are being written correctly.

As soon as the problem gets resolved, we’ll let you know.

In other news, we’ve been looking at Android M. There are a lot of interesting changes coming, such as the difference in the way that Android M manages app permissions. We’ve decided to delay releasing the design document for the next version of MobileXPRT so that we can make sure that the design deals with these changes appropriately.

Eric

Posted in Android, Benchmarking, Chrome OS, Collaborative benchmark development, MobileXPRT, WebXPRT, WebXPRT 2015, Windows 10 |

What makes a good benchmark?

By Eric Hale

on May 28, 2015

As we discussed recently, we’re working on the design document for the next version of MobileXPRT, and we’re really interested in any ideas you may have. However, we haven’t talked much about what makes for a good benchmark test.

The things we measure need to be quantifiable. A reviewer can talk about the realism of game play, or the innovative look of a game, and those are valid observations. However, it is difficult to convert those kinds of subjective impressions to numbers.

The things we measure must also be repeatable. For example, the response time for an online service may depend on the time of day, number of people using the service at the time, network load, and other factors that change over time. You can measure the responsiveness of such services, but doing so requires repeating the test enough times under enough different circumstances to get a representative sample.

The possible things we can measure go beyond the speed of the device to include things such as battery life and compatibility with standards, and even fidelity or quality such as with photos or video. BatteryXPRT and CrXPRT test battery life, while the HTML5 tests in WebXPRT are among those that test compatibility. We are currently looking into quality metrics for possible future tools.

I hope this has given you some ideas. If so, let us know!

Eric

Posted in Benchmarking, Benchmarks in general, BenchmarkXPRT development community, MobileXPRT |

It’s not the same

By Eric Hale

on May 21, 2015

We sometimes get questions about comparing results from older versions of benchmarks to the current version. Unfortunately, it’s never safe to compare the results from different versions of benchmarks. This principle has been around much longer than the XPRTs. A major update will use different workloads and test data, and will probably be built with updated or different tools.

To avoid confusion, we rescale the results every time we release a new version of an existing benchmark. By making the results significantly different, we hope to reduce the likelihood that results from two different versions will get mixed together.

As an example, we scaled the results from WebXPRT 2015 to be significantly lower than those from WebXPRT 2013. Here are some scores from the published results for WebXPRT 2013 and WebXPRT 2015.

Please note that the results above are not necessarily from the same device configurations, and are meant only to illustrate the difference in results between the two versions of WebXPRT.

If you have any questions, please let us know.

Eric

Posted in BenchmarkXPRT, Collaborative benchmark development, WebXPRT, WebXPRT 2013 results, WebXPRT 2015, What makes a good benchmark? |

Looking at the data

By Eric Hale

on April 30, 2015

We’re planning the general release of WebXPRT 2015 for late next week. The testing is looking good and the response has been positive.

We’ve been looking at the hundreds of runs in the database for the community preview. As we’ve said before, we’ve been looking at the information from the JavaScript navigator object in the hope that we could improve the disclosure information from WebXPRT 2015. However, the information is not reliable enough for us to use it at this time. Hopefully, that will improve in the near future.

For now, we will continue to use the information from the user agent string. We’ve discussed the user agent string before. It does give us some information about the device, although not as much as we are able to gather in some of the XPRTs.

Looking at the data, the most common OS has been Windows. This may be in part because you needed to be logged in to run the CP. However, Android devices represented over a third of the runs, and Chrome OS represented about 25 percent of the runs. While we had healthy numbers of iOS devices, there were only a handful of Mac OSX runs. Chrome was the most common browser. Other browsers identified themselves as Safari, Firefox, Opera, and MS IE.

As you can see, the new WebXPRT continues the tradition that WebXPRT 2013 started of running everywhere. We can’t wait to make it available to general public!

Eric

Posted in BenchmarkXPRT development community, Community Preview, WebXPRT, WebXPRT 2013 release, WebXPRT 2015 |

Search Results for: webxprt

Mystery solved

Bit by bit

Rolling with the changes

What makes a good benchmark?

It’s not the same

Looking at the data

Check out the other XPRTs: