Google lets PC users search the web using voice in Chrome

The latest beta release of the Chrome browser for Linux, Mac and Windows introduces support for voice commands, as well as improved support for responsive design for web developers.
Written by Nick Heath, Contributor

Google has added support for searching the web by voice to the latest beta build of its Chrome browser on PCs.

Users of the Chrome Beta on Linux, Mac and Windows will be able to open a new tab or visit Google.com and say "OK Google", followed by what they want to search for.

Aside from search Google said voice commands could also be used to set a timer, "OK Google, set a timer for 30 minutes", or create a reminder for its Google Now service, "OK Google, remind me to pick up dessert at 6pm tonight".

The voice command feature has to be manually enabled by clicking Enable "Ok Google" on the google.com page.

Here is a demo of the system in action:

Google has already added voice control to its services on other platforms. Owners of phones and tablets running Android 4.4 KitKat can also use "OK Google" for web search and Google Now commands.

The Chrome desktop beta also adds the ability for the browser to import existing supervised user profiles, which allow a third party to control which sites a user can visit.

Developer changes: Responsive images and web audio API updates

For developers, the beta introduces a new HTML attribute designed to make it easier to serve images that suit the device running the browser.

Because web content is served to a variety of different devices — phones, tablets, laptops, desktops, TVs — with varying screen sizes and network constraints, it can be tricky to serve images in an appropriate size. Serving the wrong size image can result in slower page loading, excess bandwidth consumption and improper formatting.

By adding the new srcset attribute to the img element Chrome allows developers to point the browser to the same image stored in a variety of resolutions and have it serve the resolution that better matches the device browsing the web.

Google provides this example:

<img alt="A rad wolf." src="pic1x.jpg" srcset="pic1x.jpg 1x, pic2x.jpg 2x, pic4x.jpg 4x">

As seen here, srcset takes a comma-separated list of URLs each pointing to a different resolution version of the image. After each URL is a value relating to what screen size or pixel density a device should have for it to be served this version of the image. In this example these values are 1x, 2x and 4x — measures relating to screen pixel density.

The release also removes the need to use the webkit prefix when calling the Web Audio JavaScript API, a high-level JavaScript API for processing and synthesising audio in web applications. The move will offer cross-browser compatibility with Firefox for scripts using the Web Audio API.

Starting with this release, the unprefixed API entry points audioContext and offlineAudioContext will be available. Google recommends switching to the unprefixed versions soon, as the prefixed versions will be removed in a future release.

Other platform changes include the introduction of the font-variant-ligature CSS property to allow developers to control ligatures in text.

Google has also removed a variety of infrequently used web platform features, listed here. Chrome will also now offer to remember and fill password fields in the presence of autocomplete=off.

A complete list of Chrome's developer features is available at visit chromestatus.com/features.

More on Chrome

Editorial standards