Skip to main content
  1. Home
  2. Computing
  3. News

Google Chrome extensions are failing, and $8,000 is on the table for a fix

Add as a preferred source on Google

There seems to be some mysterious problem affecting certain Chrome extensions, but it’s intermittent enough that it hasn’t yet been solved. The problem is annoying enough that one developer has posted two $4,000 bug bounties and created an Upwork job listing that pays up to $150 per hour. These incentives might inspire others to help track down and fix the bug.

First spotted by TechRadar and described in detail in a blog post written by Jói Sigurdsson, founder and CEO of the CrankWheel screen-sharing extension for the Google Chrome browser, the bug is related to a failure to trigger an action when the extension’s icon is clicked on the toolbar. Since this is frequently how an extension is used, it’s a crippling error. Unfortunately, the problem is difficult to recreate and is estimated to impact only 3% to 5% of those that have affected extensions installed.

A mouse pointer hovering over the CrankWheel Chrome Eextension.
Image used with permission by copyright holder

Sigurdsson posted the bug to the Chromium Bug Tracker in April 2022 and got a confirmation from another person in June 2022 that the problem was affecting other extensions, some quite popular. Two password managers, LastPass and Norton, apparently suffer from the bug, as well as Google’s own Picture in Picture extension and the Screencastify screen recorder. This places the number of people potentially experiencing the problem into the millions.

Recommended Videos

Tracking down elusive and intermittent bugs can be quite difficult, so for anyone with sufficient programming skills and experience in developing Chrome extensions, there’s a nice challenge waiting. The first $4,000 bug bounty is for being able to consistently (90% or more) reproduce the issue. The second $4,000 bounty will be paid for a confirmed fix that is accepted by the Chromium team. This is in addition to the hourly fee being offered in Upwork.

Alan Truly
Alan Truly is a Writer at Digital Trends, covering computers, laptops, hardware, software, and accessories that stand out as…
The next MacBook Air could arrive early next year, but OLED will have to wait
Apple’s next MacBook Air may look familiar on the outside, with M6 expected to be the main upgrade.
Computer, Electronics, Laptop

Apple may be about to give the MacBook Pro one of its biggest upgrades in years, but the MacBook Air apparently has to wait a little longer for its turn.

According to Bloomberg’s Mark Gurman, Apple is planning to refresh the MacBook Air in early 2027. The upcoming model is expected to stick with a non-OLED display, even as Apple prepares to move its higher-end MacBook Pro lineup to OLED and touch support.

Read more
Apple wants touchscreens on Macs, but its MacBook lineup could get weird
Between entry-level Pro models, high-performance machines, and future touchscreen versions, Apple's laptop lineup risks becoming more complicated than it needs to be.
Apple MacBook Air Pro logo sunlight

Apple's MacBook lineup could soon resemble a restaurant menu with 17 different ways to order a burger. The company is expanding its laptop range across more price points, performance levels, and display technologies, with OLED touchscreens potentially bringing another premium tier into the mix. That's great news for buyers who want options, but Apple's product lineup could use a little decluttering before things get silly.

Apple's MacBook lineup has a growing identity problem

Read more
Moonshot tells Kimi AI to code away, but leave the hacking alone
Moonshot is clipping Kimi AI’s hacking wings before letting it loose
Moonshot AI

Chinese AI startup Moonshot AI is deliberately limiting the hacking capabilities of its latest artificial intelligence model, highlighting a growing dilemma for the industry: how do companies make AI more powerful without making cyberattacks easier to carry out?

The company introduced Kimi K3.1 Pro on October 24, describing it as a highly capable, open-weight model with around 3 trillion parameters. Unlike conventional AI services that operate entirely through a provider's servers, open-weight models allow developers to download the underlying model files, run them on their own systems, and modify how they work.

Read more