Moving Beyond Traditional Development

Even if you prefer not to use Google Chrome, Apple Safari, or Mozilla Firefox, chances are you are still utilizing their rendering engines.

Browsers like Vivaldi, Brave, Opera, and Microsoft Edge are built on Google’s Blink engine and Chromium open-source code. This reliance stems from the increasing complexity of the web, making browsers intricate software systems. For instance, Chromium comprises over 25 million lines of code, contributed by more than 8,100 developers.

“We’ve shifted the idea of the web from merely being a collection of pages with some interaction and animation to the browser evolving into an operating system,” states John Allsopp, a seasoned web designer and founder of the Web Directions conference.

Consequently, many browser developers have retreated from creating and maintaining their own engines. Microsoft notably abandoned its EdgeHTML engine a few years ago in favor of a Chromium-based Edge in early 2020. Similarly, Opera transitioned from its longstanding Presto engine to Chromium in 2013.

Apple further complicates the situation by mandating that all third-party browsers on iOS utilize its WebKit engine for security reasons. Even Mozilla, which upholds its Gecko engine for Firefox, must comply with WebKit on iOS. This restriction could deter developers from attempting to create their own engines.

Chris Coyier, cofounder of CodePen and creator of CSS-Tricks, argues that due to the significant head start of major browsers, developing a competitive engine would require a billion-dollar investment with uncertain returns. He believes that browser developers should concentrate on user-facing features, such as Brave’s emphasis on privacy or Vivaldi’s extensive customization options, rather than on rendering-engine enhancements.

“It’s not a worthwhile competition,” Coyier remarks via email. “A more beneficial approach is to improve the existing browsers.”

A Unique Approach to Browsing

So, why is Ekioh pursuing this endeavor? With Flow, the company aims to offer a distinct solution. Rather than competing head-to-head with major browsers, it is designing a browser that caters to specific applications where a new rendering engine can provide significant advantages.

Ekioh specializes in delivering web applications for embedded systems like connected TV boxes, smart displays, and automotive dashboards. The company believes that a feature known as multithreaded layout can dramatically enhance performance, particularly for animations and effects.

“What sets Flow apart from other browsers is its performance,” asserts Stephen Reeder from Ekioh.

For instance, consider a button that enlarges and reveals explanatory text upon hovering. On a low-power device, such animations can be challenging, especially if a single processing core handles all tasks. Flow’s architecture allows applications to utilize multiple cores on devices like the Raspberry Pi, facilitating more complex animations.

“With our browser, we can simultaneously adjust text layout, change sizes, and animate, resulting in a richer user interface,” he explains.

Flow also implements GPU rendering, where the graphics processor takes charge of rendering page elements. This approach consumes significantly less memory compared to relying on the main processor, leading to improved performance on budget devices with limited memory.

While other engines can use multithreading to manage browser tabs and some have begun exploring GPU rendering, they are not designed to apply multiple cores to a single webpage. Reeder notes that achieving this would require substantial code rewrites.

“That would necessitate a redesign of the core browser engine,” he remarks. “It’s not something you can simply add on.”

Perhaps even more critically, Ekioh’s business model for Flow bypasses conventional browser monetization challenges. Instead of scaling up and generating revenue through search partners or advertisements, it intends to license the software to electronics manufacturers, promoting it as a cost-effective way to create faster, more responsive products.

“If a product features any graphical user interface, there’s potential for HTML,” Reeder explains, referencing the foundational language of web pages. “If you can produce a device with slightly less memory and a slower processor while maintaining a similar user experience, that product will be more affordable.”

Not Yet Ready for Desktop

Ekioh anticipates that Flow will appear in commercial products later this year, but users can experiment with it on a Raspberry Pi right now. (The company recently added compatibility for earlier Raspberry Pi models.)

However, it’s not a replacement for Chrome or Firefox at this stage. The current iteration lacks support for tabs, bookmarks, or extensions and relies on keyboard navigation for basic operations like moving forward or backward. After navigating past Flow’s welcome page, users won’t even encounter an address bar.

Reeder emphasizes that the focus is on perfecting the rendering engine before considering additional user-facing features. The engine alone is a significant undertaking, and every time Ekioh tests Flow on new websites, it discovers new functionalities that need to be implemented.

“The direction this could take may very well lead to a desktop browser, but we’re not there yet,” he states.

Nostalgia for the Past

Nonetheless, the mere concept of a new rendering engine excites some experts in the web browsing field.

Rachel Nabors, a former program manager for Microsoft’s Edge browser and author of a book on web animation, notes that despite the rise of web-based applications, graphics and animations often feel secondary for browser developers. For her, Flow’s multithreaded layout and GPU rendering are refreshing innovations.

“Browser development still largely centers on the browser as a document reader,” she observes. “It’s curious that browser creators have lagged in delivering efficient graphics for the web.”

While Flow may currently be just a minor player in the browser landscape, its future remains uncertain.

Part of the intrigue lies in the broader context: with the fundamental development of web browsers largely under the control of Apple and Google, concerns arise about a potential “monoculture” that could erode the web's independent spirit. Chris Beard, former CEO of Mozilla, voiced this concern in late 2018 when Microsoft shifted from its own engine to Chromium and Blink.

“Ceding control of essential online infrastructure to a single entity is detrimental from a social, civic, and individual empowerment standpoint,” Beard asserted at the time.

In practice, this concern may be somewhat exaggerated today. As a contributor to Chromium, Microsoft now has a degree of influence over the browser’s direction, as do external entities like Igalia, which assists companies in implementing new features in major browsers.

Yet, as Brian Kardell, Igalia’s developer advocate, highlights, a diverse browser landscape still holds value. Even with ample external contributions, developing browser engines is a costly, time-consuming endeavor primarily funded by tech giants such as Apple and Google. There’s no guarantee they will maintain their interest and adequately support these efforts.

“There’s an intrinsic goodness in diversity,” he states. “For the same reason, you wouldn’t want to cultivate a single crop.”

While Flow may currently occupy a small niche in the browser ecosystem, its trajectory remains unpredictable. As web designer John Allsopp points out, Apple surprised many by launching Safari in 2003 with the lesser-known KHTML engine, rather than basing it on Firefox. An Apple engineer noted that KHTML was considerably leaner and more efficient, and Steve Jobs highlighted Safari’s speed during its debut at that year’s Macworld Expo.

That engine eventually evolved into WebKit, which became the foundation for Safari on iOS. Google later adopted WebKit for Chrome before developing its variant, Blink. Allsopp acknowledges this historical context and sees a parallel in Flow’s emphasis on performance.

“Perhaps it will unlock the web for a new category of devices that we haven’t yet explored,” he speculates.