In October 2014, some 25 years after Tim Berners-Lee first sketched the outline of what would become the World Wide Web in 1989, the W3C HTML Working Group voted to pass the final “Recommendation” of the Open Web Platform standard known as HTML5. Five releases in 25 years is not a fast pace of evolution by today’s software standards, especially when you consider that the predecessor standard, HTML4 (HTML 4.01), was released in 1999.
OK, that is not totally fair, because the standardization process moved more quickly earlier and gave us HTML2 in 1994 under the aegis of the IETF, and HTML3.2 in 1997 under the W3C. What is fair to say is that HTML5 has had a fitful and somewhat tortuous path, and in its current state, two competing standards bodies appear to define it: the W3C (which is focused on the “snapshot” standard), and WHATWG (which is focused on the “living standard”).
None of this should take anything away from the importance of HTML5, because it plays an incredible power-balancing role in the industry. HTML5 is the industry’s key insurance policy against total subjugation of computing by a dominant oligopoly of platform vendors. That is, HTML5 bestows strategic safety and a semblance of control to all vendors and enterprises—the rest of us. HTML5 is the only cross-platform human-machine interaction technology managed and evolved by a broad and open industry-consensus process. It is a miracle that we have arrived at a standard. (Well, almost one standard.)
(Related: Debunking five common myths and misconceptions about HTML5)
The finalization of HTML5 is of minor relevance to everyday developers, but it is one of several trends that sets up 2015 as a pivotal year for adoption:
Interoperability and patent assurances. A final version of the HTML5 standard is important for certain enterprises and vendors that depend on interoperability guarantees bestowed by such standards. It also provides patent protections to those implementing the technology. HTML5 is an expansive and complex set of technologies with contributions from many players. The standard’s finality confers the rights for compliance with this standard without being sued.
Hybrid apps will reach parity. Browser HTML5 adoption has improved significantly in the last three years, but we are about to witness a major uplift in the impact of HTML5 on mobile applications thanks to improvements in the embedded WebView component in mobile platforms. With the release of iOS 8 and Android L, new WebView components will that implement JIT JavaScript compilers and WebGL.
Until recently, hybrid apps that ran on mobile devices had to contend with an inferior WebView, where JavaScript performed at about a quarter the speed of the Web browser on the same device. This was first remedied with Android KitKat about a year ago, and with iOS 8.x it is now effectively remedied in Apple’s mobile devices. The greatest use of HTML5 technologies on mobile devices will be found in hybrid apps.
HTML5 tools suck, but are improving. The key issue hindering HTML5 today is the availability of capable tools and skills needed to use this expansive technology effectively. We are far from where deep HTML5 skills or tools are available in the market to the same degree as Java or .NET, and we may not reach this level until the end of the decade. The arrival of the standard will feed the virtuous cycle of increased adoption and skills growth. Improved editors and debuggers are already beginning to make a dent in making developers more productive.
It’s HTML5 inside. Application development tools, frameworks and middleware aimed at cross-platform development are increasingly relying on HTML5 and JavaScript to deliver the goods. The majority of enterprises developing HTML5 applications will likely do so through third-party cross-platform application development tools that provide client-side or end-to-end abstraction models to improve development agility and productivity. Model-driven tools today rely on Web technologies internally, allowing developers to leverage HTML5 innovation while leaving its complexities to specialist platform builders.
Is JavaScript up to the task? JavaScript was designed for writing quick-and-dirty application glue in an era when browsers did not run much application logic. It is amazingly powerful given its design focus. Despite its performance issues, the greatest enterprise doubts around HTML5 are about the suitability of JavaScript for large bodies of complex application code. The lack of built-in dependency-management systems and modularity constructs, along with difficult-to-read and debug dynamic typing, are key areas of work for the ECMAScript governing body. Finalizing ECMAScript 6 by mid-2015 will be key to the success of the Web platform in the long run.
The HTML5 standard is as much a political triumph as it is a technical triumph. HTML5 holds the promise of a unified set of technologies that can provide skills, code and effort sharing across multiple devices and operating systems from all walks of tech-land. Even if the promise of writing once and running everywhere cannot be realized fully (and it is clear that it cannot), HTML5 can provide time and cost savings by sharing aspects of the code or development effort.
Al Hilwa is program director of application development software research at IDC.