WebAnywhere: A Screen Reading Interface for the Web on Any Computer
Jeffrey P. Bigham, Craig M. Prince, Sangyun Hahn, Richard E. Ladner · 2008 · Proceedings of the 2008 International Cross-Disciplinary Conference on Web Accessibility (W4A) · doi:10.1145/1368044.1368075
Summary
This short paper introduces WebAnywhere, a free web-based screen reader from the University of Washington that enables blind users to access the web from any computer with a standard browser and sound capability, without installing any software. The system addresses a critical gap: while web-enabled computers are ubiquitous in libraries, internet cafes, airports, hotels, and schools, screen readers are installed on virtually none of them. Commercial screen readers like JAWS and Window-Eyes were expensive, and even free alternatives like NVDA and Fire Vox required installation permissions that public terminals rarely grant. Accessible mobile devices cost \,000-5,000. WebAnywhere runs entirely as a web page with three components: client-side JavaScript that handles user interaction and coordinates speech; a server-side text-to-speech engine with caching; and a server-side web proxy that overcomes cross-site scripting restrictions by making external pages appear to come from the local server. Audio playback used the SoundManager 2 Flash object (installed on 98.8% of desktops at the time). The system supported rich navigation including browsing by paragraph, sentence, word, or character, and quick jumps between headings, form elements, links, and table rows/columns using standard keyboard shortcuts. Form input worked without requiring a separate forms mode.
Key findings
WebAnywhere required less than 100 kB to load and became usable in seconds, compared to over 10 MB and minutes for the closest alternative (Serotek System Access To Go). A survey of 15 public computer terminals in the Seattle area found that 14 could run WebAnywhere (9 required headphones); the only failure was a malfunctioning sound card. More than half of these terminals did not allow software installation. Sound latency for retrieving new multi-word speech was under 300 ms on high-bandwidth connections, and a predictive prefetching strategy reduced latency by nearly 20% by anticipating what users would request next. In a user evaluation with 8 blind participants (4 female), all successfully completed four real-world tasks: checking a Gmail account, finding a bus arrival time, looking up a restaurant phone number, and completing a survey. No participants raised concerns about system responsiveness. The authors also noted WebAnywhere's value for web developers, citing research showing that developers create more accessible pages when they review them with a screen reader — WebAnywhere provided a free, instant mechanism to do so.
Relevance
WebAnywhere represented a pioneering approach to making assistive technology itself more accessible and portable, reframing screen reading from an installed desktop application to a cloud service. The core problem it addressed — that blind users cannot access computers where screen readers are not installed — connected accessibility directly to digital equity: people who could not afford Windows plus a commercial screen reader were doubly excluded. While the specific technology has been overtaken by built-in screen readers on modern platforms (VoiceOver on macOS/iOS, TalkBack on Android, Narrator on Windows) and the decline of Flash, the architectural pattern of web-based assistive technology influenced subsequent cloud-based accessibility tools. The concept of a zero-install screen reader also anticipated the shift toward browser-based computing that has only accelerated. For practitioners, the paper demonstrates that accessibility barriers often compound — the challenge is not just making web content accessible, but ensuring the tools needed to access it are themselves universally available. Jeffrey Bigham went on to become one of the most influential accessibility researchers, co-founding accessibility research initiatives at Carnegie Mellon.
Tags: screen readers · blindness · web-based assistive technology · text-to-speech · digital equity · cloud computing · JavaScript · proxy server · portable accessibility