W3C publishes draft guidance on consistent text matching across the web

The World Wide Web Consortium (W3C) published the first public working draft of Character Model for the World Wide Web: String Matching on 16 July 2026.

Prepared by the Internationalization Working Group, the document provides guidance for technical standards authors, software developers and content creators. Its aim is to make text comparison more consistent across web services, languages and writing systems.

Computers compare text using the underlying digital values assigned to each character. However, two pieces of text that look the same to a person can sometimes be stored differently and may therefore be treated as different by software.

For example, an accented letter can be stored as a single character or as a base letter followed by a separate accent. Uppercase and lowercase letters, invisible characters and different ways of representing the same symbol can also affect whether two strings match.

The draft explains how technical standards should address these differences through consistent character conversion, case-insensitive comparison and Unicode normalisation. Normalisation converts certain alternative representations of the same character into a common form before they are compared.

The document also considers characters that look identical but belong to different writing systems. For example, some Latin, Greek and Cyrillic letters may look alike while remaining separate characters. Such similarities can cause errors and may also be used to create misleading names or identifiers.

The guidance is mainly intended for people developing web standards and software, but it may also help content developers understand why searches, identifiers, passwords or document references sometimes fail to match as expected.

The publication is an initial draft on the W3C Recommendation track. It is not a final web standard and may be revised or replaced following public comments submitted through the project’s GitHub repository.

What does this mean?

In practical terms, the draft is intended to help different websites and applications reach the same result when comparing text. A name typed on one device should not be rejected on another simply because the accented letters were stored differently.

It could improve the reliability of searches, links, document formats, programming identifiers and other systems that depend on exact text matching. It is particularly relevant for languages and writing systems that use accented letters, combined characters or scripts with more complex character rules.