Language codes (e.g., AR or ARA for Arabic) are critical for values of XML "LANG" ("Language") to tag documents, audio, video, web pages, email, and other material in order to efficiently use locale information, grammar checking, spell checking, hyphenation, dictionaries, thesauri, terminologies, machine translation, search tools, knowledge management and visualization capabilities, and an increasing number of other tools. Language codes are also required to refer to people with certain background and/or skills, language training, material content, and intellectual property. Language codes are particularly important with the use of Unicode, where the encoding is no longer indicative of a single language. However, the current standard-ISO 639-does not cover many of the languages required by industry, academia, and government. It also has inherent conflicts (i.e., one two-letter code and two three-letter codes that in some cases cover the same languages). In addition, there is no provision for providing other needed information, such as modality (e.g., voice vs. text), register (e.g., textbook vs. chat room), orthography (e.g., Cyrillic vs. Latin script for languages that have written material in both), transcription system (possibly a subset of orthography), and time period (e.g., 16th Century French). Designation of dialect and country is also not always adequately covered or consistently defined. To provide a code that better meets user needs, ISO in coordination with W3C is working on defining a new code and architecture, as well as a possible interim standard. This panel discussion will describe the requirements, the major options under consideration, the issues and questions associated with such options, and the progress of the working group during the fall. Input from developers, managers, end users, and other audience members will be solicited on all points.