What is the issue with the Infra Standard?

TL;DR: Of the terms ‘user agent’ and ‘web browser’, exactly one term is defined, yet both of these terms are used extensively and inconsistently; sometimes they are explicitly treated differently; sometimes they seem to be interchangeable.

This issue discusses use of the terms ‘user agent’ and ‘web browser’ across all WHATWG specifications.

The term ‘user agent’ is defined in Infra as shown here:

A user agent is any software entity that acts on behalf of a user, for example by retrieving and rendering web content and facilitating end user interaction with it.

This example sounds a lot like a ‘web browser’, although that term does not seem to be defined anywhere.

The term ‘user agent’ is so important that the very same paragraph that defines it includes cross-references:

A person can use many different user agents in their day-to-day life, including by configuring an implementation to act as several user agents at once, for example by using multiple profiles or the implementation’s private browsing mode.

Despite that, it seems like none of the other uses of the term ‘user agent’ (including in Infra) are cross-referenced in this way.

Despite everything said so far, WHATWG specifications include many normative uses of the term ‘web browser’ (other than as an example). Some cases explicitly state that ‘web browsers’ are subject to different requirements from other user agents:

• HTML:

  • Section 2.1.8 (‘Conformance classes’):

    Web browsers that support the XML syntax must process elements and attributes from the HTML namespace found in XML documents as described in this specification, so that users can interact with them ….

    Web browsers that support the HTML syntax must process documents labeled with an HTML MIME type as described in this specification, so that users can interact with them.

    User agents that process HTML and XML documents purely to render non-interactive versions of them must comply to the same conformance criteria as web browsers, except that they are exempt from requirements regarding user interaction.

    This implies that the defining characteristic of ‘web browsers’ is interactivity, but that contradicts the subheading ‘Web browsers and other interactive user agents’.

  • Section 4.6.7 (‘Link types’):

    New link types that are to be implemented by web browsers are to be added to this standard. The remainder can be registered as extensions.

  • Section 7 (‘Loading web pages’):

    This section describes features that apply most directly to web browsers. Having said that, except where specified otherwise, the requirements defined in this section do apply to all user agents, whether they are web browsers or not.

    It seems that nothing in this section actually ‘specifie[s] otherwise’, but the possibility is there.

  • Section 9.2.6 (‘Interpreting an event stream’):

    This section defines the algorithm ‘dispatch the event’ differently, depending on whether the user agent is a ‘web browser’, even though the algorithm given for ‘web browsers’ seems entirely generic and applicable to any user agent that uses the EventSource interface.

Some uses, although non-normative, add to this confusion:

• URL:

  • Section 4.4 (‘URL parsing’):

    Note: Non-web-browser implementations only need to implement the basic URL parser.

    In other words, for ‘non-web-browser’ implementations, parsing of blob URLs does not need to be implemented, although everything else about blob URLs (including their origins, which are also defined in URL) still applies.

Other uses of the term ‘web browser’ imply that it is interchangeable with ‘user agent’ (but then the latter term, which has a definition, should be used exclusively):

• Compatibility abstract:

  This standard describes a collection of web platform features that web browsers need to support for compatibility with the de facto web.

• HTML:

  • Section 17.4 (‘text/ping’):

    Applications that use this media type:

    Web browsers.

    Restrictions on usage:

    Only intended for use with HTTP POST requests generated as part of a web browser's processing of the ping attribute.

    But section 4.6.6 (‘Hyperlink auditing’) uses the term ‘user agent’ exclusively.

• MIME Sniffing introduction (this section is probably meant to be non-normative):

  Historically, web browsers have tolerated [servers that supply a Content-Type header field value that does not match the actual contents of the response] by examining the content of HTTP responses in addition to the Content-Type header field ….

  …

  The algorithm originated from research conducted … on content sniffing algorithms present in popular user agents ….

• Quirks Mode abstract:

  Quirks Mode defines quirks in CSS and Selectors that are necessary to support for Web browsers for compatibility with the Web.

• Web IDL abstract:

  This standard defines an interface definition language, Web IDL, that can be used to describe interfaces that are intended to be implemented in web browsers.

• WebSockets:

  • Section 2.2 (‘Opening handshake’):

    Warning: The reason redirects are not followed and this handshake is generally restricted is because it could introduce serious security problems in a web browser context.

    (This seems analogous to an example or note and therefore non-normative.)

    This issue seems to apply equally to any user agent.