RFC 2324 was written by Larry Masinter, who describes it as a satire, saying "This has a serious purpose – it identifies many of the ways in which HTTP has been extended inappropriately." The wording of the protocol made it clear that it was not entirely serious; for example, it notes that "there is a strong, dark, rich requirement for a protocol designed espressoly [sic] for the brewing of coffee".
Despite the joking nature of its origins, or perhaps because of it, the protocol has remained as a minor presence online. The editor Emacs includes a fully functional client side implementation of it, and a number of bug reports exist complaining about Mozilla’s lack of support for the protocol. Ten years after the publication of HTCPCP, the Web-Controlled Coffee Consortium (WC3) published a first draft of "HTCPCP Vocabulary in RDF" in parody of the World Wide Web Consortium's (W3C) "HTTP Vocabulary in RDF".
On April 1, 2014, RFC 7168 extended HTCPCP to fully handle teapots.
There are also makers that rigged actual coffee or tea pots with HTCPCP-compatible control interface, either in a humorous and satirical but simple way, or in a massively extended, fully functional Internet of Things way.
HTCPCP is an extension of HTTP. HTCPCP requests are identified with the uniform resource identifier (URI) scheme coffee (or the corresponding word in any other of the 29 listed languages) and contain several additions to the HTTP methods:
BREW or POST
Causes the HTCPCP server to brew coffee. Using POST for this purpose is deprecated. A new HTTP Request header field "Accept-Additions" is proposed, supporting optional additions including Cream, Whole-milk, Vanilla, Raspberry, Whisky, Aquavit, etc.
The HTCPCP server is unable to provide the requested addition for some reason; the response should indicate a list of available additions. The RFC observes that "In practice, most automated coffee pots cannot currently provide additions."