One thought is that we could have two mandated versions of every website: (a) the text-retrieval optimised version and (b) the web developer engagement version.

One thought is that we could have two mandated versions of every website: (a) the text-retrieval optimised version and (b) the web developer engagement version. When this idea has come up on HN before it has been met by angry web developers claiming they "would have to maintain two websites" or some such nonsense. However, consider that no one composes content in JSON, or even HTML. Textual content always begins as readable text. Even if only in the mind of the person creating it. There is nothing to maintain as there is nothing to "develop". All that needs to be done is add the HTTP header and deliver the text. One of the most popular examples of text retrieval I can think of to illustrate: https://www.kernel.org/doc/Documentation/ https://www.kernel.org/doc/Documentation/ Personally I prefer the directory-style "Index of" to the HTML-ised "Up / Next / Previous". Same functionality. Even if the complaint "we would have to maintain two websites" had any substance, which website do you think would be easier to maintain? That is a rhetorical question. That means I already know the answer.