> And I had to explain that it’s probably not possible, except if ios or Android implemented it directly (more so iOS, I’m not up familiar with android).
PDFs with embedded bitmaps can actually do that (and are currently supported widely, but PDFs are a terrible format with way too much attack surface). Same for any other composite html+whatever formats like epub. Standardizing a non-book format name and a spec that mandates nothing more than bare xhtml & image support with no JS or CSS interpreters would probably be quite sufficient.
> duckduckgo
Duckduckgo caches/proxies the images on its desktop site, which is useful when it gives results for sites that are since dead. Often at a lower resolution, but that's what reverse image searching is for.
PDFs with embedded bitmaps can actually do that (and are currently supported widely, but PDFs are a terrible format with way too much attack surface). Same for any other composite html+whatever formats like epub. Standardizing a non-book format name and a spec that mandates nothing more than bare xhtml & image support with no JS or CSS interpreters would probably be quite sufficient.
> duckduckgo
Duckduckgo caches/proxies the images on its desktop site, which is useful when it gives results for sites that are since dead. Often at a lower resolution, but that's what reverse image searching is for.