search

whatwg / html

HTML Standard
https://html.spec.whatwg.org/multipage/
Other
8.01k stars 2.62k forks source link

Allow link rel=copyright to be conforming #3724

Open reschke opened 6 years ago

reschke commented 6 years ago

see https://github.com/validator/validator/issues/654

Deprecating "Copyright" in favor of "License", when UAs need to handle "Copyright" anyway, and "Copyright" is valid in HTML4, seems to be not productive.

sideshowbarker commented 6 years ago

I agree there doesn’t seem to be any compelling reason to make rel=copyright a document-conformance error (as the spec currently does) — and in general, other such “synonyms” either.

That would mean removing the following requirement:

https://html.spec.whatwg.org/multipage/links.html#linkTypes:attr-hyperlink-rel

Some of the sections that follow the table below list synonyms for certain keywords. The indicated synonyms are to be handled as specified by user agents, but must not be used in documents (for example, the keyword "copyright").

annevk commented 6 years ago

It's not quite that simple, the first synonym is:

Synonyms: For historical reasons, user agents must also treat link, a, and area elements that have a rev attribute with the value "made" as having the author keyword specified as a link relationship.

An alternative solution might be to just add them to the table as a row under the thing they're a synonym for and jointly share the remainder of cells on those rows.

domenic commented 6 years ago

Is there any motivation for encouraging such rel proliferation?

In general we have plenty of features UAs must handle that authors must not use, e.g. misnested HTML tags, appcache, etc.

sideshowbarker commented 6 years ago

Is there any motivation for encouraging such rel proliferation?

In the case of “copyright”, it’s not proliferation, since it has been widely used all along and was part of HTML4, etc. In that sense the intent of restriction in the spec is prescriptive rather than descriptive — that is, it’s trying to change existing author behavior that’s not objectively bad. What I mean is, it’s not clear why “license” is intrinsically better than “copyright”, or why both can’t be used. It’s not necessary for it to be strict either-or.

And more broadly speaking it’s not clear to me that we are gaining much of anything by putting restrictions on rel to begin with. I can’t see that it’s had any real effect in practice on reducing the proliferation of rel values.

People still use a bunch of “invalid” rel values for a bunch of different reasons and to them the errors from the checker about them are just noise that risks drowning out much more important mistakes their documents might have — or risks giving them incentive to just nto bother checking their documents at all, since they can’t get them to “pass” the checker anyway due to the rel errors.