5 ms·

I don't plan on implementing these, and neither should you, unless they get added as part of an official HTML specification. Maybe they are, but I don't see the

I don't plan on implementing these, and neither should you, unless they get added as part of an official HTML specification. Maybe they are, but I don't see them on this list of link types [0]. Until they get formally specified, to me this is just Google doing Google things and fragmenting the web. [0] https://developer.mozilla.org/en-US/docs/Web/HTML/Link_types https://developer.mozilla.org/en-US/docs/Web/HTML/Link_types

Spivak 7y ago
I mean we collectively did this to ourselves by letting browsers be super permissive about invalid markup. The only reason that Google is able to do this is because things that don’t know what rel=“usg” means will silently ignore it.
- amelius 7y ago
  They would have found another, probably less elegant, way to do the same thing. E.g. they'd put the extra info in HTML comments.
  - zonidjan 7y ago
    Oh, hi Microsoft. 
- the_pwner224 7y ago
  You're confusing forward-compatibility (ignore markup with tags you don't recognize, which is awesome) with accepting invalid markup (which is stupid).
magicalist 7y ago
> unless they get added as part of an official HTML specification. Maybe they are, but I don't see them on this list of link types [0] Extending link types is in the html specification: https://html.spec.whatwg.org/multipage/links.html#other-link-types https://html.spec.whatwg.org/multipage/links.html#other-link... > Extensions to the predefined set of link types may be registered in the microformats wiki existing-rel-values page [http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions http://microformats.org/wiki/existing-rel-values#HTML5_link_...]. where "may" is defined in the usual way (they aren't in there, but the page hasn't been updated for over a year, so I'm not sure if anyone cares about it anymore after already adding so many extension values). It's common to have user agent (in this case googlebot) specific rel values.
- akersten 7y ago
  That's a fair point, thanks. I do still feel that the attribute is woefully underspecified (evidenced by the lack of the microformats list being updated, and the fact that it's a wiki) so in my opinion the least Google could do here is to shore that up. If they contributed to the wiki or formalized it a little bit (like the HSTS list or public suffix list) it would go a long way as far as how I feel about their additions to the attribute. I realize that's quite a lot to ask for "just" a few new link types though.
  - tannhaeuser 7y ago
    To be fair, the WHATWG HTML "Standard" is also nothing more than a Wiki, or even just a collaborative space where everything can change at any time.
    - magicalist 7y ago
      > To be fair, the WHATWG HTML "Standard" is also nothing more than a Wiki, or even just a collaborative space where everything can change at any time. well unless you're waiting for the second coming for someone official to come along and bless an RFC, all standards are born of a collaborative space where everything can change at any time. Versions are just snapshots, and good luck if you want to stick with one forever (hows that TLS 1.0 server going to work out after March 2020?). Anyways, if you want an official© W3 spec, here it is in HTML5 https://www.w3.org/TR/html50/links.html#other-link-types https://www.w3.org/TR/html50/links.html#other-link-types, HTML 4.01 https://www.w3.org/TR/html401/types.html#type-links https://www.w3.org/TR/html401/types.html#type-links, or hey, even back to HTML 2 https://tools.ietf.org/html/rfc1866#section-5.7.3 https://tools.ietf.org/html/rfc1866#section-5.7.3
      - tannhaeuser 7y ago
        I'm well aware of the various W3C's and WHATWG HTML specs (see my site at [1]); that was kindof my point ;) [1]: http://sgmljs.net/ http://sgmljs.net/
reaperducer 7y ago
It feels like yet another way for Google to offload the hard work of indexing the web into other people. See also; microdata, and the monthly alerts I get from Google about the content of my sites being malformed, yet they validate just fine in a dozen other tools. If this keeps up, eventually we'll have to log in to Google to provide it with with the URLs of new content, and fill in all of the meta data about each page. All in the name of whatever buzzword Big G comes up with that month.
- gpm 7y ago
  Dare I say that a better indexed web is a public good that we all benefit from immensely. Logging into Google and providing it with the URLs isn't analogous because that only benefits Google, it provides a barrier to the entry of competition, which isn't good for us users.
  - zonidjan 7y ago
    > Dare I say that a better indexed web is a public good that we all benefit from immensely. Sure, if it were used for the benefit of the public, rather than the benefit of Google.
- notatoad 7y ago
  >If this keeps up, eventually we'll have to log in to Google to provide it with with the URLs of new content, and fill in all of the meta data about each page FWIW, most people who care about their search ranking would absolutely love a tool like this. The point of the automated crawl isn't just to find the information, it's because if you let people submit their own content to the index they'll lie about what it is.
- tannhaeuser 7y ago
  > hard work of indexing the web IMO if you mean that Google should check linked-to content rather than relying on link qualifiers, Google has been doing a bad job in recent years, as I can't find useful material among an ocean of low-effort clickbait most of the time. OTOH, relying on metadata by publishers won't solve this problem.

Spivak 7y ago

I mean we collectively did this to ourselves by letting browsers be super permissive about invalid markup. The only reason that Google is able to do this is because things that don’t know what rel=“usg” means will silently ignore it.

amelius 7y ago
They would have found another, probably less elegant, way to do the same thing. E.g. they'd put the extra info in HTML comments.
- zonidjan 7y ago
  Oh, hi Microsoft.
the_pwner224 7y ago
You're confusing forward-compatibility (ignore markup with tags you don't recognize, which is awesome) with accepting invalid markup (which is stupid).

magicalist 7y ago

> unless they get added as part of an official HTML specification. Maybe they are, but I don't see them on this list of link types [0] Extending link types is in the html specification: https://html.spec.whatwg.org/multipage/links.html#other-link-types https://html.spec.whatwg.org/multipage/links.html#other-link... > Extensions to the predefined set of link types may be registered in the microformats wiki existing-rel-values page [http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions http://microformats.org/wiki/existing-rel-values#HTML5_link_...]. where "may" is defined in the usual way (they aren't in there, but the page hasn't been updated for over a year, so I'm not sure if anyone cares about it anymore after already adding so many extension values). It's common to have user agent (in this case googlebot) specific rel values.

akersten 7y ago
That's a fair point, thanks. I do still feel that the attribute is woefully underspecified (evidenced by the lack of the microformats list being updated, and the fact that it's a wiki) so in my opinion the least Google could do here is to shore that up. If they contributed to the wiki or formalized it a little bit (like the HSTS list or public suffix list) it would go a long way as far as how I feel about their additions to the attribute. I realize that's quite a lot to ask for "just" a few new link types though.
- tannhaeuser 7y ago
  To be fair, the WHATWG HTML "Standard" is also nothing more than a Wiki, or even just a collaborative space where everything can change at any time.
  - magicalist 7y ago
    > To be fair, the WHATWG HTML "Standard" is also nothing more than a Wiki, or even just a collaborative space where everything can change at any time. well unless you're waiting for the second coming for someone official to come along and bless an RFC, all standards are born of a collaborative space where everything can change at any time. Versions are just snapshots, and good luck if you want to stick with one forever (hows that TLS 1.0 server going to work out after March 2020?). Anyways, if you want an official© W3 spec, here it is in HTML5 https://www.w3.org/TR/html50/links.html#other-link-types https://www.w3.org/TR/html50/links.html#other-link-types, HTML 4.01 https://www.w3.org/TR/html401/types.html#type-links https://www.w3.org/TR/html401/types.html#type-links, or hey, even back to HTML 2 https://tools.ietf.org/html/rfc1866#section-5.7.3 https://tools.ietf.org/html/rfc1866#section-5.7.3
    - tannhaeuser 7y ago
      I'm well aware of the various W3C's and WHATWG HTML specs (see my site at [1]); that was kindof my point ;) [1]: http://sgmljs.net/ http://sgmljs.net/

reaperducer 7y ago

It feels like yet another way for Google to offload the hard work of indexing the web into other people. See also; microdata, and the monthly alerts I get from Google about the content of my sites being malformed, yet they validate just fine in a dozen other tools. If this keeps up, eventually we'll have to log in to Google to provide it with with the URLs of new content, and fill in all of the meta data about each page. All in the name of whatever buzzword Big G comes up with that month.

gpm 7y ago
Dare I say that a better indexed web is a public good that we all benefit from immensely. Logging into Google and providing it with the URLs isn't analogous because that only benefits Google, it provides a barrier to the entry of competition, which isn't good for us users.
- zonidjan 7y ago
  > Dare I say that a better indexed web is a public good that we all benefit from immensely. Sure, if it were used for the benefit of the public, rather than the benefit of Google.
notatoad 7y ago
>If this keeps up, eventually we'll have to log in to Google to provide it with with the URLs of new content, and fill in all of the meta data about each page FWIW, most people who care about their search ranking would absolutely love a tool like this. The point of the automated crawl isn't just to find the information, it's because if you let people submit their own content to the index they'll lie about what it is.
tannhaeuser 7y ago
> hard work of indexing the web IMO if you mean that Google should check linked-to content rather than relying on link qualifiers, Google has been doing a bad job in recent years, as I can't find useful material among an ocean of low-effort clickbait most of the time. OTOH, relying on metadata by publishers won't solve this problem.