Hacker News new | past | comments | ask | show | jobs | submit login
Google confirms microformats are still a supported metadata format for content (jvt.me)
102 points by cxr on March 9, 2020 | hide | past | favorite | 40 comments

No. Just no.

I've removed all of the microdata from the web sites I manage. I'm not giving Google any information for free.

Google uses microdata to show information from web sites without people actually visiting the site. Google gets free labor. Publishers get squat. Adding microdata to content pages consolidates information with Google, hurting everyone else.

Google has billions of dollars to spend on scraping and searching and whatever it wants to do. It could mechanical turk the entire internet if it wanted.

I am not Google's free labor.

If you really want to stick it to google, you can always put a robots.txt file with disallow: * directive.

But you probably don't. Presumably because you rely on google to direct users to your website. I find it hard to claim that publishers get squat out of this arrangement when google is probably the primary method relevant users find your content.

Which to be clear, I'm not saying that website operators have any sort of duty to assist third parties in extracting just the useful part of their website. They obviously don't. They can make their website however they feel like. I just don't think the disdain of google over the practice of extracting and redisplaying content provided in a format explicitly for the purpose of letting people do that, is warranted. If you don't want to opt-in, that's fine, but google is doing nothing wrong by using data provided to the world for the general purpose of enabling them to do exactly what they are doing.

Presumably because you rely on google to direct users to your website.

You presume incorrectly.

The sites I'm talking about have no advertising. They are not ad-supported. They do not rely on Google or any other search engine in any way.

There are other business models.


Many of those business models still rely on people finding your site (although there are exceptions)

microdata and microformats aren’t the same thing.

Ok, but what's the difference in relation what the parent poster said? You go through the trouble of adding microformats to your site you are helping Google analyze that site and potentially present data from that site without you ever going there.

However microformats potentially help others, because of Google's actions it is to a site owner's benefit to not provide beneficial metadata past a certain point, assuming of course they want visits provided by Google.

On the off chance that you're adding microformats to your site in good faith rather than just trying to make it accessible to google, there are non google tools, utilities, search engines, applications, browser extensions and so on that make use of the microformat data, to, for instance provide "add to calendar" buttons to marked up events, or automatically convert an html blog to an rss feed. I don't know if they ever got there but there was a vision to have a standard markup for shopping cart items so browser extensions could assist with price comparisons.

That's my point exactly, if you are adding microformats to your site in good faith because you believe it is the best thing for the web and enables peoples tools to work with your site better, google penalizes you by reducing your site traffic and increases their sites utility for users, effectively increasing their profits off of your labor while starving you at the same time.

i don’t know how your hosting plan works, but on every plan I’ve looked at, traffic costs money. if your web based business depends on traffic to serve advertising, then no, you’re not doing it in good faith to make the web better.

I don't think stores want to facilitate price comparisons.

depends on which store you are in the comparison

Would Google penalize you for showing microdata to everyone but googlebot without changing anything else?

Supported, but not recommended. I switched to JSON-LD last year after using microdata for years.

Agreed. This post should be 100% be titled as supported instead of recommended.

I recall that John Mueller recommended[1] JSON-LD over the other options during a Google Webmasters office hours session.

Many of Google Search's Structured Data features[2] primarily have JSON-LD examples.

[1]: https://youtu.be/gS4_JH-QqSg?t=2085

[2]: https://developers.google.com/search/docs/guides/search-gall...

I hadn't heard of JSON-LD, but seems sensible, would sure be easier to work with than the traditional microformats.

The name seems unfortunate, though, as I thought that was "line delimited JSON"[1]!

[1] https://en.wikipedia.org/wiki/JSON_streaming#Line-delimited_...

No where does it say they are still a recommended metadata format. The only recommended format that they list on their website is JSON-LD:


Just wait, soon Google will announce that HTML is deprecated.

AMP and Chrome are Google's HTML "embrace, extend, extinguish". In twenty years the web will be nothing but silos and ad funnels.

Google is doing what Bill Gates wished Ballmer would have done. Thankfully the Justice Department stepped in to save the fledgling web. Will they be willing to do so to save it this time around?

Google needs to be forced to abandon Chrome and AMP.

Google took what MS did and made it absurd. Chrome is essentially adware, they just hide it too well. And Google is no longer a positive actor in our internet.

Google does a lot of things. Some negative some positive. Personally I am pretty concerned about the near monopoly that chrome has. However they have had an extremely positive impact on (for example) the security of web PKI infrastructure. I find it hard to call them not a positive actor on the internet, when they are one of the more responsible parties for the internet being secure from surveillance - a huge boon to the health of the internet.

I'm more worried about cross business optimization than single products like chrome.

Android is way more nefarious than a web browser. And they give it away for free to device manufacturers. I almost wish for the old days where everyone had their own OS, monoculture is terrifying.

Google recommends that developers use AMP instead.

I rather apply to Google and get rejected, for the experience, than even considering AMP.

Soon we will have only Chrome developer job ads, and the anti-IE 6 that drank the "Do no evil look aid" are the ones to blame.

If this take care of JavaScript together, I'll not be so mad.

What advantages do/did microformats provide over microdata/schema.org/ld-json?

Microformats was based on standard html. That comes with pros and cons. Pros being its very easy to adopt, including in legacy CMS systems that might not support the other fancier stuff. For that matter, it also means web developers already understand it (Especially compared to writing XML RDF documents...)

It also allowed mixing the presentation with the format. that meant people only had to maintain one copy of data, so its less likely to get out of sync, and people are less likely to forget to add the data later on.

Downside is you're using something not really meant to support the use case of annotating data, so it ends up being a bit hacky.

These things also did not come out all at the same time. Future techniques were partially made in response to microformats, just like (I'm assuming a bit here) that microformats were made as a response to SemanticWeb RDF-XML stuff.

I did a comparison here a while back: Simpler, more terse, use class rather than extra properties or giant json blobs. http://www.kevinmarks.com/microformatschema.html

One key thing I've found useful of it is that it attempts to reduce duplication, by marking up existing data with CSS classes that then provide meaning

predating them. everything you listed is inspired by/based on microformats

I assumed, it seems like a very 2008 solution. That's why I asked what advantages it had to justify retaining support for it for this long.

I've been doing web development since the commercial web started and somehow this thing eluded me.

there should be a mailing list for these things or something. "oembed", "jwt" and "opengraph" are also things that eluded me for a very very long time.

Are there any CMS's create with schema.org in mind. ex. All of the schema.org types are represented and the website is build using a GUI generated by schema.org definitions?

Thanks for sharing my post!

Hang on a sec... so Google specifies which vocabularies it accepts?! That goes completely against the idea of structured data. That's nasty and bad.

Getting a PHP OOM error on that page.

That site does a really poor job demonstrating what a Microformat is. There’s good descriptors but I was looking for an example or scenario to better elucidate their use...

The site has since been changed above (submitted url was http://microformats.org/2020/03/04/google-confirms-microform...).

Url changed from http://microformats.org/2020/03/04/google-confirms-microform..., which points to this.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact