Conversation

twitter.com/DanielMicay/st This issue was finally resolved after Googlebot being stuck on Chromium 41 since 2015. Finally not going see errors in their webmaster tools from modern Content-Security-Policy, CSS, ES6, etc. webmasters.googleblog.com/2019/05/the-ne (pointed out to me by

@saleemrash1d

)

Quote Tweet

Daniel Micay

@DanielMicay

Apr 28, 2019

developers.google.com/search/docs/gu > Googlebot uses a web rendering service (WRS) that is based on Chrome 41 (M41). So that's why it's incapable of handling grapheneos.org/releases.js for grapheneos.org/releases. Rendering with Chrome 41 in 2019 embarrassingly bad. It's from early 2015...

Replying to

and

I wonder how much of the awfulness of their search results this was responsible for.

Replying to

and

Honesty their results would be so much better if they used lynx to crawl.

Replying to

and

Too many sites are using tons of JavaScript, so they would be entirely opaque to it. There's a lot of blog and forum software that requires JavaScript to display any results at all. Earlier versions of Discourse couldn't show anything without JavaScript and it had wide adoption.

Replying to

and

Exactly. Penalize sites that aren't accessible with -100% pagerank.

Replying to

and

This hell of content hidden in client-side expansion could never have happened if Google hadn't enabled it by processing it for indexing.

Replying to

and

It did happen without them being able to process huge amounts of it though. Sites worry about all kinds of weird SEO hacks instead of making the content properly structured in a way that makes it more accessible for both people and search engines. It apparently doesn't matter.

Replying to

and

Google cares far too much about domain authority, etc. now. It takes a couple years for a site to even start being able to rank properly in searches. Their approach to anti-spam has massively favoured major sites. My Reddit posts on GrapheneOS rank way above the official site...

4:59 PM · May 8, 2019Twitter Web Client

Replying to

and

Example: google.com/search?q=graph. The Reddit post linking to the build documentation at grapheneos.org/build ranks first. The actual official documentation that it links to is ranked as #7. Chromium build metadata #2 and other links are all nonsense. It loves big, old sites.

Replying to

and

They've pretty killed off being able to have small independent sites actually rank well in searches especially if it hasn't been aware of them for at least 2 years. You'll just see Wikipedia, StackOverflow, etc. It trusts user generated content on major sites and buries the rest.