Google embraces, extends, and extinguishes

Published 2018-05-03 on Drew DeVault's blog

Microsoft infamously coined the euphemism “embrace, extend, extinguish” to describe their strategy for disrupting markets dominated by open standards. These days, Microsoft seems to have turned the other leaf, contributing to a huge amount of open source and supporting open standards, and is becoming a good citizen of the technology community. It’s time to turn our concerns to Google.

Google famously “embraced” email on April Fool’s day, 2004, which is of course based on an open standard and federates with the rest of the world. If you’ve read the news lately, you might have seen that Google is shipping a big update to GMail soon, which adds “self-destructing” emails that vanish from the recipient’s inbox after a time. Leaving aside that this promise is impossible to deliver, look at the implementation - Google emails a link to a webpage with the actual email content, and does magic in their client to make it look seamless. Thus, they “extend” email. The “extinguish” with GMail is also well underway - it’s infamous for having an extremely strict spam filter for incoming emails from people who run personal or niche mail servers.

Then there’s AMP. It’s an understatement to say Google embraced the web - but AMP is how they enter the “extend” phase. AMP is a “standard”, but they don’t listen to any external feedback on it and it serves as a vehicle for keeping users on their platform even when reading content from other websites. This is thought to be the main intention of the service, as there are plenty of other (and more effective) ways of rewarding lightweight pages in their search results. The “extinguish” phase comes as sites that don’t play ball get pushed out of Google search results and into obscurity. AMP is perhaps the most blatant of Google’s strategies, serving only to further Google’s agenda at the expense of everyone else.

The list of grievances continues. Consider Google’s dizzying collection of chat applications. In its initial form, gtalk supported XMPP, an open and federated standard for chat applications. Google dropped support for XMPP in 2014 and continued the development of their proprietary platform up thru today’s Hangouts and Google Chat platforms - neither of which support any open standards. Slack is also evidently taking cues from Google here, recently shutting down their own IRC and XMPP bridges.

Google Reader’s discontinuation fits too. RSS’s decline was evident before Google axed it, but killing Reader dealt a huge blow to any of RSS’s remaining momentum. Google said themselves they wanted to consolidate users onto the rest of their services - none of which, I should add, support any open syndication standards.

What of Google’s role as a participant in open source? Sure, they make a lot of software open source, but they don’t collaborate with anyone. They forked from WebKit to get Apple out of the picture, and contributing to Chromium as a non-Googler is notoriously difficult. Android is the same story - open source in principle, but non-Googler AOSP contributors bemoan their awful approach to external patches. It took Google over a decade to start making headway on upstreaming their Linux patches for Android, too. Google writes papers about AI, presumably to incentivize their academics with recognition for their work. This is great until you notice that the crucial piece, the trained models, is always absent.

For many people, the alluring convenience of Google’s services is overwhelming. It’s hard to hear these things. But we must face facts: embrace, extend, extinguish is a core part of Google’s playbook today. It’s important that we work to diversify the internet and fight the monoculture they’re fostering.

2018-05-04 18:12 UTC: I retract my criticism of Google’s open source portfolio as a whole, and acknowledge their positive impact on many projects. However, of the projects explicitly mentioned I maintain that my criticism is valid.

2018-05-05 11:17 UTC: Apparently the previous retraction caused some confusion. I am only retracting the insinuation that Google isn’t a good actor in open source, namely the first sentence of paragraph 6. The rest of the article has not been retracted.