Voting on Codeberg's new privacy policy

Nobody cares about privacy policies. Except when they do.

Deadline tomorrow !!! Everything you've ever posted becomes public from tomorrow. Even messages that have been deleted or the photos not allowed It costs nothing for a simple copy and paste, better safe than sorry. Channel
13 News talked about the change in Facebook's privacy policy. I do not give Facebook or any entities associated with Facebook permission to use my pictures, information, messages or posts, both past and future. With this statement, I give notice to Facebook it is strictly forbidden to disclose, copy, distribute, or take any other action against me based on this profile and/or it: contents. The content of this profile is private and confidential information.
The violation of privacy can be punished by law (UCC 1-308- 1 1 308-103 and the Rome Statute). NOTE: Facebook is now a public entity.

Once in a while some big tech platform updates their privacy policy to award themselves even more access to our personal data. WhatsApp’s 2021 privacy policy update springs to mind, as well as any number of scandals about privacy on platforms like Facebook or TikTok.

Every time it happens, communities scramble to understand the changes after the fact and pressure the company to walk back some aspect of the changes. Sometimes, like with the β€œDeadline tomorrow!!!” copypasta screenshot above, there wasn’t even a policy change to begin with, and peoples’ latent sense of distrust and powerlessness just led them to hedge their bets and copypaste the magic words onto their wall.

In that context I thought it was actually pretty cool receiving this email from Codeberg this week inviting members to vote on their new privacy policy.

After a lot of work and long discussions, the draft for the new privacy policy is in its final form and ready for voting. An update to our privacy policy is absolutely required legally due to the DSGVO since quite a long time ago - the first draft was presented at the last annual assembly, and now we are finally at a state that's legally checked and reviewed in-depth by some of you.

As a member of Codeberg e.V. do you support the following resolution:

”The members of Codeberg e. V. hereby agree to update the privacy policy according to https://codeberg.org/Codeberg/org/pulls/54"

I wrote about joining Codeberg earlier this year. This type of vote is the payoff for pooling your resources in democratic, community-led infrastructure. You get a meaningful say in how it evolves instead of just playing catch-up after corporate headquarters drops their latest announcement.

More than ever I’m convinced this is the future we should choose for the free software and open source world. And it’s easy to join the revolution yourself over at join.codeberg.org once you’ve created your account.

Is metadata streaming really SEO armageddon?

This blog post called Next.js 15.1+ is unusable outside of Vercel has been doing the rounds this past week.

Starting with version 15.1.8, Next.js might break metadata handling for non-Vercel deployments, potentially devastating your search rankings.

It's about metadata streaming, which is this new TTFB optimisation where Next can load in meta tags asynchronously after the initial page load. The post presents this feature as an SEO doomsday scenario. It has a tinge of "Hackers Can Turn Your Home Computer into a BOMB" energy.

I'm not convinced. We've had this version of Next running in production at work for months. We have asynchronous generateMetadata functions all over the place. Not running on Vercel. No SEO doomsday has materialised as of yet.

I spent a good while rereading the article and the linked GitHub issues and couldn't find anything suggesting that metadata streaming behaves differently depending on whether you host your app on Vercel or elsewhere either. The claim that this new feature represents an escalation to a new level of vendor lock-in seems to be based on vibes as far as I can tell.

There was also this part about static sites.

Here's where it gets absurd. Even with static builds, metadata tags aren't included in the HTML head anymore. They're bundled with React Server Components and require JavaScript execution. Your static site server now needs crawler detection logic just to serve basic HTML metadata.

This claim seemed both unlikely and easy to check. So I created a quick test app by running create-next-app and then configuring it to be a static build and adding an asynchronous metadata fetch. You can load the resulting webpage yourself and see the result in its source code.

<title>Example Title</title><meta name="description" content="Example Description"/>

Metadata tags are indeed still included in the HTML head in static Next sites. The claim seems to simply be false.

There's plenty to dislike about React and Next. Even though I've been paying my bills by writing JSX for the better part of a decade, Heydon Pickering's What Is React? sums up my feelings about the whole thing fairly well.

This is the dominant web development tech stack of our time and it's flawed in lots of really fundamental ways. There's so much that's problematic about it that I don't think we need to stretch quite this hard in search of new complaints. Metadata streaming itself seems to embody everything that's become so silly about contemporary React, always growing ever more complex in order to solve increasingly niche problems. But it does at least seem to work.