Techdirt will stop embedding tweets in articles

A sign attached to Twitter’s global headquarters is viewed from a sidewalk on Market Street in San Francisco, California. June 18, 2014. (Photo: Matthew Keys/The Desk/Creative Commons)

Technology publication Techdirt says it has rolled out a new policy that prohibits writers from embedding social media posts from Twitter in its news and commentary stories.

In a post on Tuesday, Techdirt editor Mike Masnick said the decision was made due to shifting policy and feature decisions made by Twitter’s new owner and chief executive, Elon Musk.

Twitter has become something of an unstable business since Musk took over: Even before the acquisition was finalized, Musk fired several top executives from the company, some of whom learned of their dismissal over e-mail.

In the first few days of his ownership, Musk proclaimed that Twitter’s verification feature would be rolled into a subscription service, hired two actors to impersonate laid-off Twitter employees and troll reporters who were covering actual layoffs, and then launched a separate verification system that works independent of its traditional one (he then ceremoniously “killed off” the new feature, only to resurrect it a few days later).

All of this change is a bit much for Masnick, who told writers at Techdirt that they are not to embed tweets in their articles for the foreseeable future.

As Masnick explains, Twitter’s volatility raises questions over whether embedded tweets will disappear at some point in the future, or whether they’ll only be available to embed if Musk decides to roll that feature into a subscription tier.

Techdirt has been burned like this before: Masnick explained that the website once used a document repository called Docstoc to store and embed court decisions and other legal documents, so they could be easily read within Techdirt’s articles.

“We’ve had plenty of old embedded content disappear around here,” Masnick wrote. “There are tradeoffs with any approach.”

The biggest problem with embedding tweets at the moment is that they can easily disappear when a user deletes them from their timeline or profile. At the moment, some content management systems like WordPress handle deleted tweets by showing the text of the tweet as a pull quote, followed by a link to the user’s profile. At the very least, it gives readers an idea of what was written in a tweet before it was deleted.

But it’s not clear how things will shake out at Twitter in the future: Musk could always decide that tweets should disappear in a way that makes the content difficult to obtain should a post be deleted by a user or the company. He could also roll out a new embedding system that breaks the current one, and then do with that feature whatever he wishes.

To get around this problem, Masnick has instructed Techdirt writers to take screen captures of a tweet, and then embed the screen capture within a story, whenever it is practical to do so. The image should be accompanied by the text of a tweet or a description of the post within the article itself.

“I’m disappointed that we need to do this, as embedding is a nice feature, and a core part of how the internet should work,” Masnick wrote. “And while I’m not expecting Twitter to just disappear overnight…the odds of it happening have jumped up to a degree that we need to plan ahead, and this is how we’ll be doing it going forward.”

Journalists have been expressing their angst about Twitter in another way: Moving to a relatively new service called Mastodon, which replicates (and, in some ways, improves upon) the Twitter experienced, but with a network of decentralized servers that are independently run, yet interoperable.

Not everyone who was using the Mastodon protocol before is psyched by the idea that journalists are now embracing it: A server called “journa.host” that has amassed around 1,000 mainstream journalists has been blocked by other 70 other Mastodon servers, making it difficult — and, sometimes, impossible — for reporters and editors whose accounts were created through journa.host to interact with other users.

On a website that tracks the interoperability of Mastodon nodes, the administrators of servers who have blocked “journa.host” have given any number of reasons for doing so — from the belief that journalists on the server will engage in “surveillance capitalism,” to concerns that the moderators of the server were allowing abusive behavior in flagrant violation of certain rules that were set but not enforced.

Speaking from experience, this reporter attempted to join “journa.host” and went through the motions of applying based on a set of criteria laid out in the server’s application: Signing up from a work or branded e-mail address, verifying my identity by putting a piece of code on an already-verified third-party social media profile, and explaining my occupation as a journalist. An invitation was initially extended, only to be immediately revoked; in an e-mail, journa.host moderator and Slate.com community manager Evan Urquhart said a decade-old legal issue was raised as a concern “by some higher ups,” and ultimately became a disqualifying factor (Urquhart suggested I “go for a less uptight server” instead).

Ultimately, I joined Mastodon’s newsie.social server, which has three times the number of journalism participants and allows for external verification through a tool called PressCheck. As of Thursday, newsie.social is only blocked by four other Mastodon servers.