Careful observers posted to Hacker News this week – and the Washington Post reported – that the X formerly known as Twitter (XFKAT?) appeared to be deliberately introducing a delay in loading links to sites the owner is known to dislike or views as competitors. These would be things like the New York Times and selected other news organizations, and rival social media and publishing services like Facebook, Instagram, Bluesky, and Substack.
The 4.8 seconds users clocked doesn’t sound like much until you remember, as the Post does, that a 2016 Google study found that 53% of mobile users will abandon a website that takes longer than three seconds to load. Not sure whether desktop users are more or less patient, but it’s generally agreed that delay is the enemy.
The mechanism by which XFKAT was able to do this is its built-in link shortener, t.co, through which it routes all the links users post. You can see this for yourself if you right-click on a posted link and copy the results. You can only find the original link by letting the t.co links resolve and copying the real link out of the browser address bar after the page has loaded.
Whether or not the company was deliberately delaying these connections, the fact is that it *can* – as can Meta’s platforms and many others. This in itself is a problem; essentially it’s a failure of network neutrality. This is the principle that a telecoms company should treat all traffic equally, and it is the basis of the egalitarian nature of the Internet. Regulatory insistence on network neutrality is why you can run a voice over Internet Protocol connection over broadband supplied by a telco or telco-owned ISP even though the services are competitors. Social media platforms are not subject to these rules, but the delaying links story suggests maybe they should be once they reach a certain size.
Link shorteners have faded into the landscape these days, but they were controversial for years after the first such service – TinyURL – was launched in 2002 (per Wikipedia). Critics cited several main issues: privacy, persistence, and obscurity. The latter refers to users’ inability to know where their clicks are taking them; I feel strongly about this myself. The privacy issue is that the link shorteners-in-the-middle are in a position to collect traffic data and exploit it (bad actors could also divert links from their intended destination). The ability to collect that data and chart “impact” is, of course, one reason shorteners were widely adopted by media sites of all types. The persistence issue is that intermediating links in this way creates one or more central points of failure. When the link shortener’s server goes down for any reason – failed Internet connection, technical fault, bankrupt owner company – the URL the shortener encodes becomes unreachable, even if the page itself is available as normal. You can’t go directly to the page, or even located a cached copy at the Internet Archive, without the original URL.
Nonetheless, shortened links are still widely used, for the same reasons why they were invented. Many URLs are very long and complicated. In print publications, they are visually overwhelming, and unwieldy to copy into a web address bar; they are near-impossible to proofread in footnotes and citations. They’re even worse to read out on broadcast media. Shortened links solve all that. No longer germane is the 140-character limit Twitter had in its early years; because the URL counted toward that maximum, short was crucial. Since then, the character count has gotten bigger, and URLs aren’t included in the count any more.
If you do online research of any kind you have probably long since internalized the routine of loading the linked content and saving the actual URL rather than the shortened version. This turns out to be one of the benefits of moving to Mastodon: the link you get is the link you see.
So to network neutrality. Logically, its equivalent for social media services ought to include the principle that users can post whatever content or links they choose (law and regulation permitting), whether that’s reposted TikTok videos, a list of my IDs on other systems, or a link to a blog advocating that all social media companies be forced to become public utilities. Most have in fact operated that way until now, infected just enough with the early Internet ethos of openness. Changing that unwritten social contract is very bad news even though no one believed XFKAT’s CEO when he insisted he was a champion of free speech and called the now-his site the “town square”.
If that’s what we want social media platforms to be, someone’s going to have to force them, especially if they begin shrinking and their owners start to feel the chill wind of an existential threat. You could even – though no one is, to the best of my knowledge – make the argument that swapping in a site-created shortened URL is a violation of the spirit of data protection legislation. After all, no one posts links on a social media site with the view that their tastes in content should be collected, analyzed, and used to target ads. Librarians have long been stalwarts in resisting pressure to disclose what their patrons read and access. In the move online in general, and to corporate social media in particular, we have utterly lost sight of the principle of the right to our own thoughts.
Illustrations: The New York City public library in 2006..
Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series she is a contributing editor for the Plutopia News Network podcast. Follow on