Old fights between Google and Microsoft, again in Europe

With Google searches increasingly not generating traffic anymore, I can see how the news publishers are upset. But these laws are just plain wrong and the solution should be a lot simpler.

I think it’d be easier and better to force Google, Twitter, Facebook, Slack, Telegram and all these not to pull too much metadata from a URL. Let’s all go back to the time when adding a link was just that, a link, and you needed to manually write what that page was about.

Let me explain.

In the old days when you wanted to share a URL with your friends, you pasted it in a message. If you sent a message that looked like this:

You wouldn’t make a friend happy: it’s an empty message with a link to an obscure website. Lack of bandwidth was a thing and clicks were dedicated only to things that mattered. How can someone judge if a URL is worth clicking on? The netiquette asks you to add context to help the recipient understand why your message is worth your friend’s time and attention.

Time and attention is the most valuable currency in internet time. But humans are also lazy and that’s why modern clients extrapolate metadata from the URL to automatically add context. For example, that same URL in Telegram looks like this:

What happens when you click on that image in Telegram? You see that image magnified. Thus, depriving the original author of a click. Granted, XKCD doesn’t care for clicks.

But the New York Times does. This is what an article from the New York Times looks on Telegram:

The relevant chart is already there, with the lede: my friend may not click on the link after all and start arguing about it. By the way, this feature may be responsible for a lot of disinformation out there: people don’t read what they see and repost stuff based on what’s pulled automatically. Twitter often asks you to read something before retweeting it. I bet that people would read more if links on Twitter looked more dull.

The easiest way to fix this issue is the stop Google, Microsoft, Facebook, Telegram, Twitter, Slack, etc to mess up with the URL. Use them for what they are, and not pull extra metadata besides the title and dates of the page.