FWIW, startpage (like scroogle.org before it) re-enables the behavior that is so familiar to most web-knowledgeable people. Startpage is just a Google proxy.

I've been thinking it might be interesting to allow for easy hosting of Google proxies, like a one-click install on digitalocean that could be shared among a group of friends (~100 users at most).

Interesting: it looks to switch out the value for the href on mouse down...

I don't understand why they do it this way. If they're using javascript to replace the value anyway, why not just leave the value alone, and capture the data they want behind the scenes with javascript? What does running the traffic through a redirect get them that javascript itself cannot?

well, i guess if the click triggers a page reload, the onclick-async event won't fire (or at least complete).

Reddit does the same thing.

Do they still bother to do this in Chrome, even though Chrome supports the <a> ping attribute that obviates the need for it (and which was presumably added to Chrome exactly for Google Search's own use-case)?

I don't know, I don't use Chrome.

> even though Chrome supports the <a> ping attribute that obviates the need for it (and which was presumably added to Chrome exactly for Google Search's own use-case)?

And that's one reason why I don't use Chrome.

There is so much wrong with a company adding features to their browser in order to facilitate special cases for their own website.

> There is so much wrong with a company adding features to their browser in order to facilitate special cases for their own website.

Isn't <a> ping good for the end user? Unless you disagree with any ability to track outbound clicks - in which case we can politely agree to disagree.

Even if you disagree with the ability to track outbound clicks, ping would be good as if it becomes the main way of tracking clicks you'd be able to stop tracking by turning it off. When people do tracking by rewriting links server-side, on the other hand, you don't have an easy out.

I don't see this as a special case at all - a substantial proportion of link shortener usage, for example, is not for link shortening, but because most of them offer analytics.

The url that you copy is not unusable, it's a redirect url.

But that does make it unusable for anything but directly posting it into the address field of your own browser window.

Which means you end up having to open the URL first yourself just to copy and share it. And that doesn't even always work, since URLs triggering downloads don't update the URL in the address field (at least in Firefox).

Lets try.

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&c...

I'm sharing it with you. Is this unusable? Did you not get to the news article I was sharing with you by clicking on the link? So what's your point?

It's unusable. Before I clicked the link I couldn't tell if it was going to take me to chicagotribune.com (it does) or go*tse.cx.

That's my problem. I cannot in good faith give this kind of link to my friends.

"before you clicked" the link you haven't used it.

I'd argue that link user experience starts before you actually click a link. Maybe when you're deciding whether or not to click?

Same reason that URL shorteners are bad UX. Source (creator of Delicious): http://joshua.schachter.org/2009/04/on-url-shorteners

I'd argue that the vast majority of users have no idea what you're talking about.