As long as DDG uses Bing's results then they also implement Bing's censorship policies. That's a choice DDG makes.

Less a choice, more a consequence. It’s not like DDG wants to replicate MS’s search policies, it’s just that DDG lacks the infrastructure to do it on their own. Until they do, this is going to keep happening.

I don’t think ddg was consulted or informed here. Choice is a strong phrasing

I don't blame DDG, but I argue that fallback into Google (self operated search engine, and not operated in China) somewhat makes sense for this.

That's because Microsoft got caught censoring the results, then they lied and claimed it was a "human error," which absolutely nobody believes.

Then Microsoft manually took an action to restore the image results in Bing for the query "tank man." Which then filtered down to DDG since it relies on Bing (previously DDG's image results were similarly devoid of the proper results).

> That's because Microsoft got caught censoring the results, then they lied and claimed it was a "human error," which absolutely nobody believes.

Oh, I don't know. If we can label "accidentally using the wrong scope" as a human error, I could easily see a situation where someone meant to remove something from Chinese search results but accidentally did it worldwide. I don't know what their infrastructure is like, what kind of safeguards they have in place, or even what they do to block a result from the all-seeing algorithm. Which is more likely: that Microsoft would allow China to dictate what the rest of the world can/cannot see or that someone made a mistake and blocked something globally instead of regionally?

I find it baffled that a company of that caliber don't have regression tests that can verify region-bound changes wouldn't affect results intended for other regions.

It was a mistake not an accident. When you make an accident, test safety checks can save you.

However although you don’t intentionally make a mistake, you do intentionally carry it out.

In that case, no amount of safety checking will stop a human being from implementing bad logic.

If someone accidentally didn't scope the change as region bound, what makes you think they couldn't accidentally scope the test wrong the same way?

Human error knows no bounds.

>Then Microsoft manually took an action to restore the image results in Bing for the query "tank man."

Did they? https://www.bing.com/images/search?q=tank+man has no results of Tank Man. https://imgur.com/a/U5fDs7P (If you do see Tank Man images, see the edit at the end of this comment.)

If you press enter on the search box with the pre-filled "tank man", then you get a URL with more query-string parameters - https://www.bing.com/images/search?q=tank+man&imgurl=&cbir=s... , and that page does contain the actual Tank Man.

Seems it can be narrowed down to the form= parameter. https://www.bing.com/images/search?q=tank+man&form=QBILPG has Tank Man, without the form= it doesn't. The actual value of the form= parameter doesn't value either, as long as it's not empty.

I have no idea what they're doing in the backend.

Edit: Another weird thing - https://www.bing.com/images/search?q=tank+man with JS enabled redirects to a different page with additional querystring parameters, and this page does have Tank Man. My testing above was done with JS disabled, which is why I got the pre-redirect page that doesn't have his pictures.

they probably need the query string for pagination and if you don't have it they redirect you with some javascript which you turning it off breaks. i don't think there's any grand conspiracy there.

Yes, I didn't claim they were doing it intentionally, just that the website's behavior is weird. That's why I said:

>I have no idea what they're doing in the backend.

The pagination theory doesn't make sense to me though. The relevant parameter is called "form", its actual value doesn't matter as long as it's non-empty, and surely a sane implementation would treat the lack of a page parameter as page 1 and not page whatever-the-stormtrooper-image-is-on.