Tuesday, May 10, 2022
Monday, May 2, 2022
Googlefd resolves to -- Bing? SHENANIGANS.
UPDATE: a zipped file of the offending page! available for download here.
WTF is this "googlefd" domain that is suggesting I try some searches instead that are the same exact search that I just did, but on Bing? Is this a Bing thing? If so, is dirty pool, so I think not. Inspired by an intrepid Redditor, I took screenshots & decided to trace it -- both with tracert & cURL.
The phenomenon:
When I mouse over it it resolves to a rather shady sounding "googlefd.com." To quantify how shady, I did some searching — had to use a different browser though, after getting suspiciously sparse results.

Apparently not much. Ipqualityscore.com reaffirmed my suspicions, based on supposed pattern matching to other suspicious websites.
![]() |
All these are suspicious if you ask me. |
![]() |
Whoxy on googlefd. Click to enlarge, should open in new tab. |
I got a 'That's all we know" Google error. Sure I did... Wait, no. That's not what it was doing in the middle -- how'd we get bback to Google?
Besides, the source says it is the "That's all we know " with the picture of the robot -- and there are no images on. the actual error message I got.
![]() |
As you can see, there were supposed to be images on the page below. |
I did, however, save the HTML of that page about eight times. So I could have a look. And Archive.IS is nice enough to both render the page and provide a screenshot. So we can see that Bing logo perfectly rendered on the latter, missing on the former, where the people requesting obviously didn't have permission. You can go there to look at that.
I will leave you with the image of that loading page, which I reconstructed from my saved html. Maybe this will help someone who has an idea of what sort of shenanigans to which these guys from the UK are up.
Subscribe to:
Posts (Atom)