Cliqz Search Engine is no longer in operation (2020) | Hacker News Reader

Cliqz Search Engine is no longer in operation (2020)

cliqz.com

19 pointsbelter24 comments

Think of this what you will but in my opinion a privacy respecting search engine that is built on the ruins of a terrible privacy offender is not what I want.

In other words: It is easy to offer privacy respecting search that beats your privacy respecting competitors when you sit on a ton of high quality clickstream data, that has been acquired by not so privacy respecting means.

At least it is only a temporary advantage and we will see if Brave manages to still provide high quality search when their Cliqz legacy goes stale.

Is the terrible privacy offender you're referring to Brave, or Cliqz? I'm aware of some somewhat questionable things that Brave has done but I've never heard of Cliqz before this.

Cliqz. Regardless of what their marketing copy made you believe they collected the full clickstream of all their browsers users.

That's not true at all. I worked with Cliqz behind the scenes and know you are incorrect. Think what you want. Everyone at Cliqz was of high integrity and wanted a product that they themselves would be proud to use, as individuals with privacy at the center of their worlds.

So Mozilla wasn't sending browsing history to Cliqz?

Cliqz had a lot of docs at the time about ways to preserve privacy with client side systems that clearly took a lot of developer effort to research. Their reputation is undeserved unless new information came out

Do you have a reference to documentation that shows that it was intended that these measures were to taken on the client side? Do you have proof that it was implemented on the client side.

I have proof that the users clickstream ended up at Cliqz backend. What happened there is out of my knowledge and I don't care. If you respect my privacy you don't send the data that you claim you don't collect to your backend in the first place.

Could you share the proof?

See this old thread:

https://news.ycombinator.com/item?id=21677853

...and no, I don't buy the "this ends up at servers we have no shared ownership with" someone wrote in the answers.

As I said in other answers here, there are the client binaries out there (and I have them archived). Try them, dump the traffic. Maybe the backend servers don't exist anymore but I'm sure you can work around that.

This is not quite the same as sending the whole clickstream, is it? Sending what you type into the URL off somewhere to autocomplete is what pretty much all browsers do out of the box, and it’s been the norm for many years. I’m not saying it’s a good norm, or that Cliqz is good, but this doesn’t seem to be an indicator of them being uniquely bad.

Clickstream data was not sent. He's probably referring to telemetry, all of which was sent to proxy servers run by FoxyProxy then forwarded to Cliqz servers (this removed ip addresses of the source).

I have archived the old client binaries and I have the pcaps of the traffic they send to backend servers. It is full clickstream data. I stand by my point that Cliqz collected the clickstream of their browsers users. Sue me if I'm wrong.

If you have evidence, post it. Let's see the pcaps.

Do you have any evidence for this accusation?

I have network traffic dumps from the Cliqz browser to their backend servers. They contain the full clickstream.

Oh, the Cliqz browser; thought you meant the search engine. Yeah, that's no secret (https://cliqz.com/en/privacy-browser, scroll down to “1) Telemetry”) and you can turn it off in the settings.

> April 29, 2020

I'm confused as to why this is being reposted now? Is this because of Brave Search acquiring their technology and being available for Beta?

Yeap...

"Brave acquires search engine to offer the first private alternative to Google Search and Google Chrome on both mobile and desktop"

https://brave.com/brave-search/

"A New Search Engine" (2019)

https://0x65.dev/blog/2019-12-05/a-new-search-engine.html

"Building a search engine from scratch" (2019 )

https://0x65.dev/blog/2019-12-06/building-a-search-engine-fr...

Out of curiosity, does anybody know why they didn't build upon an existing open-source search engine like YaCy? Is the YaCy codebase bad? Did/Does it not do what they wanted? Did they not know about it?

Startpage, Duckduckgo, Qwant, and a few others decided to write their own stuff around Bing instead of further or fork an existing OSS project. Why?

https://yacy.net/

I've run a YaCy instance at various points, but I don't think it has much to offer a company that doesn't intend to peer with instances run by other people. It's a really cool idea, but the relevance of the results it returns seems pretty rough, especially for stuff that isn't in your local index. Development also seems pretty slow, at least from my subjective perspective; I come back to it once every couple of years and everything about it looks exactly the same as when I first looked at it. It's also a royal pain to run behind a reverse proxy.

Yeah, right now YaCy doesn't seem amazing, but those companies have/had millions thrown at them. Those could've gone into the development and improvement of an existing product.

Plus, peering would allow them to improve each others results and they could build their own plugins to differentiate. It wouldn't be necessary to pay Bing.

Maybe another company will do it, maybe not.

I've never heard of Cliqz but it always makes me sad when an ambitious project fails. Good luck in your future endeavors, gents.

...this is from April 2020, and I still have never heard of Cliqz until now.

They were most popular in Germany due to their Munich offices. They had both a browser and a search engine. The search engine was built from the ground up with its own scraping tech.

Reply on news.ycombinator.com