Well yeah, it doesn’t make sense to open your own instace and use it alone, but if you make it public it won’t matter as much because they can’t know which queries are yours. It would force you to get another ip address for the server though.
The only complain I have about searx is that the owner can log stuff, which forces you to choose a public instance carefully. If instances couldn’t log it would be the perfect option IMO.
There’s no such thing as “couldn’t log” when you’re talking about an open source project. The source code can always be modified. Even if something is open-source, you’re always trusting the one hosting it with your data.
It’s cool that they have that info (hadn’t noticed it before), though it only looks at the front end code, there is no way to check for any malicious code on the back end. At the very least, logging the IP addresses that use the search engine is trivial without any changes to the front end. Actually correlating search queries to IPs would probably be harder since the query isn’t in the URLs, but maybe not that much harder, I’m not sure.
Well yeah, it doesn’t make sense to open your own instace and use it alone, but if you make it public it won’t matter as much because they can’t know which queries are yours. It would force you to get another ip address for the server though.
The only complain I have about searx is that the owner can log stuff, which forces you to choose a public instance carefully. If instances couldn’t log it would be the perfect option IMO.
There’s no such thing as “couldn’t log” when you’re talking about an open source project. The source code can always be modified. Even if something is open-source, you’re always trusting the one hosting it with your data.
You can see at searx.space if an instance changed the code he uses.
It’s cool that they have that info (hadn’t noticed it before), though it only looks at the front end code, there is no way to check for any malicious code on the back end. At the very least, logging the IP addresses that use the search engine is trivial without any changes to the front end. Actually correlating search queries to IPs would probably be harder since the query isn’t in the URLs, but maybe not that much harder, I’m not sure.
Oh right. Well yeah that’s a shame. I bet there’s a way to make the frontend select an instance randomaly from a list and send the query to it.