I can’t argue about the historic relevance; The article you linked is from 2020, the issues from early 2019. The original matrix developing company seems to have deep ties as described, yes.
But:
If you follow the very first link I. The article you can read the history of the matrix protocol itself. It shows where and when the matrix protocol was separated from this company and what the status quo seems to be:
https://en.m.wikipedia.org/wiki/Matrix_(protocol)#History
From this it seems clear to me that the information from this article are by now obviously outdated with KDE and Mozilla two big mentioned community projects that are involved.
Wikipedia as primary source is not well suited, but the fact that the article linked to it themselves seem to show that they relied on the back then status quo.
In short: in 2017 they would be absolutely right, in 2020 there were still huge issues - but by now those are mostly addressed or are unknown.
All the claims concerning telemetry in that thread are made by the person “asking” the question.
I don’t understand the intent of this post. It looks like you’re trying to start a rumor to be honest.
Telemetry in Thunderbird exists and is well documented. If someone is not happy with their engine or approach they need to switch clients and not basically tell them “I want you to rebuild everything around my nerds”.
That’s what Arthur describes: you’re comparing from my point of view two non issues against each other.
If you don’t have the profile that either warrants:
- Google would risk business revenue by having to handle a mittm attack done by them or
- someone investing in an elaborate scheme to get a fingerprint copy
And that being worth it instead of just getting you personally is a very specific threat model where I lack the fantasy on what would warrant that.
Or to give the relevant xkcd:
And to answer your specific question: I personally went with keepass2android and have neither issues nor concerns so far.
According to their page it’s a pure searxng instance. I didn’t see anything on my own instance changing so there are three options I see:
- The mentioned server side changes (e.g. A server move you mentioned but could also be server settings, provider settings, etc).
- client side changes: somehow your Firefox provides different information to alter the results
- subjective change: it’s always a possibility that either what you searched or your perception was more fine tuned to Cyrillic.
And then there’s the obligatory “none or all of the above”.
Personally I’d guess it’s just a fluke. I gave it a few searches from Firefox mobile on “all languages” and had a mix of mainly English and a bit of German und French in there as results.
Edit: if you’re comfortable with that feel free to share some search terms and we can compare results. Would be curious myself!
As they are closed source no one can tell you their true privacy policy. It seems better than average from what I’ve read but you never know…
Personally I use logseq and sync the files via a Nextcloud instance. I can only recommend it, although I also recommend spending an hour to learn the tagging and linking logic and reading through their guide on what’s possible. I still only leverage a minor part of the potential myself.
One that is closer to onenote (I think, never used onenote) is Joplin.
In the sense of okhams razor it’s also possible that you’re just more sensibilized to the term.
It would be a fun experiment to next time first check YouTube before looking it up elsewhere, just to eliminate the chance that the information vector is before the search.
From there then come various other possibilities (from behaviour based prediction to Lemmy profe linking).
Just to widen the search area!