we’re launching our own privacy-preserving approach: one that reflects your feedback.
Not that I trust their privacy-preserving approach considering what got leaked the last time they tried it.
we’re launching our own privacy-preserving approach: one that reflects your feedback.
Not that I trust their privacy-preserving approach considering what got leaked the last time they tried it.
It is pure Ai slop that won’t be around in 3 years. I am open to counter arguments.
I looked up their GitHub repo and I don’t see the “agents.md” file, AI/LLM disclosures nor signs of LLMs in the commit history. Is there any source of information on Fluxer being slop?
When I read the repo a few months ago, I saw one guy added more than 1m of code. They are deceiving everyone and should be punished.
Isn’t this because Fluxer source code was published after the app has been launched?
I checked again today and it is 11m lines of code added. Do people not understand what it takes to code 1 million lines? PostgreSQL is around 1m. PHP the language and runtime are 1m. The man hours required for that much code is ridiculous.
People suspected he’d uploaded Fluxer in one big chunk to hide LLM usage. He admitted to using some AI help initially but claimed that he would use it sparingly.
This new issue is that he seems to be using LLMs heavily, based on the fact that he uploaded a million lines of code last month alone.
https://github.com/fluxerapp/fluxer/graphs/contributors?from=8%2F22%2F2026
It’s free and open source, and I don’t think perfect should stand in the way of good. If my group needs to migrate in 3 years, so be it; it’s working now and isn’t a Microsoft product.
Okay, trust the security of something with over 11m lines of code added in 2026 by one person.