19M from Germany

  • 0 Posts
  • 91 Comments
Joined 3 years ago
cake
Cake day: June 17th, 2023

help-circle



  • But how is that significantly more secure than LineageOS? I have read through countless blog posts from GrapheneOS developers and have not yet encountered an explanation that is sufficiently convincing. Outside of additional security hardening, which is definitely a big pro, GrapheneOS doesn’t have many things that LineageOS doesn’t. LineageOS is fully FOSS and telemetry-free. They introduced the “Trust” control panel for managing all sorts of privacy and security matters. They have PIN scramble.

    The only major, obvious security vulnerability lies in the proprietary driver blobs from the device vendors / OEMs. But AFAIK Google Pixels also have those, right? So outside of doubtlessly valuable measures like restricting malicious reprogramming / access through the USB port, in what ways is GrapheneOS actually more secure than LineageOS?





  • Sorry for the downvote, but I see this take repeated here on Lemmy so often and it just makes no sense. This will not kill the FOSS app “ecosystem”. Nothing whatsoever changes for FOSS ROMs like LineageOS or GrapheneOS. And as long as there are FOSS operating systems, apps will be developed for them. If anything, this could drive mainstream adoption of free/libre Android forward, re-invigorating the scene through public outcry.

    And to the people who propose fully jumping ship from Android to “Linux phones” because of Google’s recent changes, you would only make the app support matter worse. As someone who daily drives both a phone with LineageOS and one with postmarketOS (mainline-ish Linux), mobile app support is endlessly worse on Linux than the fallout from Google’s developer registration could ever be. That is not to say that Linux phones will not eventually get to a point of reasonable maturity, but it is way too early and frankly utterly irrational to bury AOSP Android or needlessly hate on it.



  • This is not just one of those ivory tower papers with their actual applications far away in time and eventually ending up in some obscure industrial process never heard of again in lay circles; this could have an immediate impact on the maker culture and makerspaces right now and in the near future. The preprint describes the process in a very understandable, digestible manner and provides actual implementation examples, as well as detailed recipes for all of the compounds. If you are even remotely interested in the subject matter, I’d recommend you to try it out for yourself. The “ingredients” are all easily obtainable and handleable. Yes, gallium and indium might be a bit expensive, but it is worth it imo. They literally used consumer kitchen equipment for some of the steps, to demonstrate how this is feasible for tinkerers, makerspaces and prototypes. No expensive machinery required (except for an FFF 3d-printer, of course).






  • No. Education will always remain the way out of poverty. And regardless, no “AI” can replace actual programmers at present. Their code and code quality are entirely unreliable and not suitable for serious, production use. They may be sufficient for hobbyist applications, but for software that is actually getting deployed, LLM outputs vary too widely, and you will always need experienced programmers to monitor them and correct errors. Also, an AI can not come up with a coherent design principle for you, the individual modules and moving parts it spits out will invariably fail to work together at scale. Creating software is much more than just churning out code. It requires advanced reasoning and specific knowledge, and AI is not there yet, and who knows if it ever will. All companies that are firing (or not hiring) programmers over current day LLMs are going to fail.