(2023-12-18) On passion ----------------------- There is a huge difference between passion and motivation, however big the latter is. With motivation, you can clearly explain why you're doing what you're doing. Passion is when you generally can't do this. You still understand why, but can't explain. At least in a way that general public will get your point. But if you really don't know why you're doing that, that's not even passion, that's called stupidity. There are some things I'm motivated to do (e.g. in order to survive) and some things I'm just passionate about. For instance, I can logically explain why I started learning Toki Pona but I can't explain why I recently created the Toki Pimeja encoding for arbitrary binary data. If I try to explain this, I'll look like a crazy person to anyone who doesn't know the premise of what I'm living in now, and even to those who do. This is an actual problem for me because I can't find same-thinking people to help me with some of my initiatives, just because I never can fully explain what they are for. This can lead to me running out of passion for a particular thing that completes an already long list of abandoned projects of mine. GerdaOS is an example of such a thing. It started as a project to essentially make the Nokia 8110 4G stock ROMs more secure and give KaiOS back to the community in terms of liberating it from the KaiStore lock-in. The eventual idea was to turn it into a custom and more secure KaiOS distribution that gives its users much more control over their devices than they have, making it a keypad smartphone platform to hack and build upon. But, while being overwhelmed by work, I ultimately couldn't deliver that idea to other community members. Most of them didn't even want that degree of freedom and security, they just wanted a less laggy system that could also run WhatsApp and other proprietary BS. Meanwhile, I tried to integrate at least some VoIP but the KaiOS 2.5.x API is so buggy that I couldn't even acquire the earpiece control the way it should have worked 99 times out of 100, only the multimedia speaker. Even WhatsApp does this via some system .so library, not the Gecko API. But that was the least of my problems, and eventually, I couldn't even understand who my target audience was and whether it was worth to continue any effort in that direction. By the way, Kopher and RCVD are relatively new but too niche, so my last "big impact" KaiOS applications appreciated by the community were CrossTweak, FastContact and FastLog. But I wrote them for myself in the first place, and I did so before finally realizing KaiOS was a dead end from the start, at least for my true goals. But why? One word: complexity. Linux kernel is extremely complex. Android is extremely complex. Any modern browser engine is extremely complex. So, KaiOS, being a combination of Linux kernel + Android base system + a suboptimal UI running on top of a browser engine, is a nightmare to even start thinking of building any really secure communication system with. And let's not forget it is also running on a piece of smartphone hardware which itself is obscure enough to not trust it to build such a system, not to mention how complex it is as well. This is the closest I could get to any logical explanation about why I ran out of passion of developing anything for KaiOS anymore, and I tried to share these thoughts with the KaiOS community in their Matrix bridged with Discord. I was met with dead silence. No one was interested in this. I even tried joining some other Discord chat dedicated to featurephones, but it turned out to consist of straightaway noobs who couldn't even get that any commercial non-UMS9117(L)-based "featurephone" that supports LTE nowadays actually is running a stripped-down Android version on fully smartphone hardware. There was nothing to discuss with them. Other groups who claim to develop "secure phones" use the same approach of putting modified Androids onto smartphone hardware. If you ask me, I really am stuck alone with my vision right now. I think some answers might have been given by the first luxury XOR phones and their firmware, but this is something I really doubt will appear in any leaks anytime soon, and, of course, you can't order the first model from the official sources anymore. SC6531E isn't something I'd view as a platform of perspective, and SC770x and UMS9117, as I said, really are hard to crack even on the handshake level as of now. On the other hand, I do have plenty of MT6261-based devices to work with (and even some gigabytes of leaked MAUI sources to study) but don't have a reliable open-source way to flash any memory area yet, only the BROM method to load and run any code from RAM. I also have some SIM800 modules that can be soldered into a phone of my own design... but that's exactly what I want to avoid. I want my solution to eventually be able to fit into some of the cheapest featurephones on the market, with these official Chinese Nokias like 130-2023 being a good target as they are available internationally. Don't get me wrong, they still are incredibly complex, but first, alas, only MIPS-based CT8851/SC6533G (which no major brand uses and they are being phased out) are simpler than that, and second, they still are orders of magnitude simpler than any smartphone hardware and corresponding firmware. Maybe there really is no logic in this vision and doing all this. Maybe there's no logic in researches aiming to create my own firmware for such phones (with the planned application scripting language being based on TRAC T64 like nntrac). Maybe no one will appreciate it. But, as I said before, I'm sort of a dreamer, and this particular passion of mine takes a lot to put off. --- Luxferre ---