Hacker News new | ask | show | jobs
by phh 1929 days ago
The offers of GrapheneOS and /e/ are very different. /e/ targets to be a usable alternative to a google-owned software, which includes for instance cloud storage and backup, access to third party proprietary apps, etc. While GrapheneOS is harder to use, but it does provide much better security hardening.
1 comments

hey there phh; slightly tangential, but: What are your plans for phh-treble? To be honest, if you've got the bandwidth, I'd love for you to start something comparative to /e/ and CalyxOS.

With Generic Kernel Images (and APEX to an extent), do you see phh-treble likely replacing LineageOS as the preferred base for projects like /e/? Merci.

My goal for Phh-Treble remains the same, which is provide a good basis for a pure unmodified AOSP with good hardware support, so that other people can make whichever ROM they want with it. As such, I'm saddened that many ROMs, including /e/ still doesn't officially provide GSIs...

I haven't had a good look at either /e/ or CalyxOS, but /e/ looks like they are too focused on the marketing for my taste, /e/ really isn't known to be a contributor in term of source code to the community, their only contribution is bringing users (I do have a bit of respect for that). CalyxOS has nice things, it is much closer to what I'd want than GrapheneOS. CalyxOS tries to makes features closer to what you can get on a "standard" google-owner Android, and they are definitely a great value to the community, like SeedVault is a really nice addition for de-googlizing one-self. I feel that CalyxOS goes too much into security, while privacy/data-owning are already very nice features, and they restrict their user-base because of that.

I doubt that GKI would help phh-treble replace LineageOS. First, I'm happy there are LineageOS GSIs based on my work :-) But the real reason LineageOS persist is that most users want a ROM tailored for their own device. Sure one of the biggest real issue of phh-treble is the lack of security upgrade of the kernel, but from my experience discussing with many people asking me for device-specific ROM, the reason they want device-specific ROMs is really about feelings.

Most users feel a device-specific ROM will be more optimized. (no they are usually not Gentoo users). Usually they'll come to me with "hey, you have this feature that works on my device in GSI, but not in my device-specific ROM, could you help fix it?", when I ask why not just use my GSI, the answer goes with optimization, and the obvious answer is "It's so much more optimized that hardware features doesn't even work.".