Avatar
sommerfeld
d0debf9fb12def81f43d7c69429bb784812ac1e4d2d53a202db6aac7ea4b466c
Sovereignty, freedom, privacy, self-hosting and software dev 8BETLojm7zCfGV8DKKiRWLAXrFGbkZxDAgzpGmkrDvafTB18QFUg9cQ1KKDtcECrekV14yZDdGEEQbGy8ShXq1pDBVAh2Hw

I've been piecing it together in the past months by talking with people. Before I had no idea how common this was.

It's crazy that most IRL people around me are secret gambling addicts who regularly waste thousands per year on lotteries, sports betting and online casinos.

These people pretend to go to the bathroom just to secretly place a bet on their phone. Pretty sad, but also shows how miserable and desperate most are.

True, didn't think about that. Maybe that forces regular visits, which might be a good thing 😄

1. Buy pixel

2. Install grapheneOS

3. Do not install google playstore, it's optional

4. Create a non-admin profile for him

5. Done, nothing else can be installed by him

Regarding phone calls, I don't know if that is possible but it would be an interesting feature to implement (a whitelist for calls instead of a blacklist), you can suggest that to the devs, they are very responsive.

Hear me out: with NIP46 remote signing, it's safe for your own account to sign received zap events, i.e. your npub could be the zapper npub.

The implication is: we can make zaps more trustworthy if their signed by the npub that claims to have received it. Right now lnurl provider zappers usually use their own random keys.

This could even work for custodial wallets! A custodial user can provider their custodian a nsecbunker URI that only has permission to sign zap events.

Am I crazy? nostr:nprofile1qqsr9cvzwc652r4m83d86ykplrnm9dg5gwdvzzn8ameanlvut35wy3gprpmhxw309akk7mnpvshx5c34x5hxxmmd8gurqwpsqyxhwumn8ghj7mn0wvhxcmmvqyvhwumn8ghj7urewfsk66ty9enxjct5dfskvtnrdaks7vj9m5

I gave you a straight answer: storage is suboptimal and fragile, not my fault if you don't believe me.

If you want more reasons:

* aarch64 is a shitty architecture: bootloaders are different for each hardware and are super proprietary, hard to configure and just terrible in general compared to standard x86 bootloaders.

* there's less software support overall for aarch64

* rpis heavily depend on proprietary firmware for most basic functionality, almost nothing is upstreamed to the linux kernel

Making it public involves some extra dev work and docs to make it pretty and minimally acceptable 🤭

Should I publicly release a super niche piece of software that is only useful to me? 🤔

Let's make it take 100 years on purpose then 🫠

Storage is usually the first thing to go bust, specially if using an microSD card for the OS. But even USB connected storage is slightly better but still shitty.

I'm speaking from my personal experience and many others. You're just running it for 1 month. I was like you before, check my earliest notes, I would die defending my rpi.

Won't cut it because:

* it's not as cheap as before. For the current retail price, you can buy decent x86 servers

* you can run it on a pi, I did it up until recently, but it's a shitty node and a shitty experience and bound to break in less than 2 years

You can't run an indexer or block explorer with a pruned node.

I wouldn't count that as a success. A success would be if it was the standard today to distribute large media instead of large CDNs and server farms.