😂
yeah, in general every practice that makes clients smarter with regard to where to read and write events instead of just connecting to everything they've been explicitly configured to and disregarding NIP-65, tags, etc
by centralization I don't mean blastr itself, I just mean that if all events go everywhere costs of running a relay go up = less people can afford to
There are some event kinds that are very interesting to be blasted, like kind 0s, or 10002.
Ideally, getting all events blased everywhere will be unnecessary
that's gorgeous 🧡
I see blastr like my https://t.me/pingstr_bot bot... a stop-gap solution to patch something that will be better solved later
100%
right now there's an obvious incentive to do it because almost all clients are super naive; being not-naive is extremely costly
once this is not the case, the incentive goes away
(that said, you *can* identify pubkeys using it for everything and ban/rate-limit their writes), but that's costly on it's own; I doubt that will be needed
writing everything to blastr is an anti-pattern and leads to centralization
key rotation/delegation is _the_ problem to solve
other than that, all other major features (including identity) can/should be solved in higher layers
never heard of caddy before; looks very cool! going to spawn out a new server and play with it
Thanks!
Pretty sure #madeira is a simulation
Yeah, I should definitely do that 😅
Let them go ahead and accelerate this
Apache server is an absolute clusterfuck; using nginx was such an improvement in my life
😂😂😂😂
Haha NOICE I registered nostrdevkit but that’s a mindful
Ndk.fyi 🙌
Gossip protocol and other sane behaviors to easily make non-naive/abusive clients and relays or other pieces of infra
Maybe banks should try social distancing?
?resize=1200,600
You need an extension with nip-26 support (ie not a single one 😂)
Use my PR on the getalby extension repo (I don’t have the link handy but should be easy to find)
Hopefully that’ll be merged soon 🤞
highly unlikely, but that would make my life so much easier