eris
banner
nihili.st
eris
@nihili.st
they/them
oh look it's the guy who took dark money and didn't disclose it
September 12, 2025 at 2:24 PM
shut up loser
September 11, 2025 at 5:24 AM
is there real recourse for Mississippi users at this time? do we have any reasonably functional third party bsky appviews? I'm not sure how much the situation has changed since I last looked into everything
August 23, 2025 at 1:37 AM
for now, they bypass the protocol until the devs implement private stuff. so they work like any other centralised web app
August 13, 2025 at 9:20 AM
it doesn't support private anything for any reason at this time
August 12, 2025 at 11:01 PM
I mean if you can't properly resolve and verify a handle, you probably shouldn't be interacting with it as a safety thing. you can also probably rewrite the DID to be more friendly, like <plc id>.plc.directory
August 12, 2025 at 8:48 PM
probably just pretend something to the front of the handle since it's most of the way there already. either some static thing like 'at' or the user's DID or something. would depend more on the use case tbh
August 12, 2025 at 8:31 PM
am I using it right now?
July 29, 2025 at 4:11 PM
I'm not entirely sure what the point is in soft blocking and why it would be used over regular blocking. surely some of these annoying people would eventually create a troll's appview that doesn't respect these records
July 29, 2025 at 4:11 PM
which may be why it's all a bad idea
July 29, 2025 at 3:57 PM
I assume they mean the relay?
July 26, 2025 at 9:39 PM
support for private data is a prerequisite for on-protocol DMs, so who knows when that'll happen, but afaik it is planned
July 24, 2025 at 10:14 AM
if you would like to become a sysadmin in exchange for purely theoretical benefits, then hosting a PDS is a great option
July 11, 2025 at 2:38 AM
I will ask my grandma to verify my PGP keys then
July 8, 2025 at 1:08 PM
Why
July 2, 2025 at 10:58 PM
I don't see #2 as a viable path since the short form video space is already crowded with big players and it's going to be hard to make something substantially better than tiktok (which succeeds imo because their algorithm is so good)
June 30, 2025 at 4:14 AM
feed algorithms have always been black boxes and selling points for social media apps so the assumption that people would provide good quality feeds for the network for no reason really never made sense to me. maybe monetisation needs to be better proven for protocol services?
June 30, 2025 at 2:14 AM
yeah, afaik you can't filter the firehose. there are some tricks you can do for some optimization, like if you want to listen to a certain DID you could look up its PDS and listen to that event stream instead of a relay. I'm kind of surprised the relay doesn't have that functionality by now tbh
June 24, 2025 at 8:10 PM
I'm not currently aware of any PoC for malicious Jetstream but I'm also not super in the loop lol
June 24, 2025 at 7:57 PM
I'm sure there are ways to verify the data from Jetstream but imo once you start asking these questions, it makes more sense to consume & verify the binary firehose instead. then you don't need to worry as much about hammering PDSes hosted on pis, etc. and ideally libraries would make it easy
June 24, 2025 at 7:57 PM
personally, I think a malicious PDS would be bad since it has full control over your identity & keys and also should act as a mitm for all on-protocol traffic. so it could modify/refuse incoming or outgoing posts, inject ads into feeds, lock you out of migrating, post or delete for you, etc
June 24, 2025 at 7:19 PM
yeah, Jetstream provides an easier way to consume protocol events at the cost of stripping out all of the crypto stuff that allows the events to be self certifying. pretty much any untrusted server or any other way to interfere with those events can go undetected.
June 24, 2025 at 7:19 PM