Joel

testing this out

Public Key

npub13mzyqpxtuv4sj8pf6mpjwn6grsfm3dhy6lylcz6jdkzfya6ffexs6cj894

Profile Code

nprofile1qqsga3zqqn97x2cfrs5adse8faypcyackmjd0j0updfxmpyjway5ungjdd7x0

Author Public Key

npub13mzyqpxtuv4sj8pf6mpjwn6grsfm3dhy6lylcz6jdkzfya6ffexs6cj894

Show more details

Last Notes

2024-11-26 19:35:11

- reply

I can understand at the appview level, but I don't think they're moderating anything at the relay level, nor planning to. (And that's putting aside the possibility of multiple, though expensive, relays.)

2024-11-25 20:24:37

- reply

I'd argue the microblogging use-case is not very comparable to the bitcoin use case. Nobody splits their Twitter into several Twitters. The ratio (usually) is one microblog per one human. You can't have a "burner nsec" to be used for just one note in your timeline posted when something about where you are makes you feel a bit more exposed. And this one microblog is meant to be constantly updated, from all over the place, and integrated into lots of spots. Cold wallets for bitcoin make sense. Nothing about microblogging is cold—it's constantly warm by nature
I get that FROST represents a reasonable trade off for certain enthusiast types, but it's still very much a trade off and very much in the enthusiast realm, doesn't solve the adoption problem—because that problem is a function of the wrong use case (microblogging), not the protocol itself.

2024-11-25 17:52:22

- reply

Good points. I think you're touching on something fundamental, and that is that on Nostr you have to accept that all accounts are throwaway. And act accordingly.
The architecture mandates this. Call it architecture-mandated psychology if you want a fancy phrase. Anyway FROST isn't a solution, it just highlights the problem from another direction.
Sadly, conventional micro-blogging is a bad use case for a world in which all accounts are throwaway. Conventional long-form blogging even worse.
So the biggest problem isn't the architecture. It's the current leading use case.

2024-11-25 17:10:15

- reply

Important to differentiate microblogging and long-form. Bluesky, with its character limit, can get away with no edits. Substack, being longform, cannot, and so they have always allowed edits. Not seeing comments here make that distinction, and its critical.
Long-form: there is just no way for Nostr to present itself as a viable option for creators and not allow edits.
Microblogging: yes, a case can be made.

2024-11-25 12:29:55

- reply

I believe whtwnd.com is a functioning ATProto client, could be mistaken but worth a look.

2024-11-25 06:31:44

- reply

Sort and filter for reads within feed categories: most recent, most commented, etc. Also seems a bug with all reads showing as 0 minute reads.

2024-11-24 15:41:15

One thing that nags me testing Nostr out is I can never be sure if I'm seeing all the replies to an interesting post. I've setup a few clients, each with different (but pretty standard) relay setups, some overlapping and some not, and I've been using them to examine popular posts. Each client often presents a separate story of the life of a given post, some replies are here but not there, some are there but not here, some are missing certain types of replies (like quote replies), some aren't, like numbers can jump around a lot, and so on. I also realise there is filtering at the client level too, keeping out replies from new users and such, and that further complicates things. I hear this outbox implementation is slowly being adopted, but I don't think that really solves the problem to anywhere near the degree I would hope, as there are just too many variables at play. Maybe the idea is to just accept that—i.e when it comes to strangers' replies on strangers' posts you'll never really know how much of what you're seeing represents the totality of the story of the post?
https://image.nostr.build/fd596061c76a76eb81fefe1bc8f1f8eeca92f49b9ee7c7da2732cf93e1b858de.jpg

2024-11-24 12:50:40

- reply

Context: I've setup a different mix of popular relays on each to try and simulate the experience of a group of users. But also each client does filtering too, so that adds another "can't be sure" layer.

2024-11-24 12:35:21

- reply

You could make a partial relay, but then anyone subscribed to that relay would only know about the posts, likes replies from other people also subscribed to it. But the post would still be out in the wild and able to be commented on by strangers, so there would be lots of missed reactions.
Bluesky's thinking is a lot of people would really hate it not knowing how many likes their post got, or if there are replies that they are not seeing. And posts can get a lot of replies and likes from strangers, and from there good conversations can start. The only way to ensure everyone sees 100% of the likes and replies on a post is if there is a single relay. I kind of agree. For example your reply is not showing up on clients Primal or Iris, but it is showing on Snort and Nostrudel. I would never have seen your reply if I didn't come to Snort to check out this same post after seeing it on Primal. That is very annoying, on Nostr I always suspect I'm missing interesting replies.

2024-11-24 11:07:14

- reply

NFC cards, rings, etc., are pretty insecure though, all things considered.

2024-11-24 08:04:56

- reply

He left out to live my life as if there is a god and die to find out that, yes, in fact there is—but it's a different one and it's not pleased.

2024-11-23 15:23:27

- reply

Can't force clients to respect outbox, and if even if all clients did some users might not want to be forced by their client to publish (and reveal their IP address) to unknown relays and demand to be given a choice. Other thing is there are two types of missed replies, Type A is replies that the replying party does not want the poster to be aware of (think screenshots of X posts with chatter underneath). Type B is posts that the replying party assumes/hopes the poster is aware of but actually the poster isn't aware and never will be. On nostr many users want "Type A Native" (no screenshotting, referencing the actual live post), and that complicates things, since you need a way for users to opt for Type A Native which isn't misunderstood by the client to be Type B.

2024-11-23 15:11:49

- reply

Or maybe the article was ... suspenseful music ... edited.

2024-11-23 13:01:41

- reply

Well yeah, but on the other hand nostr's current censorship resistance, in my opinion, is due in very large part to it being off the radar, small—and not the architecture. I get the whack-a-mole argument, but I feel people are greatly exaggerating the difficulty of playing whack-a-mole in this day and age. Any government, sufficiently motivated and resourced, could make it the protocol so painful to make practical use of that, while it would still hobble along for the die-hards, the nostr vision as it stands now would be dead. (In other words it'd be a heck of a lot easier to cripple nostr than people are assuming.) Rather than censorship resistance I see the real value of nostr being in portability, though for very specific use cases that are not affected by nostr's limitations.

2024-11-23 12:42:14

Test time! I'm going to post this note here on Primal with Relay set A and then use some non-following test accounts to like it in a few other clients with certain other test relay sets, including non-overlapping, and then let's see how many of those likes total up here.

2024-11-23 09:47:27

- reply

That's not a bad thing. First of all, the fact that a nostr post might have a reply that you don't know about, or that the reply you send might never be seen by the poster, or that a post might have more likes than you can see, or fewer, or that you can't ever really know how many followers you have—all these things are seen by normal people as a ridiculous step backwards (what kind of bullshit social media works like that?). And one for which no censorship resilience can make up for. Total deal breakers. One big-ass relay fixes that. Also that article makes it seem like only fortune 500 corporations can spin up that big-ass relay, which is nonsense.

2024-11-22 12:28:46

- reply

Exactly this. In Vietnamese pronouns are determined in large part by how old you are versus how old the person you're speaking to is, and many other languages have pronouns that cannot be determined without taking into account both parties in the dialog. A more global solution would be some sort of open ended "how can I address you" field, a sort of sub-bio.

2024-11-22 12:16:32

- reply

Fair but I think it almost never costs more to censor than to allow. Humans like to be in spaces that other humans, either implicitly or explicitly, aren't allowed to enter. We'll gladly pay for that. And most humans also don't like labour, including the labour of having to create and maintain such spaces themselves (those that do are outliers). Humans love to outsource. Therefore, for an online business, setting rules drives revenue, and even if it costs more in absolute terms to censor this or that, those costs are offset by the gains in revenue from people willing to outsource the labour of curation.

2024-11-21 15:41:36

- reply

Yeah but businesses have the right not to go out of business, and free speech is, by and large, very bad for business. It's extremely hard to be in favour of the right of individuals to near-total free speech in commercial domains and also the right of businesses to make a profit. You can be in favour of near-total free speech in a charity/NGO/expensively-manifested-thought-experiment domain, that's cool.

2024-11-21 09:33:49

- reply

Bluesky isn't a protocol. Sigh.

2024-11-20 14:54:17

- reply

oops, I thought the first note didn't send. getting used to this!

2024-11-20 14:52:00

- reply

On ATproto, if self-hosting you wouldn't actually need to do that though, unless moving cloud providers or if your bare metal server died and you want to restore to a new one. The same PDS works for multiple Appviews (clients), so no need to move anything. Just that right now there is really only one Appview.

2024-11-20 10:46:57

- reply

Thing is most people run everything through a subconscious "How does this affect my food supply?" filter. Someone in 1950s America might be mortified if a nude photo of them was passed around the office, whereas someone from some remote tribe somewhere would have no issue with everyone in their world having seen them naked. Why is one mortified and one not? Because to be exposed in the first instance is to be socially looked down upon, and our brands are hard-wired to interpret being socially looked down upon as a risk of diminished food supply (red alert). Other words, all comes down to food.