david

That’s always a possibility. If it becomes a problem, like if you see your extended WoT network grow by 20k bc someone 4 hops away from you followed 20k spammers, you just unfollow and mute that pubkey. Or maybe one or more others in your network find the problem before you do and unfollow / mute the bad actor, and the problem is solved before you’re even affected.

2024-09-08 19:14:14

- reply

I’ve watched this movie almost as many times as I’ve watched the Godfather.

2024-09-08 18:53:11

- reply

Just ran it, did a little mini zapathon and seems to be working so far! ⚡️ ⚡️ ⚡️
🙏🏼

2024-09-08 16:13:47

- reply

Mostly use it in the US without a VPN.
I just changed Paste from Other Apps from ask to allow — wonder if that has anything to do with it (maybe WoS not allowing damus to paste the sats amount?)

2024-09-05 00:25:38

- reply

I gotta start working in some archer pushups 🏹

2024-09-01 22:54:27

- reply

I was excited to watch it — so many great actors — but ended up very disappointed. Seems like the moral compass of the show is all over the place

2024-08-30 10:30:32

- reply

The idea of the orange check is that you pay some sats to the platform, right? Like if he owned Xitter that’s what he’d do. Not sure how to do that on nostr where there is no single dominant platform. I suppose zaps could work, although they’re not verifiable (except for ecash, if you trust the mint?)
Ideally I’d want the ability to synthesize zaps + other forms of social proof into a single score. Which is how #grapevine works.

2024-08-29 05:03:12

- reply

Yup I do the same! I think I’ve had to do up to 4 days at a time 😅

2024-08-28 16:01:56

- reply

Totally fucked up

2024-08-28 13:49:26

- reply

Start out going to listr.lol and see if you have a list with “Grapevine” in the title

2024-08-28 12:33:28

- reply

You’re 🎯 about the hot data mess. No way around that, ever. We may as well come to grips.
Which means that as a consumer of the hot mess, you’ll have to interpret data that comes packaged in multiple formats. Each external format needs to be interpreted into a *single common format of your choosing* so that you can synthesize them all together in a way that is of maximum utility to you.

2024-08-27 09:43:24

- reply

Content discovery is a huge pain point on nostr, and this is a problem I want the grapevine to address. The grapevine score is designed to serve as a weight, useful for weighted averages or vote tabulation. If we have a system where you can either self-attest you like to write about topic X or recommend others as good follows for some topic (maybe badges can do this?), the grapevine can synthesize that info into recommendation lists. If Alice wants to find great follows for SomeTopic, her grapevine will have a rec list for that, and you’ll probably be on it if there are indicators you should be.
#WoT
#grapevine

2024-08-26 09:55:12

- reply

I know there are quest and labcorp labs that will take any paying customer, and you can prepay at walkinlabs. You select the labs you want and select labcorp vs quest. But other labs are tied with specific offices so you can’t do that. I had presumed that general purpose quest and labcorps are everywhere in the US but maybe not in your area? 😬

2024-08-25 22:17:59

- reply

Andreas was the GOAT back in the day. His Money as a System of Control is one of those vids I have to rewatch from time to time.

2024-08-25 16:30:24

- reply

allcapsplzicanthearyou

2024-08-25 16:25:23

- reply

walkinlab should allow you to order and pay for a big list of tests at quest or labcorp. Although some states won’t let you because of, ya know, reasons. In which case I think it works just to change your home address to a diff state.

2024-08-25 09:04:53

Does nostr have an app to create polls?
#askNostr #note1sl3…g02k

2024-08-25 09:04:36

Wishing I could do a poll of @npub1the…4dha members:
- do you run a lightning node using BTCPayServer?
- do you use a VPS or host the node at home?

2024-08-23 16:30:14

Demo of the Grapevine is now LIVE!!
Check it out at:
https://brainstorm.ninja/#/grapevine/actions/export
🍇🔥🍇🔥 LFG 🔥🍇🔥🍇

2024-08-22 16:54:13

Awesome car show on a beautiful day in Riga #nostriga https://i.nostr.build/lhc2O7Xrsjxp4KdF.jpg https://i.nostr.build/YBYUtVVDmo2PMzkN.jpg

2024-08-22 05:06:45

GM and PV! ☕️ 🌧️ #nostriga

2024-08-21 05:30:19

- reply

GM Derek!

2024-08-19 13:34:30

- reply

Can you mute a string with a leading or a trailing space? “nostr “

2024-08-18 23:15:23

- reply

You can learn about the Grapevine WoT Score (aka the “Influence Score” — terminology is in flux) at https://brainstorm.ninja/#/grapevine/influenceScore. @npub1man…n9tn is currently working on streamlining the calculation of those scores so they can be more easily exported as a NIP-51 list. Right now the list export page is hidden under developer mode but hopefully we’ll be able to unveil it soon.
If I’m understanding what you mean by follows + follows WoT implementation, I’m guessing you’re making a list of all pubkeys who are either 1 or 2 degrees of separation (DoS) away from me? Brainstorm actually calculates three types of scores: the main Grapevine Score, the DoS score and the more traditional WoT score. (Calculations page is at https://brainstorm.ninja/#/grapevine) I think each score type will prove to have its own set of utilities, with the Grapevine Score being ideally suited as a foundation to build contextual recommendations. Definitely better than the traditional WoT score, which is too much of a popularity contest.
https://brainstorm.ninja/#/grapevine/wotScore
https://brainstorm.ninja/#/grapevine/dosScore

2024-08-18 22:59:06

- reply

I’m def hoping to learn more about DVMs and their applications at Riga (I’ll be there).
In the meantime, to keep things simple, we’re streamlining the export of NIP-51 kind 30000 lists. Right now, we’re generating two types of lists:
1. the Grapevine WoT List
2. New Follow Recommendations
List 1 will be the top N pubkeys as determined by the Grapevine WoT Score. This is fully personalized. A user in a community centered in China or Japan will get a different list than someone embedded in the mainstream nostr community.
List 2 will be the same as list 1 but minus pubkeys you already follow.
So any client that supports NIP-51 lists will be able to benefit from Grapevine-generated lists.
Once sufficiently validated, next step will be contextual recommendations, eg: top 10 experts on homesteading (or whatever) as per my Grapevine.

2024-08-18 00:25:46

#100aDayUntil100k
#100pushups
50 + 25 + 25 regular
35 + 35 + 30 decline
Days 196✅, 197✔️

2024-08-17 22:15:55

- reply

This is why I see it as a Venn diagram. Super easy to meet one requirement without the other:
cypherpunk but no PMF
or
PMF but not cypherpunk
The challenge is to achieve both at the same time.

2024-08-17 22:12:51

- reply

💯

2024-08-17 21:57:01

- reply

It’s always a challenge to balance the “purist” approach with the practical realities. I imagine a Venn diagram, a small set of actions that allow us to be completely true to both.

2024-08-17 20:41:58

I’ve been thinking a lot lately about how to merge the Grapevine with a relay service. In other words: a relay that monitors the network, calculates your Grapevine WoT scores, makes them available as NIP-51 lists, and updates them periodically.
A feature like this would really make a relay stand apart from the rest, I think. And it’s a natural pairing, because these scores will help you decide which notes to seek and store and which to ignore.
This could work with centralized relay services, like nostr.wine, or personal relays, like relay.tools.
Free tier = basic Grapevine WoT Scores
Paid tiers = track scores of more pubkeys, in more contexts, with more frequent updates (once a day? once an hour? Continuous?)
Any relay gurus who have an interest in spitballing and hammering out a product spec — send me a DM. I’ll be in Riga if you’d like to meet in person.
#WoT
#relays
#nostriga
@npub18kz…x5sz
@npub10np…tl5h

2024-08-17 19:57:46

- reply

But suppose I’m just a lowly pleb, not a dev or a builder. How do you convince me to pay for a personal relay? Without appeal to abstract principles like censorship resistance? What will I be able to do with a personal relay that I can’t do without one?
Or perhaps you’re not necessarily targeting lowly plebs or devs, but somewhere in the middle: regular business owners who want to run a relay for their business and need to customize it and want to make it as painless as possible. Not sure I see the need for that but maybe I just need a little imagination. Pretend I am running a restaurant. Will I some day have a need for a customized business relay?
I ask bc I see web of trust as an answer to the question: I will want a personal relay to keep track of my grapevine WoT scores. But maybe there is more that I don’t yet appreciate.

2024-08-17 19:31:25

- reply

My question is this: why would anyone go to the trouble of running a personal relay? Assuming that the altruistic motivation to decentralize nostr doesn’t count, bc it isn’t sustainable.
Here’s my best answer: build a personal relay that calculates your Grapevine WoT Lists and updates them in real time. I can see this becoming a MUST HAVE.

2024-08-17 19:08:10

- reply

I use damus, and I can update my relay list on damus. Works as far as I can tell.
Not long ago I realized I had to quit and restart damus for the updated relays list to take effect. Not sure if that’s still necessary or not.

2024-08-17 19:04:53

- reply

Def needs to be easier to do home serving. But those tools won’t exist without a demand to support them. Perhaps nostr could add to that demand. Get users to run personal relays in the cloud, then get users to set up relay backups at home on start9 or umbrel, then watch users start asking for the tooling you mention.

2024-08-17 18:54:35

- reply

You can’t use other relays on damus?

2024-08-17 18:27:09

- reply

What additional relay configuration options would you like to see as a user?

2024-08-17 18:21:29

- reply

This is a valid and important point. But we must constantly remind ourselves that users will not do things like run personal relays at home for ideological reasons. They’ll do it if it’s easy, cheap, and awesome as a consumer.
How about this as a bridge:
1. I set up something like relay.tools which is quick and easy
2. I set up a mirror relay at home on an umbrel or start9, as a backup to my relay in the cloud.
If my home relay goes out and I don’t notice for months or years (which has a high likelihood of happening), that’s ok bc the cloud relay is doing the actual work. If the cloud service rugs me or comes under the thumb of the state, that’s probably ok bc I probably have it all backed up at home, and maybe now I’ll be motivated to make sure my home relay stays connected.

2024-08-17 17:55:14

- reply

This would be a potential use case for the NIP-51 Grapevine WoT List. Two ways to do it:
1. Completely ignore zaps by npubs that are not on the list.
or
2. Ascribe a weight to each zap, where the weight equals the Grapevine WoT Score of the zapper. Optional: make the weight an also function of the amount of the zap. Maybe weight = sats * Grapevine WoT Score.

2024-08-17 17:41:47

Got my first relay set up using relay.tools by @npub10np…tl5h . Woot!
Relatively easy to set up. I love the configurability. Some of the options include:
- public or private
- free or paid
- keywords, tags, and pubkeys to include or block
Validation: I see there are at least ~ 30 searchable paying customers so far.
Here’s what I’m trying to figure out. For this to be a profitable business, who will be the customers? Businesses running RaaS (relays as a service)? Or everyday plebs?
If I wanted to run a RaaS, looks like this would be a super easy way to get it up and running quickly. So that seems a feasible model. Although I wonder how many customers like that exist. Probably not a lot. Enough to sustain relay.tools as a business? Maybe not.
As a pleb … why would I want this service? Maybe to have a guaranteed backup of the stuff I want. Nice to have. But do I NEED it? Not really.
So here’s the question: How do we turn a service like relay.tools into something every nostr user MUST HAVE?
cc:
@npub12zq…99h8
@npub1man…n9tn

2024-08-17 14:35:05

Hello relay.tools 🛠️

2024-08-17 12:41:19

GM and PV! ☕️ ☁️
Who else is excited for Riga? 🇱🇻 it’s coming up soon!!

2024-08-15 21:23:56

- reply

“In the web” means connected by follows? brainstorm.ninja calculates several scores: the Grapevine WoT Score, the standard WoT score, and the degree of separation (DoS) score which equals the minimum number of hops it takes to get from me to Alice, where a hop = a follow.
It takes a lot of resources to calculate these scores, as it sounds like you’ve observed. That’s one of the biggest challenges to bringing WoT to nostr.
And it’s why I am building brainstorm.ninja: we calculate these scores and export them as NIP-51 lists so apps like yours don’t all have to calculate them from scratch. That way, users will be able to experiment with the various methods of list generation and see which methods work best for them. All with no extra work on the part of clients like Voyage! 😃

2024-08-15 20:49:55

- reply

If you mean NIP-51 lists, then cool — it means Grapevine WoT lists should work for Voyage feeds 👍🏻
btw how does Voyage calculate WoT? Is it the same standard method as most other clients? Which is: calculate my follows & your followers; calculate the intersection of those two sets; the resulting number of npubs = your WoT score; define a cutoff and anyone above the cutoff makes the list.
I ask bc Grapevine WoT lists use a different method than the standard one. I’m hoping to demonstrate that it will have greater versatility and utility for discovering new content.

2024-08-15 20:03:38

- reply

Yup, @npub10np…tl5h has tested custom Grapevine WoT lists on Amethyst and they appear to work!
#nevent1q…zlqu

2024-08-15 19:33:03

This is going to be an important step towards web of trust adoption. I’m excited to see it happen!
#WoT #note1hwv…qr85

2024-08-15 19:12:31

- reply

Woot woot!! 👏🏻

2024-08-15 19:11:37

- reply

I 💜 #postr!

2024-08-15 19:10:57

#100aDayUntil100k
#100pushups
25 x 4 regular
Day 195 ✔️
Interspersed with outdoor sprints and back exercises
#postr

2024-08-14 19:45:38

- reply

Cool. Have you built stuff with unity before?

2024-08-14 19:21:01

- reply

I tried making a 2D visualization tool at one point but it slowed down a lot once I got over something like 500 npubs
#note1ylq…jpyj

2024-08-14 19:09:20

- reply

The amount of data available to us is almost overwhelming. And so the challenge is how to wade through it. No single answer to that question. But the method I’m using to calculate Grapevine WoT Scores is designed specifically to synthesize multiple sources of data into a single score. What you can export right now at brainstorm.ninja is calculated using follows and mutes. But my plan is to add the ability to export lists generated where the scores incorporate different types of data - zaps, reactions, etc - and generate contextual scores. You can’t export them yet but you can see them at the Content Discovery section of the site. Eventually you’ll have a control panel so you can decide which data sources you want to use and how much weight you want to give to each data source. I call it “interpretation” bc you’re basically deciding what each type of data *means*.

2024-08-14 18:52:24

- reply

Something like that would be super cool 😎

2024-08-14 18:43:23

- reply

Cool. Gonna make it an option

2024-08-14 18:30:10

- reply

What would you say about an option to make a list of the top N pubkeys, but first remove everyone you’re already following? As a way to discover new accounts. The top N new people to follow as recommended by your Grapevine. View the feed on your fav client and see who catches your eye.

2024-08-14 17:35:07

- reply

Manually would take waaaayyyy too long! I have a tool that publishes the list — currently it’s hidden bc it’s still in development. But if you go to settings, Developer, toggle on Dev Mode, then under your user icon top right go to Hello World, then Test Page 7 you’ll find the Export Influence Scores test page. Click the Publish button and a popup should tell tell you it’s been published. Then go to listr and you should see it. Right now it publishes the 1000 highest ranking pubkeys.

2024-08-14 14:33:36

#100aDayUntil100k
#100pushups
25 + 35 regular
40 x 3 incline with weights
20 decline
Days 193 ✔️, 194 ✅

2024-08-14 14:31:35

GM all good people! ☕️ 🌞 💜
https://music.apple.com/us/album/ive-seen-all-good-people-a-your-move-b-all-good-people/1049017157

2024-08-14 13:32:18

- reply

My plan eventually is to make a service that updates the list dynamically. For now you redo it by hand.

2024-08-13 19:49:57

- reply

Cool, makes sense. I def like the idea of using these scores for new/different content. Have you checked out the Content Discovery app on brainstorm? The system is set up to make recommendations in any context. I just need to add a button to export the lists. These lists should be much shorter but much more focused. Right now for example there are 4 users on the list for the “philosophy” context, generated bc people wrote nostrapedia articles on the topic of philosophy and then other trusted users liked those articles.

2024-08-13 19:02:38

- reply

Awesome! I think it’s going to be fun to think about what to do with scores in ranges like that. Your goal is to eliminate spam from your relays, which means prohibiting anyone with a score below some cutoff. But what is the cutoff that will best serve your purposes? Do you find out empirically? Pick a cutoff from first principles? Or use some other method? And what other data besides follows and mutes would you like to incorporate into the score? That’s the beauty of this method: it can synthesize data from multiple sources, and you can weight each source in whatever manner best suits your purpose. For example, I haven’t included zaps yet but might do that next.

2024-08-13 18:45:21

- reply

Let me know if you want to increase the number of profiles per list. Or any other changes or features you want to see.

2024-08-13 18:30:49

- reply

WOO HOO!!!! 🔥 😃

2024-08-13 18:24:02

- reply

If it still doesn’t work, go to Grapevine Settings and make sure none of the parameters are NaN — I think that is what was causing the problem. Reset button at bottom should do the trick.

2024-08-13 16:48:11

- reply

I think I’ve fixed the NaN bug - one of the variables used to calculate influence scores was not being updated properly. Log out and back in and should be fixed now.

2024-08-13 13:28:59

- reply

How do you calculate what’s trending? Will Alice and Bob see the same set of trending notes or does it depend on their follows list?

2024-08-13 13:25:33

- reply

It’s not always obvious what a sensible bounty ought to be. Would be nice to have some sort of price discovery system. You define the goal, offer to pay X, and ppl can counter saying they’ll do it for Y. Or maybe people can just chime in saying “a fair price for this job would be such and such” even without necessarily offering to take the job themselves.

2024-08-12 11:37:56

#100aDayUntil100k
#100pushups
30 + 40 + 30 regular
Day 192✅

2024-08-12 11:37:05

GM and PV all good people! ☕️ 🌞

2024-08-12 03:05:30

- reply

Usually I’ll end up with something like 35k downloaded, maybe 34k with a nonzero influence score. Anyone connected to you by follows will have a nonzero score, even if many many hops away

2024-08-12 02:34:24

- reply

One thing you can do: on the table, filter using multiple columns. Influence above 0.5, DoS less than 4, stuff like that. And top left of table shows how many profiles made it through your filter. Perhaps I could add a button to export the filtered table as you see it, as a nip-51 list … ?

2024-08-12 02:31:41

- reply

Once I figure out how to fix the NaN bug, maybe I’ll work on a control panel so you can better tailor the lists to your needs. Let me know if there’s any particular feature that you’d like to see!

2024-08-12 02:28:07

- reply

Ok cool, glad you found that! And looks like I’ve recreated the same error as you (see my other post).
DoS and WoT scores should be integers, but Influence, Average Scores, Certainty and Input should be float. A lot of influence scores will be either 1 (almost certainly not a bot/scam/bad actor; ie a “regular user”) or 0 (bad actor until proven otherwise). A lot of influence scores will be 0.9xxx, meaning probably not a regular user, but maybe less than 1 bc got dinged by a mute or two, or maybe just doesn’t have enough high quality follows yet to have achieved a score of 1.000

2024-08-12 02:19:44

- reply

I just tried using a different profile on chrome and there is an error calculating the influence score — all are 0, except for my main profile which is NaN. Strange. I bet this is encountering the same error you’re getting. Maybe bc I had previously logged into chrome with my main profile? Some memory problem (I’m using redux) and it’s not being sufficiently cleared at logout, like maybe redux is showing two profiles labelled as logged in … Also I see an error calculating the certainty, which is a step in calculating the influence score, bc all the certainty values are NaN (except for the logged in user).
I’ll have to dig in to see why that’s happening. Won’t be able to tonight though. In the meantime you could perhaps try several diff browsers — I’ve been using brave on macOS mostly and that calculates scores correctly for me. Or try a browser you haven’t used before for brainstorm. https://i.nostr.build/jijQxtDG2XnNQFk3.jpg

2024-08-12 01:54:40

- reply

https://i.nostr.build/D5446KiD2aJAblP7.jpg

2024-08-12 01:53:27

- reply

Are they all NaN or null in the table?

2024-08-12 01:51:40

- reply

I see why you wanted the NIP-51 list — you already had NIP-51 imports working.
Perhaps I should build a control panel so you can decide who goes on the exported list, either top N or cutoff by some threshold. And I can add the DoS and WoT scores to the list — maybe optionally since that will increase file size.

2024-08-12 01:41:04

- reply

Although we still gotta get the scores fixed … those are probably the wrong 1000 pubkeys!

2024-08-12 01:40:09

- reply

WOOHOOO!!!

2024-08-12 01:35:15

- reply

Oops, wrong url
This should work:
https://njump.me/naddr1qqfkjmnxd36k2mnrv4fkxmmjv4e5c6tnwsq3gamnwvaz7tmjv4kxz7fwv3sk6atn9e5k7q3q0npj3gydmv40m70ehemmal6vsdyfl7tewgvz043g54p0x23y0s8qxpqqqp6nqhgqxju

2024-08-12 01:33:48

- reply

And on njump.me/#naddr1qq…qxju

2024-08-12 01:32:26

- reply

I can see your list on listr.lol

2024-08-12 01:29:05

- reply

Ok progress!
Try navigating to some profile pages and see what the influence score is. Should all be in the range between 0 and 1, mostly close to 1.
Or go to the table of all users (icon top right next to settings icon) and look at the scores there.

2024-08-12 01:17:16

- reply

I don’t think that should be it.
Change is live, tell me what you get now.

2024-08-12 01:15:46

- reply

One pubkey in the tags field that is

2024-08-12 01:15:21

- reply

The fact that you have one pubkey means that the function in question is being called. Hmm.
I have an idea what it might be. Just made a minor edit to the function and pushed it. I’ll tell you when it’s live.

2024-08-12 01:09:06

- reply

Yup, might be worth trying. I have only tested it with Alby, not nos2x. Perhaps it’s an event signing issue? Do you see any console messages?

2024-08-12 01:05:54

- reply

From what you see on page 7, the last update def pushed. It’s as if a function didn’t get called that should have on page load. But the publish button should call it again. If you click publish, does anything happen? You should see a popup and also see a message in the console, both of which should show the event id.

2024-08-12 00:59:01

- reply

What does your grapevine calculations page look like? https://i.nostr.build/V550enxIo46ZYFNH.jpg

2024-08-12 00:48:45

- reply

But I see that you did turn the toggles all off afterwards, which is probably needed to keep from slowing down the score calculations.

2024-08-12 00:47:16

- reply

Go ahead and give step 3 a try - it’s the simplest way to get some profiles that are more than 2 hops away.

2024-08-12 00:27:54

- reply

What does your settings page look like? https://i.nostr.build/M8i092Yw6b01YYDR.jpg

2024-08-11 21:46:56

- reply

brainstorm.ninja is updated to export scores in NIP-51 format, top 1000 scores only, with title: Grapevine WoT Scores List. You can see my list at listr.lol

2024-08-11 18:43:15

#meatstr part deux https://i.nostr.build/pXWkaIIoIUYryws7.jpg #note1hev…2csd

2024-08-11 18:13:46

#meatstr https://i.nostr.build/ZfDxEmrl7EjhIPNE.jpg

2024-08-11 18:10:11

- reply

No real hesitation — I could very easily do both formats. Or make it optional. I’ll go ahead and change it to NIP-51, top 1000, push changes and let you know when it’s done (hopefully within a few hours)

2024-08-11 17:09:08

#100aDayUntil100k
#100pushups
35 + 35 + 30 incline with weights
Day 191 ✔️

2024-08-11 17:02:56

- reply

I made some changes locally on test page 7 to generate these reports, including setting it to publish only the top 1000 scores. I can push those today if you’d like to try it out.

2024-08-11 03:48:34

- reply

I published influence scores using the below two naddrs if you’d like to take a look and see if they would work for your purposes. The first one I just published, and the second one I published about a week ago. These are kind 39902; I still may change them to be a modified NIP-51 format. But it should be straightforward for you to extract the scores and use them as a filter.
njump.me/#naddr1qq…3qr4
njump.me/#naddr1qq…9yzg