pippellia
Pip the social graph guy

pippellia.com

I demystify the social graph so you can learn faster and use this powerful tool more effectively.

Public Key

npub176p7sup477k5738qhxx0hk2n0cty2k5je5uvalzvkvwmw4tltmeqw7vgup

NIP-05 Address

pippellia@verified-nostr.com

NIP-57 Address

whitebat1@primal.net

Profile Code

nprofile1qqs0dqlgwq6l0t20gnstnr8mm9fhu9j9t2fv6wxwl3xtx8dh24l4auspzemhxue69uhhyetvv9ujuurjd9kkzmpwdejhgqgkwaehxw309aex2mrp0yhxummnw3ezucnpdejq4z3cha

Publishing to

relay.primal.net

relay.nostr.band

nostr21.com

Author Public Key

npub176p7sup477k5738qhxx0hk2n0cty2k5je5uvalzvkvwmw4tltmeqw7vgup

Show more details

Last Notes

2024-11-14 20:56:53

- reply

uff. I have mixed feelings

2024-11-11 18:34:52

- reply

and the blue of golang

2024-10-30 13:00:14

- reply

I think mints are similar to relays. Using few big relays is fastest, and if we all do it there is very low probability of missing some posts.
Same with mints, if we use few big ones it's fastest and we have lower fees (as most transactions would be intra-mint).
However, if we look at the current distribution of users per relay it's pretty good (plot from nostr.band)
Why is that?
The advantage of using multiple relays is lower probability of censorship.
Similarly, the advantage of using multiple mints is lower probability of rugs. Also, with mints there is an anti-centralisation influence, because you don't want to use the biggest mints as there is more chance of big rugs from the gov or hacks.
So I would say there is decent evidence that the distribution of mints will be more flat than that of relays.
https://image.nostr.build/9e05dbd414118bb74e9b0d1e43bd8cffede8db74bf0649350081f64bc5de725b.jpg

2024-10-30 09:17:57

- reply

There is a nip for which mints you recommend, but not one for which mint you use (which has many privacy implications)
There is a danger of feedback loop, yes, however keep in mind that your trending mints are going to be different, and possibly very different from mines, because it's personalised!
I agree community helps with these things, totally.
Another aspect to consider is that in the future reputation is going to become much more important, so who you give your attention to or who you recommend is going to be used to judge you. Trending towards anarchy requires personal responsibility and stricter societal norms, so I agree with you pointing out these problems, but I think they will become smaller in the future (still a lot of work to solve them in a convincing way)

2024-10-30 08:32:51

- reply

fair. It should be automagic like auto-selecting the first 20 mint based on their reputation to you, but yeah, reputation here should not be follows.
follows filter spam, which is the first step, but then you need something else. Like reviews! 🤣

2024-10-29 23:08:23

- reply

cool! Is the auth private, or do I need to sign something with a whitelisted npub?

2024-10-28 17:23:43

- reply

I still don't get it.
Where are these files stored? Whose computer

2024-10-27 09:53:03

- reply

People buy any lie because they have been operating on an artificially safe environment

2024-10-25 15:40:11

the UX of @nprofile…f30t is perhaps the best in nostr. Kudos

2024-10-24 19:06:56

- reply

Super interesting comments.
> Crossing mints means more complexity and trust issues maybe? But again, you can imagine niches where it might be ideal.
@nprofile…h5zr some time ago shared the "multi nut payment" or some other silly 🥜 name :)
The idea is that you can have your wealth spread across 1000s of mints, while still being able to pay a single lightning invoice. Just wanted to share that in case u didn't know.
Another very practical observation on the limit of the privacy of your scheme (which, don't get me wrong I find really interesting):
The person has to own an UTXO. Not many people today own an UTXO, so the anonset is already much smaller than the whole world. And for the future, well that'll probably remain small because of the onchain fees.

2024-10-24 17:48:28

- reply

For this particular usecase, I am not super bullish because:
1. micro payments (ecash, lightning) allow services to monetize directly. This was not possible before due to a technical limitation, so businesses got creative and explored a whole bunch of indirect monetization avenues.
2. even for public and free services, something simpler like running pagerank-ish algos (pagerank, personalized pagerank, trustrank...) on the follow graph/zap graph would do the job.
Btw I am actually building 2.
There is defo a place for "proof of rich" in identity. For example giving a free luxury service only to rich people. E.g. try this yacht for free!
Everyone would enjoy a free day on the see, so the business must filter.

2024-10-24 16:48:47

- reply

I can totally see this mechanism being integrated into a nice business model.
E.g. a company offering banks and lenders these proofs for certain users that want to use their services.
Or, some form of "proof of rich" in digital identities. Bots can't easily accumulate wealth

2024-10-24 08:29:38

- reply

totally agree

2024-10-23 21:45:50

- reply

I think @nprofile…mt90 and you Niel should do this together.
I can totally see your church ditching Facebook when they'll see something amazing like ZapChat!

2024-10-23 14:30:20

extraordinary work 🔥
#nevent1q…7p7m

2024-10-23 07:15:50

a 10x improvement in WoT analysis is coming 🔜
Plug & play, in any Nostr client or Relay

2024-10-23 07:06:26

- reply

wait until u try @npub10r8…t2p8

2024-10-22 19:16:09

super interesting!
#naddr1qq…4het

2024-10-19 07:46:44

- reply

they are just a bunch of envious goblins

2024-10-19 07:44:15

- reply

1) no one is beyond redemption, I was a shitcoiner too
2) who makes the joke is not important. The joke just has to be funny

2024-10-17 08:00:10

- reply

well, yes, but not everyone needs the multiple column feature.
I thought this was targeted to social media managers, which would benefit the most from such a tool and would probably pay for it.

2024-10-16 20:43:19

- reply

real fast. Who is the target user?

2024-10-16 19:56:46

- reply

ohh yeah

2024-10-16 14:44:30

- reply

fascinating

2024-10-15 17:19:40

- reply

No Comment/Quote boost section?

2024-10-15 17:18:58

- reply

love it. Weird and fun, as it should be!

2024-10-15 12:02:18

- reply

this is a superpower. Censorship resistant emojis! 😝

2024-10-15 11:51:00

- reply

how can I add it to my reactions on Amethyst?

2024-10-15 11:50:02

- reply

ahahahah lmao

2024-10-14 07:54:47

- reply

is this true @nprofile…qdrm?

2024-10-14 07:41:28

- reply

I absolutely agree. Even for the WoT angle, you can't build meaningful metrics if the user is not aware of what he's publishing.

2024-10-13 22:43:41

- reply

ohh, so it wasn't Amethyst. Okay, gotcha, and thank you.

2024-10-13 21:49:17

- reply

yes, but I've used it like 2 times. And to be honest both times I had problems so I wouldn't recommend.

2024-10-13 21:47:22

- reply

cool, website, but I still don't understand. I don't remember leaving a review, and in fact I haven't. No reviews there, no likes. So why am I recommending something that I don't recommend?
I have shared links from highliter, but that's it.

2024-10-13 21:38:04

hey @nprofile…vzaw, why do I know "recommend" Highlighter?
https://image.nostr.build/c5ac3f45a3fb170f061f37f74a777352522bd37d0a6228bc84c49caf19a3814d.jpg

2024-10-12 19:01:57

- reply

*do u have

2024-10-12 19:00:05

- reply

i have an online wallet with more than pocket money?

2024-10-11 16:56:55

- reply

curious to know your take @nprofile…4tc2, since you are building a relay if I remember correctly

2024-10-11 10:55:13

- reply

wooow

2024-10-11 10:09:37

loool 🤣
#nevent1q…qu5v

2024-10-11 07:37:09

- reply

he has connections

2024-10-10 20:45:13

- reply

this can be radically improved by using what I am working on.
Using the distance as the metric for filtering (if distance > 2, filter) is suboptimal, as it limits discover ability.
What's better is a filter based on personalized Pagerank ( if PP < threshold, filter). PP looks both at the distance and at the "flux", meaning how many path exists thay connects you to another npub.
So, PP can be:
- more flexible, because you can choose the threshold in a continuoum
- improves the discovery (by for example allowing npubs at 3 hops if they are "high quality")

2024-10-10 19:57:33

- reply

shouldn't the nprofile contain the ln-address, and other metadata?

2024-10-10 18:12:16

- reply

Does this effect the number of trusted active users (either daily or weekly?)

2024-10-10 11:52:48

- reply

I am not sure the data is correct.
If we look at the retention rate of trusted pubkeys, that's too low imo to just be the effect of spam. @npub1xdt…ntxy
+95% leave the first day just because they see some Reply Guy? Seems way too much compared to historical data.
https://m.primal.net/LRDd.jpg

2024-10-10 07:45:37

- reply

Gotcha.

2024-10-09 20:17:06

- reply

what are u working on? Is this for the aggregator?

2024-10-09 09:22:43

- reply

oh right, I am convinced. Do u recommend some kind of preparation before doing it?
I'll go with 3 days for the first time.

2024-10-09 09:00:08

- reply

I never did more than 1 day of fasting. Do u think it makes sense to reduce calorie intake for a few days (still eating, but not much) and than fast for let's say 2 days? This seems more approachable in mind

2024-10-05 21:08:48

- reply

@npub1zuu…c2uc u should hike in Tenerife if u haven't already. Beautiful

2024-10-05 21:07:48

- reply

@npub1zuu…ucun should hike in Tenerife if I haven't already.

2024-10-05 21:06:04

staggering
https://m.primal.net/LMPX.mp4

2024-10-04 22:14:17

- reply

It doesn't work. Any personal judging/recommendations system is useless if you don't express any opinion (who u follow, what u zap...). In that case, global Pagerank is more useful.

2024-10-04 20:22:47

- reply

that WoT score is similar (but worse) than personalized Pagerank.
Advantages of PP:
- the weight of a follow is proportional to the PP and inversely proportional to the size of the follow list. (those that follow 5k people have less impact than those that follow 500, given the same PP)
- PP give u information about everyone u are connected to, not just 2 hops. (However, by how it works it favours people closer to u). This means it can be used in recommendations systems

2024-10-04 18:53:09

- reply

amazing 😍

2024-10-02 08:11:40

- reply

I visited Munich quite a few times the last year.
Amazing city, super green, really good quality of life, not that more expensive than north Italy with the exception of houses

2024-10-01 19:55:34

super cool! cc @nprofile…q3w9
#nevent1q…7day

2024-10-01 18:42:25

- reply

GG mate. It has been and will be useful to me, to learn how to do my stuff in Go.

2024-09-30 15:29:04

- reply

the most based kid in the forest (?)

2024-09-30 12:50:47

- reply

Maybe it's possible, but it is quite complex because the mint would have to track how many times a "chain of banknotes" is moved around, and I am not sure if that's compatible with "pay to npub" types of tx, which work differently.
Even if possible, that's too complex imo.

2024-09-30 11:21:34

- reply

yes, the mint fee is to compensate for the work it does.
If two people are paying each other across mints, then they'll pay for the mint fees + lightning fees
To understand why let's look at the steps involved in a normal transaction.
1) Alice has an ecash banknote on MintA
2) Alice sends it Bob
At this point, both Alice and Bob own the ecash banknote. This is problematic, because Bob wants to have unilateral control of that banknote.
To do that Bob
3) Request a new banknote from MintA. MintA invalidate the old banknote, and gives a new one to Bob, that only he owns.
So to prevent MintA being DDoS attacked, MintA charges a fee for every request.

2024-09-30 08:58:55

- reply

unless u use your own lightning node, yes.
Nah, I don't think we need to fix this. If communications needs to be cheap, just don't use payments

2024-09-30 08:03:51

- reply

all wallets now have higher fees, even custodial ones I think

2024-09-30 08:03:25

- reply

if 1 sat is too high (right now), then it's better so send the signal in another way I think

2024-09-30 07:25:18

- reply

1 sat fee solves many DDoS scenarios, and it's not a problem for the users I think

2024-09-29 11:01:30

- reply

maybe u can get it for free is u login with the university

2024-09-28 23:26:39

- reply

each is 41.5kg, filled with half cement and half polystyrene.
Full cement would have been ~83kg each, way too much.
I've also made two wheelbarrow tires, full cement is 19.5kg

2024-09-28 09:01:57

Listening to bitcoin review without falling asleep
https://m.primal.net/LEUc.jpg

2024-09-28 08:52:06

- reply

spettacolo

2024-09-26 21:54:47

- reply

my mother does

2024-09-26 11:54:31

- reply

noice!

2024-09-26 11:51:44

Great tech, but maybe just put down the damn phone and enjoy the ride
#note1w7h…hmx4

2024-09-26 10:34:12

all good questions, reposting for visibility
cc: @nprofile…qdrm @nprofile…9zgs @nprofile…ej4c
#naddr1qq…csrq

2024-09-26 07:21:47

- reply

fair points

2024-09-25 22:31:43

- reply

why using jack?

2024-09-25 22:28:35

- reply

strange post I agree, but I hope u get better sir

2024-09-25 21:30:55

- reply

nice meme

2024-09-25 21:27:29

- reply

this too 🤣

2024-09-25 21:01:47

- reply

oh yeah, me too.
I think the crux of it is, because it's a trusted relationshp (with the admin/owner/relay/whatever), they don't agree on how to implement it.
If it were trustless (which can't be) they would agree more I think

2024-09-25 20:53:36

- reply

to recap and clarify my thoughts @nprofile…a6k8
- if one follows (= meaning gives attention to) a spammer, that will impact the view I have on him/her.
- if one follows child porn shit, that too will impact my view.
- if one follows vitalik, which I think it's a shitcoiner, that means they give him attention, which is not too bad.
My personal takes, which I think are coherent with the "gives attention to", which is what a follow is.

2024-09-25 20:49:10

- reply

mmm, you are right.
Follows just means "I am interested", doesn't mean "I trust in domain X", neither "I endorse".
And that's fine, as long as you don't use follows to derive something more than "Web Of Not Spam".
Clearly there is a line. If you follow child pornography, I'll deduce you are disgusting, but that's on the extreme.
For anything that has to do with value, I would just look at zaps.
Context specific lists? Better than trust attestations, but I doubt users will want to:
a) produce them and keep them updated
b) publish them publicly
I think we should not assume people to do things to help others altruistically. Sharing your friends list to help others find good people isn't going to be super popular I think. Zapping is different, because u help someone that has already given value to u, let's say the creator, and it's seamless.
idk, many unorganized thoughts here @nprofile…u9av

2024-09-25 20:36:19

Nothing better than Jon, Pablo and Banana talk shit 🤣
https://fountain.fm/episode/LBVT42qHgTDCJXWcnJ93
#nevent1q…62q9

2024-09-25 13:54:50

- reply

well, in meatspace your reputation is partly impacted by who your friends are.

2024-09-25 13:44:41

- reply

ouch, take care :)

2024-09-24 13:18:59

finally!
#note1ca9…nqtu

2024-09-24 12:56:30

- reply

mmm, interesting.
wdyt @npub1t3g…upxq ahahahah

2024-09-24 09:12:28

- reply

welcome! glad we are more numerous now

2024-09-24 07:31:45

- reply

my buddy @nprofile…nfxp is pretty good with illustrations.
Just saying

2024-09-24 07:29:09

- reply

I was about to ask.
It should be similar to C, ops Latin

2024-09-23 14:37:00

- reply

I would say either with spam mass estimation (http://infolab.stanford.edu/~zoltan/publications/gyongyi2006link-tr.pdf) or anti-trust rank (http://i.stanford.edu/~kvijay/krishnan-raj-airweb06.pdf)
Both are pagerank-like algos but here having a high-value means bad.
However, these black-lists aren't the solution imo, because updating them costs more than creating a new npub for the attacker.
Conversely, white-lists are the way to go, because the cost is on the attacker (work hard to be included in such lists).

2024-09-23 07:15:11

- reply

World of Tanks @npub1t3g…upxq

2024-09-23 06:19:32

shout out to @nprofile…swg4 for their top notch content.
(even though the scores can be debated)
#nevent1q…muut

2024-09-23 06:10:30

- reply

isn't it this one?

2024-09-22 16:18:47

- reply

hey stel, thanks for the suggestion!
I never heard about it

2024-09-22 00:26:16