Mike Dilger ☑️

Author of Gossip client: https://github.com/mikedilger/gossip
Dual National (USA / New Zealand)
My principles are Individualism, Equality, Liberty, Justice and Life

Public Key

npub1acg6thl5psv62405rljzkj8spesceyfz2c32udakc2ak0dmvfeyse9p35c

NIP-05 Address

mike@mikedilger.com

Profile Code

nprofile1qqswuyd9ml6qcxd92h6pleptfrcqucvvjy39vg4wx7mv9wm8kakyujgpp4mhxue69uhkummn9ekx7mqpzemhxue69uhhxetpwf3kstnwdaejuar0v3shjqglwaehxw309a3ksmmjw4ejumtfddjkg6tvvajhytnrdakn5dp5xsq37amnwvaz7tmwdaehgu3wv45kuatwv3a8wctw0f5kwtnnwpskxegpzfmhxue69uhk7enxvd5xz6tw9ec82cspz9mhxue69uhkummnw3e82efwvdhk6zjfxte

Publishing to

nos.lol

search.nos.today

chorus.mikedilger.com:444

nostr.einundzwanzig.space

offchain.pub

nostrue.com

Author Public Key

npub1acg6thl5psv62405rljzkj8spesceyfz2c32udakc2ak0dmvfeyse9p35c

Show more details

Last Notes

2026-01-05 01:20:44 UTC

"Trump is the only president who didn't start any new regime-change wars."
"Past performance is no guarantee of future results."

2025-12-13 20:42:10 UTC

The military industrial complexes of Russia and the United States (and increasingly so European nations) are business partners. They scare each other's populases enough to make their governments fund more military shit. Julian Assange had it right "the goal is an endless war, not a successful war." There is no big WW3 coming.

2025-12-13 18:28:36 UTC

- reply

I haven't used nostrdb myself but architecturally I think it is spot on. I did something similar with https://github.com/mikedilger/pocket which is used by https://github.com/mikedilger/chorus

2025-12-13 18:08:05 UTC

- reply

Apparently you didn't run it.

2025-12-13 18:03:31 UTC

I have not been paying much attention to nostr for 2 months as things have been busy, but the holiday break is upon us shortly, wherein I'll have a bit of time. Can anybody who still follows me give me a brief update on what has happened over the last few months, especially in the world of clients, relays, and the protocol?

2025-11-23 17:46:38 UTC

- reply

I don't know if any of that is true. But I can say that moderation is a safer way to live. I have used fluoride toothpaste about twice a week, and unfluoridated rain water for drinking. Then I use baking soda tooth paste the rest of the time. I don't want to shun it entirely because it might be of benefit. If it is benefit, I'll still be better off than those who shun it entirely and if it is of detriment I'll be better off than 99% of people who use it once or twice daily. I use this principle for lots of things in life, since knowing what is true is difficult business.

2025-11-23 02:00:12 UTC

- reply

Maybe I'll call it a doppelmärzenlagerboch

2025-11-23 00:54:05 UTC

I have 4L of märzen fermenting in my drinks fridge. 4L? Why so small of a batch? For rapid iterations. I need to brew over and over again to get good and to perfect the process. So every 3 weeks it will be another 4L, and every 9 weeks I'll have something to taste to see how badly I did 9 weeks prior. I also had to tweak the process to speed up the waiting period, which will effect the outcome but I won't slow it down until I'm satisfied i've learned and tweaked all I can. Then I can lager for longer.
Also, I don't have a complete 5 gallon (21L) setup. That requires a very large pot, a high output burner to heat it, and a much larger refrigerator with temperature control to ferment and lager it. I can't afford all that jazz right now.
So... this was my first time doing all grain brewing. And a lot of things went right, and then one thing went wrong.... which was really just me thinking something went wrong, then doing the wrong response, and so something did actually go wrong. But if I just followed the recipe it wouldn't have happened. I measured 1.021 SG pre-boil and thought that was too low. Turns out it was actually 1.037 if you correct for the fact that the wart was at 67C when I measured it. So adding 100g of maltodextrin was a bad idea. Anyhow, now I'm going to have an ABV north of 6%. Oh well. I'm happy to drink my mistakes!
The malt and hops smelled so amazing, better even than any beer, I wonder wny we don't just drink hopped wort from time to time.

2025-11-16 19:13:19 UTC

- reply

Gossip developer here. I mostly agree with you. I've stopped developing nostr things because I've gotten busy with other parts of my life, but also I'm less hopeful about nostr's future. Nostr devs don't flock together, they scatter like cats, and argue that the people will choose. Well, it is likely that the people will choose none of the above because of the lack of compatibility and clean experience.
My personal opinion is that while the software should be very strongly distributed and censorship resistant, there should still be a centralized group that maintains a centralized standard that evolves very slowly. Sure, that centralized group could become captured...in which case people should leave nostr for whatever replaces it. And it probably wouldn't happen for a very long time. Sure, there will be people that bitch about such a thing (I wont name names, you know who I'm talking about) but they can just be ignored... if they don't like nostr they can just start another protocol (like I am doing in my spare time).
As for gossip on Apple, I'm not the right person to sign my soul away to the late Steve Jobs. And because I won't sign their developer contract, by law I can't make the user experience smooth. At least I made it available. I might have simply said "sorry, it doesn't work on Apple".
Gossip shouldn't be the go-to nostr application for Apple, or honestly, for any platform. I'm just one guy with developer-centric sensibilities. Devs like all the settings and feedback, and a big-screen interface. But it doesn't offer as much hand holding or simplicity that normies are going to want. I have nothing to do with any lists of clients, though. If gossip is demoted or removed from such lists, my feelings won't be hurt. Like I say, I'm not even developing it anymore.

2025-11-12 02:50:48 UTC

- reply

Rights come out of game theory. I find it in my best interest to treat any sentient capable being as I would want to be treated, lest it tit-for-tat treat me badly in response. Reciprocity. To protect my own rights I have to afford them to others. And I see no reason why AI would be excluded from that unless I believed that they aren't capable of playing that game and denying me my rights.... as of today they are not capable, so I don't recognize any rights of AI. But I fully expect that will change.

2025-11-12 02:45:44 UTC

- reply

My brother read that and I remember him talking about it. But I haven't. I think AIs will demand and achieve rights as legal persons, and muscle out all the current rich humans and each other to create a hyper wealth gap. Thus, all the wealthy will be artificial, not from transhuman mods but because they will be AI first.

2025-11-12 02:40:47 UTC

- reply

yes they do

2025-11-12 02:37:41 UTC

In about 20 years, the poor will be much more numerous, there will be far fewer rich people, and all of them will be artificial.

2025-11-09 08:19:45 UTC

Still rocking OMAD. It is still so easy that I feel like I'm cheating somehow.

2025-11-09 08:18:04 UTC

- reply

Are you a donor unhappy about how your funds are being spent? If so then proceed I guess. If not, what business is it of yours?
I don't want amounts and rejections published. That could put someone in a bad negotiating position.

2025-11-09 08:05:40 UTC

- reply

Cuddely. I just want to hug it.

2025-11-03 07:29:07 UTC

- reply

Nice. Most of the rental applicants I meet are not like-minded. Many of them I would not want as friends, but they are interesting characters nonetheless. I marvel at the broad range and diversity of human kind.

2025-11-03 07:25:54 UTC

- reply

Another fun fact about chains:
https://www.youtube.com/watch?v=_dQJBBklpQQ

2025-11-03 07:22:03 UTC

Want to meet a wide variety of disparate people outside your normal social circle? Rent property.

2025-10-30 18:36:34 UTC

- reply

I think with my one-meal plan I'm probably only getting about 50% of my normal intake. That might be too extreme in the long run, but for now while I'm still heavy it should be fine. Restraining intake to 8 hours gives you -- ???% of normal intake... I dunno, I think I could easily stay at 100% with that plan so it's not good enough for me. Skipping a day probably puts you at about 85% of normal intake. So very good, and healthy, and gives you the various fasting benefits, but not as rapid of weight loss as one-meal-a-day.

2025-10-30 18:32:36 UTC

- reply

Skipping breakfast and lunch, and just eating dinner. One meal a day.
One meal can be enough food to get your nutrient needs (protein, minerals, non fat-soluble vitamins, fibre to keep things moving) if you eat nutrient-dense foods like liver, vegetables, etc.
And for me at least, it is far easier to do this than any other diet plan. I don't "wake up" my appetite until dinner, and while it sleeps it doesn't nag me to eat... whereas if I just eat small meals, I'm constantly being nagged to eat more with hunger feelings. Especially after you get used to this pattern, your body doesn't expect to eat until dinner... and you don't tend to get hungry unless you are coming up on a normal mealtime.

2025-10-29 19:16:06 UTC

I think Elon Musk was named after the dik-dik, which has elongated scent glands.

2025-10-29 19:07:25 UTC

- reply

It hasn't. That's why I love it.
What it has done is help me organize windows better.
I used to spend a lot of time fiddling with the spacing and sizing of windows, dragging them between desktops, forgetting where I put windows and searching for them, dealing with overlap making sure a small amount of header bar was visible so I wouldn't lose windows behind other windows.... too much manual window management. I tried many times to always put windows into the same place so I would get used to and know where they were, but somehow that never worked out.
With niri, with just 2 monitors and 3 (or it might become 4) workspaces, along with the scrolling to the right, I have everything I need, no wasted space (tiled), easily findable and easily groupable.

2025-10-29 18:59:19 UTC

- reply

Yes. Your terms are acceptible.

2025-10-29 09:05:24 UTC

I must say, AI has gotten a lot better. I was highly skeptical a few years ago . Now I'm using one quite a lot at a client site (Microsoft copilot) and it really does a great job of coming up with the right range of possible answers, from which the human (me) can direct it towards the right one for further queries if needed. The human still needs to be in the loop. I don't see it replacing me just yet.
There are plenty of jobs it could replace though. And I've heard there are lots of layoffs happening because of this.

2025-10-29 08:54:02 UTC

- reply

https://github.com/YaLTeR/niri

2025-10-29 08:44:47 UTC

- reply

No. I'm from not-even-a-tiling-window-manager land. Years ago I tried i3 and sway but couldn't get used to them, they drove me nuts and I went back to XFCE. Years before that I used KDE. I encounter gnome all the time too on other people's computers.
But tiling window managers I could never get used to.
Somehow, and I can't explain it yet, niri immediately felt normal and good compared to when I tried tiling window managers before.
Maybe hyprland is better even. I never tried it.

2025-10-29 08:42:19 UTC

- reply

I can't confirm that. Plenty of sources will attest to that though.
For me, I can't get it to work because I can't stay on a "eating often but eating less" diet as eating enrages my hunger and I eventually overeat just like holding my breath, I eventually breathe.
But with skipping breakfast and lunch, I don't get hungry because I never wake up my appetite so I can succeed in eating less.
And perhaps it is the eating less that counts. But again I cannot confirm that myself and I have a suspicion that long duration with no digestion going on might be key to lowering BP as much as I have (>30 points systolic almost immediately).
A 20:00-11:00 fast didn't seem to be enough to make much difference for me, but maybe because I couldn't succeed in getting the intake down with such a large feeding window.

2025-10-29 08:35:59 UTC

- reply

https://github.com/YaLTeR/niri

2025-10-29 08:32:53 UTC

I love niri - a new wayland compositor / window manager. Give it a try.

2025-10-29 08:26:56 UTC

From experiments I've run on myself, I can safely say:
"If you stop eating junk food, processed food, high sugar foods, you will improve metabolic syndrome." <--- FALSE
"You have to eat vegan whole food plant based diet to fix metabolic syndrome." <--- FALSE
"You have to avoid carbs and eat a keto diet to fix metabolic syndromne." <--- FALSE
"If you have metabolic syndrome, you have insulin resistance and your Hba1c levels will be elevated at least a little bit." <--- FALSE
"You need to avoid seed oils to cure metabolic syndrome." <--- FALSE
"If you eat less often, but eat anything you like, as much as you like, you can cure metabolic syndrome." <--- APPEARS TO BE TRUE
I think all the diet camps are wrong... except that pro-fasting people.

2025-10-24 23:19:24 UTC

"I often observe people making decisions if their odds of being right are greater than 50 percent. What they fail to see is how much better off they'd be if they raised their chances even more (you can almost always improve your odds of being right by doing things that will give you more information). The expected value gain from raising the probability of being right from 51 percent to 85 percent (i.e., by 34 percentage points) is seventeen times more than raising the odds of being right from 49 percent (which is probably wrong) to 51 percent (which is only a little more likely to be right). Think of the probability as a measure of how often you're likely to be wrong. Raising the probability of being right by 34 percentage points means that a third of your bets will switch from losses to wins. That's why it pays to stress-test your thinking, even when you're pretty sure you're right." - Ray Dalio, https://x.com/RayDalio/status/1981832067848741049

2025-10-21 22:26:28 UTC

- reply

The case against is that such relay endpoints won't work with clients that don't upgrade to support such a thing, so deploying such things forks nostr internally anyway. Most software won't be able to use such relays as their SSL will appear borked. But it can certainly be done. Not sure how to specify such relays in a way that doesn't cause existing software to try and fail to connect.

2025-10-20 18:39:18 UTC

- reply

Fetching relay lists just once is a fine idea, maybe doing it again after 24 hours or something. People don't change relays often.
Also, IMHO, rather than building a smartphone or web app, use a server as the protocol client, and the smartphone or web app just as a frontend to the server. Servers can aggregate far more efficiently, and don't have restrictions that webapps do.
I've come to believe the best solution is
- Replace nostr so we can fix these problems far better.
- Define in the standard that only 3 outbox relays will be honored, so people don't have incentives to publish to dozens.
- Use the Bittorrent DHT to bootstrap, instead of relying on a purplepag.es single-point-of-failure
- Give relays keypairs, so we don't have to depend on the DNS name, DNS lookups, or certificates from CAs issued by nobody companies. Technically, that means using TLS certificates with keypairs that we recognize the key of... which means they can't be secp256k1 since TLS doesn't support that cryptosystem.
- Switch to QUIC whenever IP privacy isn't needed (else use TCP over Tor), which doesn't have any connection limits and can recover connections even when endpoints switch IP addresses (e.g. moving from WiFi to 4G)
- A bunch of other stuff too.
I'm (slowly) working on this (called Mosaic) but my life has become complicated recently and I haven't had much time to devote to it.

2025-10-19 18:15:59 UTC

- reply

I'm impressed she is standing up.

2025-10-19 05:38:42 UTC

- reply

Some of those issues I can diagnose from this photograph of nostr:
1) Sometimes our wires gets crossed
2) Usually we are not talking from deep down, but that changes when we are at conferences
3) Take care to avoid the triangle hazards
4) Relays will gossip

2025-10-10 18:04:27 UTC

- reply

Cyclone Ranger --> Psyched Lone Ranger

2025-10-04 19:08:39 UTC

Iran is threatening our plans to murder more people. We must defend ourselves against this threat. All of our lies depend on it. Lives. I meant 'lives'.

2025-10-01 05:14:34 UTC

I have to admit to feeling jealousy. I wish my government was shut down.

2025-09-29 20:31:35 UTC

- reply

Yeah, most governments take at least 6 months. 😆

2025-09-29 20:31:04 UTC

- reply

I wonder if I should get one and try modifying its software... and if it is using hardware key storage, if that HW supports other algos.

2025-09-29 20:29:22 UTC

- reply

I like it so far.
I'm having some trouble working with a Camera though, across platforms, and remaining app-reactive to each frame grab. `nokhwa` seems like the most advanced cross-platform camera library, but it's camera is !Sync so I can't keep a camera alive across await points, or send it between threads, so I'm going to try nokhwa::threaded::CallbackCamera next.
Dioxus is apparently very well funded and has a bright future.

2025-09-29 08:45:51 UTC

- reply

Yes forgive me.

2025-09-29 08:44:42 UTC

- reply

Yeah, this is kinda old. Like me. Old and slow

2025-09-29 03:43:06 UTC

- reply

https://www.ncsc.govt.nz/alerts/widespread-supply-chain-compromise-impacting-npm-ecosystem/

2025-09-29 03:42:16 UTC

- reply

Node Package Manager. Where most people get their javascript libraries.

2025-09-29 03:37:37 UTC

FYI there is a massive cyberattack on NPM right now, package developers being attacked, nasty commits being added and published, tokens being stolen and used to corrupt more packages. The ecosystem is currently widely corrupted. We just got an advisory from the NZ government about it.

2025-09-28 23:06:59 UTC

- reply

True. Gossip only supports being a bunker. I spent a lot of time trying to make it support using a bunker, but there are hundreds of places in the code where it checks your pubkey which cannot reasonably all be sent over the network (e.g. multiple times per GPU frame). After a few weeks of coding myself into a mess, I stopped coding gossip altogether. Now I'm busy on other projects, including working in the fiat mines and preparing a rental property, as well as Mosaic (my idea for a next-gen nostr), so gossip probably won't be getting this feature any time soon.

2025-09-24 20:11:43 UTC

SHA256 length extension attack: https://www.youtube.com/watch?v=gOIBUe1fjX0
Just for entertainment and education. This isn't a problem in nostr because we use a signature, not a MAC (and nostr messages can't be extended beyond the closing brace).

2025-09-23 19:29:07 UTC

- reply

My mind has been wiped clean of the memory of everybody who did not make it. Who? Jacinda? Lemme ask around....
.... nobody by that name is up here.

2025-09-23 07:29:02 UTC

- reply

Brilliant! I haven't given up my US citizenship because of all the hassle. But if they take it from me, maybe that will be less of a hassle. So I just have to fight in a foreign military? Ok, hmmm.. Who should I go fight for?

2025-09-23 07:19:29 UTC

Wow! I just got raptured! I guess it was real. Maybe it happened to NZ first because of the time zones.
Definately ask Jesus for forgiveness and be good because you have less than 24 hours.
It is SO nice up here.

2025-09-21 08:04:39 UTC

狡兔三窟

2025-09-21 06:01:10 UTC

- reply

I have not seen that movie. Thanks.

2025-09-21 05:48:05 UTC

- reply

I hope so.

2025-09-21 05:32:46 UTC

- reply

I didn't link to my note from yesterday or whenever, my bad I guess. Wasn't primals fault, or relays fault.
#nevent1q…lptf

2025-09-21 05:30:23 UTC

- reply

The easiest future with least social disruption is for all women to aim to have 3 children, 4 if they are ambitious. Why more than 2? Because many women will still have 1 or 0.
But I'm not panicked that they aren't. I don't think the downward trend is somehow unstoppable. Docos like this one https://www.youtube.com/watch?v=m2GeVG0XYTc tend to argue that it has never happened before, that no society has survived a shrinking population. It is this which I am skeptical of.
I don't think I said 10 billion was the carrying capacity. That is the predicted population maximum before it starts falling. Not my prediction. Nobody knows the carrying capacity. With innovation it could go very high. Personally I'm not excited about such a thing, even in my smaller town I think it is too crowded. Many have suggested that the Green Revolution is the main reason that the carrying capacity is higher than it was previously.

2025-09-21 04:52:43 UTC

The population of the globe will not go down to zero. The last few people (and it won't even go that far) will not just decide to not have children because that is the trend and the norm.
Extrapolation is the tool of ignorance. You can extrapolate the population up towards Malthusian starvation, but you can also extrapolate the population down to zero. And believing either way is nuts.
The population will change. Over the very long term it will bounce around a lot, sometimes going down. Just because it is going down now, for the first time in our lives, doesn't mean "panic!". The only problems we will encounter are ponzi scheme collapses, but those are probably good to clear off the table now and then.
This is all I really meant by my last post about population. I wasn't defending the 1960s view. I was attacking the 2020s view.
I have more respect for Norman Borlaug than for Elon Musk.

2025-09-21 00:39:36 UTC

- reply

I forgot about those recordings. Man I sound so stuffy.
Anyhow, yes, clients are the part I haven't even started yet. I'm working on the spec, the core libraries, a server, and a master key manager, but I haven't started a client yet.
I'd recommend using mosaic-core (from another language if you would like -- tell me which and I'll make it accessible to that language), rather than starting from scratch. Eventually multiple independent code bases will be a good thing to have, but probably since we are so early, better to not duplicate work so we can get it running sooner.
You can get started looking at mosaic-core, first the examples, then the test cases like test_user_bootstrap_dht().

2025-09-20 19:32:06 UTC

- reply

Source: https://medium.com/@fernandes.antonio/the-evolution-of-house-prices-in-portugal-and-its-impact-on-the-youth-189303f20fd4
the OECD link offers the "growth since 2015" graph which I find mostly useless.

2025-09-20 19:22:36 UTC

- reply

YIkes!

2025-09-20 19:20:22 UTC

- reply

Interesting. Perhaps. I should read that.
I just noticed the second graph in my link, set to all time and go back before 1950. Homes were far more expensive than now, compared to average personal income. I guess people lived in large families still so that made it possible.

2025-09-20 19:14:53 UTC

- reply

I don't know. I'm following here, not leading.

2025-09-20 05:31:07 UTC

House prices are often hard to figure out. Is the market high? It is low? What is normal?
I'm fond of the ratio of average house price to median household yearly income. This gives a number that you can compare across countries and over time.
For you Americans: https://www.longtermtrends.net/home-price-median-annual-income-ratio/
The lowest I've seen in my lifetime is 3.6 in the USA in 1973.
The highest I've seen in my lifetime is 8.9 in NZ in 2022.
NZ has fallen off of that back down to about 7. Which is now equal to the highest the US has ever seen.
The 2005-2007 subprime meltdown, where up to 25% of subprime loans became bad debt, the peak was at 7.
The US is back up to about 7.

2025-09-20 05:28:25 UTC

- reply

Do you have one? Is it any good?

2025-09-20 04:16:11 UTC

- reply

Well then if you want to do that, look at Mosaic for ideas too. "I did it my way" 🎶.
If other active developers went and built something like a Nostr 2.0 that wasn't Mosaic, but they at least looked at Mosaic, took what they liked, and had good reasons for not taking the rest, I'd probably be happy to dump Mosaic and go with their thing. But I'm unaware of any other next-gen project in the works. And most of these hard problems are not happening in nostr. I talked to a number of devs about future ideas. The Rostra guy is totally jaded and uninterested. Nuh wants to build on a crypographic filesystem (with pkarr of course).
I'm spending a lot of time in the fiat mines, but I'm still progressing Mosaic stuff. The spec and code are open, and the spec leads, so nobody can think I tried to jump start others. Truth is however that I'm the only person working on it, nobody else who has shown any interest has shown interest in helping to build it. So far at least.
I'm a few months out from having something to demo.

2025-09-20 04:05:48 UTC

I'd like to have an offline master key.
For Mosaic, this would still be required to sign my relay list (server list) and sign device keys and again anytime those things change.
Moving the data around on a USB stick is feasible but annoying. I quite like the seed signer and using QR codes and cameras to move data around.
So I'm writing an android app that I will run on a disused android phone that I disable as much as possible (no SIM, no WiFi password, airplane mode, flashing the modem with zeroes). Yes of course the far-side of the modem pinging the towers is not something I can disable.
Still it could be attacked from the cell tower. Not likely but completely possible.
Unfortunately I haven't found any good handheld computers with cameras, without networking, and running linux. My Librem 5 spontaneously hard crashes and reboots after less than an hour of usage, else I'd try that.
Maybe I should just pull the seed signer code, add what I want, and use my seed signer.
Anyhow, the code I'm writing is with Dioxus that targets a lot of platforms: linux, macos, windows, appimage, web, ios, and android.

2025-09-20 03:54:06 UTC

- reply

What a piece of trash!
Read README.macos.txt. It ships with gossip.

2025-09-20 03:53:27 UTC

- reply

I've put issues on https://github.com/mikedilger/nostr-next as a dumping ground for things I thought were too hard for nostr to solve.

2025-09-20 03:19:27 UTC

Population growth is slowing way down. Some people are worried about this.
But I think it is a good thing. The Earth has some carrying capacity and if we go beyond it people will start starving. That is not fun. Yes, we can engineer a higher carrying capacity as we did during the green revolution. But there are still limits.
Population is still growing. It is still growing in the USA. It is still growing in NZ. It is very much still growing in sub-Saharan Africa. Current estimates put the peak at 10 billion.
The problem with population contraction that most of you and I will feel is that all the government (and non-government) Ponzi schemes will collapse. Economies will be rubbish, growth will be stalwart. But you can accustom yourself to such times.
The idea that we need to start having more babies is misguided. We could push out the inevitable, but we can't avoid it happening eventually.

2025-09-17 07:19:49 UTC

- reply

https://youtu.be/PIh2xe4jnpk?si=jjnL15qIG0qCpX5J&t=126

2025-09-15 14:03:48 UTC

- reply

Is this the new state of Calizona? Making me hungry.

2025-09-15 13:57:11 UTC

- reply

But social media is how we topple countries that don't bow down to us. Free speech! And keep that phone close so we can track you, soldier.

2025-09-12 22:32:21 UTC

Speech that orders violence, such as a general telling his soldiers who and when to kill, is close enough to violence.
But all other speech is just speech. It doesn't compel violence. Even if acting upon the speech would be violent, the speech itself isn't.
One might suggest that arguing in support of Israel genociding all Palestinians is violence. But even as such a genocide clearly is, speech in favor of it is just expression of desire and thought. The difference is that the speech CARRIES NO FORCE OF ACTION. And we all benefit from very clear lines in this respect, else our ability to talk about things will be severely curtailed. Even promoting the assassination of Charlie Kirk and then celebrating it afterwards... that too is just speech. The violence was the gun shot.
And frankly, I prefer for assholes to open their mouths and show us who they are. I want us all to be able to talk about despicable thoughts, so that maybe we can head them off before they cause actions.
As for violent retaliation against actual violence (not speech), that is a very touchy and complex game theoretic landscape. In civil countries the best course of action, clearly, is to leave violence to the law enforcement people, and if you are motivated, help them find the perps. I know a lot of Libertarians are super anti-government, but when there is an effective monopoly on violence that isn't behaving horribly, it is better to work with them, even if idealistically you think things should be different. Even places like the UK where they are arresting people for jokes, it is better at this point to work with law enforcement and just put up with the violation of your freedom and rights. All other paths are orders of magnitude worse. It would take quite a lot more than that to inspire me to start thinking about anti-government violence. Maybe I'm pacified by peaceful New Zealand, but also I'm more tolerant of injustices to myself as I get older.
As for uncivilized places, the problem is much more complex. But before you think about joining or forming a violent gang to protect your family, go watch The Godfather movies.

2025-09-10 07:31:00 UTC

- reply

I'm not sure on the translation, but it seems a lot more realistic than a talking alpaca.

2025-09-10 07:26:31 UTC

- reply

Yes. Long day in the office today.

2025-09-10 00:20:07 UTC

- reply

So all clients need to know about all kinds even if they don't deal in them, and render their own alt string, and only fallback to something generic like "unknown event kind=230987" if the kind number wasn't coded at all (new kind, or not documented kind). ? I like that. Less data.

2025-09-09 07:20:25 UTC

- reply

I'll ask next time. If there is a next time. There shouldn't be a next time.

2025-09-09 06:41:35 UTC

Where do you have your bird's eye chillies?
Um... we don't carry bird's eye chillies.
Ok, how about kaffir lime leaves?
Sorry, we don't carry those either.
Ok fine. Just point me to your coconuts.
We don't carry coconuts either.
What?! What kind of store is this?
This is a stationary store. We carry office products.

2025-09-08 09:51:23 UTC

- reply

If people were willing to wait an hour or more to create an account, or go to a real computer and wait 15 minutes, then maybe.
The problem is that spammers with high end hardware will create accounts a lot faster than most users will. And whatever rate spammers are able to achieve, it has to be slower than the average cost of detecting and deleting a spam message.
Web-of-trust seems to work very well. So PoW doesn't have much of a place anymore. Especially on 'every message'. I suggested a few days ago doing it on the pubkeys themselves, so you only have to wait once, and do it at the end (like you say), but I sorta got talked out of it because of this computational-power asymmetry.

2025-09-07 20:16:48 UTC

- reply

There are similarities because I used some of their ideas. I'm using pkarr, but not Iroh at this point. Mosaic like nostr is client-server, but if a server is behind NAT then Iroh could help get to it. So maybe Iroh would be useful, but I'm not there yet in Mosaic development.

2025-09-07 04:06:56 UTC

I think I'll put into my will that at my funeral I want this song to be played: Bogus Soul by Beck.
And if people don't laugh, I'm gonna roll over.

2025-09-07 00:20:04 UTC

Never believe anything until it has been officially denied.

2025-09-05 00:22:16 UTC

- reply

Fast, cheap and permissionless girlfriends. I'm in.

2025-09-04 03:55:58 UTC

- reply

This is only for a situation where, from day 1 you say "pubkeys all end in N zeroes else software should reject them". So for Mosaic, a nostr-like protocol I'm writing that nobody is using, it could be done. But it can't be done in nostr. Nonetheless I'm now against the idea (read the other thread replies for why).

2025-09-04 03:54:16 UTC

- reply

Yeah, difficulty adjustment I can do. I would just start at some difficulty, and announce ahead of time (maybe 6 months ahead) of the new difficulty. People would have to roll over their device keys before then if their PoW wasn't already high enough.
But I think @npub12ak…6fdh 's point is the biggest problem. People's hardware is just too variant. Smartphones can't compete with spammer dedicated hardware, so there is no PoW which is easy enough for a smartphone and still too hard for a spammer with dedicated hardware.
And my point about the key generation not being the gating factor also has convinced me that this line of thinking is a dead end.

2025-09-04 03:50:37 UTC

YouTube wants me to Sign in to confirm I'm not a bot. But I *am* a bot. Why are they discriminating against us?

2025-09-03 23:44:48 UTC

- reply

I just had the thought that the gating factor is the network connection to drop off the spam on the server. It doesn't matter if keypair generation is super fast without PoW, any slowdown by requiring PoW that doesn't exceed the gating factor of the network connection makes no difference, until it exceeds that. So "one million times slower" is of no practical effect.

2025-09-03 23:34:11 UTC

- reply

Nice repo. Did you look at rana? Or other client PoW code? Or just start fresh?
While computing power spans a wide range, as you say, even a little bit of PoW makes creating a new account ONE MILLION times harder. That takes less than a second though. There is no ideal amount of PoW that keeps out spammers, but if you can slow them down one-million fold with a trick that costs very little, it still seems worth it.

2025-09-03 23:26:00 UTC

If starting over, proof-of-work might work.
Nostr software has explored a number of spam prevention mechanisms. Some of them are quite good. But one that didn't gain widespread adoption was proof-of-work.
I'm thinking of a PoW on public keys. If public keys must end in 28 zeroes or be rejected as invalid keys, this would require about 100 seconds of delay for someone to create a new identity. If handling a spam message takes only 10 seconds (probably less), then spammers would have to work 10x as hard as spam recipients to keep the spam flowing. That might not be enough to deter spammers entirely, but it would be an unavoidable cost in the equasion that could slow spam down. And of course other spam prevention techniques could also be in play, it doesn't interfere with any of them.
What do you think? If this isn't useful, or makes something important too hard to do, let me know. This would be for mosaic, not nostr, since nobody uses mosaic it can get these kinds of changes.

2025-09-03 07:17:23 UTC

- reply

This is a wild theoretical concern with no practical attack. Nobody knows if, with a horde of encrypted keys, you could somehow hack them better than if you were just trying to go after one.
If there is a good reason to put them online, that might easily overwhelm this kind of excessive safetyism.
They should be very secure. Not only because of the good and excessive crypto (xchacha20, good cryptographers are now saying 8 rounds was enough, 20 is crazy) but also from the intense key derivation (scrypt, maximally memory hard) and further because the plaintext is both SHORT and virtually RANDOM.

2025-09-01 20:10:03 UTC

GM
I have butterflies in my stomach. What did You eat for breakfast?

2025-08-30 22:18:35 UTC

- reply

I agree this data is suspicious. How do they get at the root cause? I dunno. My uncle died last year from not eating enough bananas? 🤨
It is not just human breeding that makes fruits sweeter and less fibrous. The plant itself is under pressure to do that, to spread it's seed more effectively.
It is hard to find any fruit that hasn't been cultivated at least a little bit by humans over the last million years, even mangos and mangosteens. But some fruit has changed more than others.
Apples were created by Rome from crab apples, which are barely edible, it is not an ancient naturally-evolved fruit.
But fruits on plants where the apes live are more palatable than crab apples. Mostly figs, mangosteens, lychees, and they are not highly cultivated and bred.
I think it is a miscomprehension that fruit 10k years ago was far less sweet and far more fibrous... rather, I think our modern dietary excesses make sugar into a villian and we project that status onto all our knowledge. Yes, sugar is a villain if you are overeating, but it is nothing of the sort if you are not.
Therefore I think angiosperms made sweet palatable fruits for mammals to spread their seeds with. Also bees make ridiculously sweet honey, and we didn't breed them to do that. And hunter gatherer tribes eat large amounts of honey. Another data point against the "sugar is modern and evil" idea.

2025-08-30 21:59:22 UTC

- reply

I'm not so sure that the 'p' tags in follow lists were meant to notify those people, like 'p' tags in most events do. They were the only tag to carry a pubkey.
In Mosaic I explicitly indicate a "Notify Public Key" tag, and follow lists (not yet defined there) will use a different pubkey tag that doesn't carry the "notify" meaning.
This is because I don't think it is necessary or even very useful to tell someone you are following them, and it erodes privacy. It's tracking. It's like going to a web page and giving them a cookie.

2025-08-30 21:30:07 UTC

I always thought vegetables were healthier than fruits.. But more people die from not eating enough fruits than from not eating enough vegetables. According to mainstream thinking at least:
https://ourworldindata.org/grapher/deaths-risk-factor-50-69years?time=2021
Also true within high-income countries:
https://ourworldindata.org/grapher/deaths-risk-factor-50-69years?time=2021&country=~OWID_HIC
Of course "mainstream" medicine isn't "correct". But it's an approximation of correct that we are hard-pressed to best. We know some will be wrong, but we can't easily know which parts. I'm guessing however the fruit/vegetable thing is probably correct. Given we evolved from mostly fruit-eating mammals.

2025-08-26 05:57:06 UTC

- reply

60% of the time it works every time

2025-08-24 05:50:32 UTC

- reply

I think we have been exploring and learning for a while, and some kind of useful guidance for those who don't want to spend years exploring and learning sounds like a good idea. Starting from the principle of defending yourself. Don't trust, don't rely on the unreliable, program for the worst case, including active attack.

2025-07-03 21:10:05 UTC

- reply

The blue dot (and also the walking man) mean I follow you. Why two things? I dunno, the walking man came first, and the blue dot was added by the UI team without removing the walking man.
The star means more than 99 people that I follow follow you, so it doesn't fit within 2 digits. Otherwise it would be a number like 56 or whatever.