fiatjaf

https://stuff.fiatjaf.com/

Public Key

npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6

NIP-05 Address

Profile Code

nprofile1qqsrhuxx8l9ex335q7he0f09aej04zpazpl0ne2cgukyawd24mayt8gpr9mhxue69uhhq7tjv9kkjepwve5kzar2v9nzucm0d5q35amnwvaz7tmjv4kxz7fwwajhxar9wfhxyarr9e3k7mgpzamhxue69uhkjmnzdauzuun9d3shjuewd3skueqpzemhxue69uhhxetpwf3kstnwdaejuar0v3shjqg5waehxw309aex2mrp0yhxgctdw4eju6t0qy2hwumn8ghj7am0wshxummnw3ezuurpwf68jxsvj50

Publishing to

Author Public Key

npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6

Show more details

Last Notes

2026-01-19 16:21:21 UTC

- reply

(I only wrote this here from https://fevela.me because I wanted to see if my reply was going to show up in the "chat #nevent1q…6hed" window of https://grimoire.rocks, but the bug is real.)

2026-01-19 16:20:01 UTC

- reply

Bug report: clicking on the "chat" option in a note's menu opens the note in a chat, but its associated command is just "CHAT", no id. So when we try to open it in a new window it breaks.

2026-01-18 17:59:41 UTC

- reply

Yes, everything is in that ./data folder.
The mmmm.lock file should get removed automatically when the server stops, but it might hang there if it crashed or other mysterious circumstances. You can just delete it (as long as you're sure you're not going to run two instances of pyramid at the same time with the same database).

2026-01-17 16:10:42 UTC

- reply

What is "raw blossom"?
If you done the installation using the easy method then you probably don't have to do anything, SFTP should just work.
Otherwise I think you must ensure you're listening on 0.0.0.0, I'm not sure.

2026-01-17 14:51:47 UTC

https://www.youtube.com/watch?v=_Sls_QCMG2Y

2026-01-17 14:43:59 UTC

Thanks to @npub1937…haj6's neverending efforts `nak fs` now works on Windows:
https://image.nostr.build/cb625c00e08422dde83ecd72e285c20290dc566fbd342bf9bf81603d6b8c5efb.png

2026-01-17 14:33:05 UTC

- reply

Pretty good solution indeed.
By the way, this is something Radicle probably can't do without overcomplicating everything with encryption, but on Nostr it can be easily done on the relay side. I have that flow on https://pyramid.fiatjaf.com/moderated/ (disabled on that instance). It's not integrated with the git/grasp server yet, but will be.

2026-01-17 14:04:46 UTC

- reply

Or maybe I don't understand what exactly you mean by "moderation tools". You said it can't be just the ability to ban a specific person, so what could it be?
What comes to my mind is something like:
- allow friends-of-friends of maintainers to send issues/patches/etc;
- if you're not in the friend-of-friend whitelist you have to pass some arbitrary criteria like, I don't know, paying some satoshis, burning your CPU a little, passing a captcha or doing a phone KYC with some trusted server (all bad solutions, but what can I do);
- allow blocking specific people that pass the criteria above.

2026-01-17 14:04:01 UTC

- reply

Doesn't letting anyone submit an issue by just filling a captcha negate all the moderation stuff?

2026-01-16 23:22:39 UTC

- reply

All I see is a moving purple dot. I think it is trying to hypnotize me.

2026-01-16 23:20:22 UTC

- reply

ahahah, how was your experience?

2026-01-16 23:19:28 UTC

- reply

This stuff happens by default on GitHub (and similar places) because GitHub (and similar places) gives everybody the illusion that every user is equally legitimate once they've passed the initial phone KYC sign up process.
In Nostr there is no entity to say who is a valid user and who isn't, so in the long run everybody is forced to use some kind of filter, otherwise they will drown in spam.
I believe the best simple method for such filtering, when browsing an unknown place like a thread on someone else's profile (or git repository), is to only fetch stuff (replies, reactions, proposed patches etc) from the relays announced by the profile/repository you're browsing.
Assuming software is made to use that filtering method the default, then it becomes trivial for the repository owners to control whose comments get to show up to an unwary passerby.

2026-01-16 23:10:02 UTC

- reply

Has anyone tried to run Radicle? Am I the only incapable person who can't even clone their main source repository using their own product following the tutorial?

2026-01-16 23:09:22 UTC

- reply

What is that one?

2026-01-16 21:47:37 UTC

- reply

No, the default is the list that shows up when you click that button.

2026-01-16 11:59:44 UTC

- reply

`top` shows it around 1~2% for me right now, and my pyramid is actually constantly being hammered with requests.
The infinite loop thing was a bug in COUNT handling. It's fixed in the latest release, but I don't know if it's related to what you're seeing.

2026-01-15 18:55:15 UTC

- reply

Another good idea. I'll look into that, but it will depend on whether there is already some library that does the hard parts of streaming in Go, I don't know anything about the subject yet.

2026-01-15 15:05:19 UTC

- reply

Not yet, but will.

2026-01-15 15:04:54 UTC

- reply

Looks wrong, it usually does less than 1% CPU for me.
I just noticed the last release is buggy for some reason. There is an infinite loop somewhere.

2026-01-15 12:36:47 UTC

- reply

No, the only "changelogs" are the commits in the repository and these notes I publish, which are all nestedly quoted here. I never liked these changelogs in open-source projects, they're often useless autogenerated garbage that I assumed no one read.
But if I can help you in any other way please let me know. You can also join our group at https://grimoire.rocks/run?cmd=chat%20pyramid.fiatjaf.com%27Tnq7x2ZTgrPZWFrC

2026-01-15 12:31:36 UTC

- reply

It's too intertwined. But I'm curious: why do you want to run it in two different machines?
You can always run two copies and only enable the blossom server on the other, perhaps? There is also a simple pure blossom server at https://viewsource.win/fiatjaf.com/myriad

2026-01-15 12:28:08 UTC

- reply

It exposes a NIP-43 interface, but the tree stuff is custom and lives in the server only, there is no need to make these things "native" to Nostr.

2026-01-15 01:23:46 UTC

The new #pyramid v1.0.18 includes:
A way to customize the kinds the relay accepts:
https://image.nostr.build/58f48d349b18363e5113465ecf826c89018292ac1552db75a60800ef0a8fba22.gif
SFTP access for full control over events and blossom blobs (dangerous):
https://image.nostr.build/38af15f528ac309db701547312cc9a3a4573e50bb2c23eb9ccb2ebbf3b5ccd57.gif
NIP-86 support in all sub-relays (should work on https://landlubber.coracle.social and on https://nostrapps.com/relay-tools):
https://image.nostr.build/2de64c6c2556b8584186b54700bb0fc6f0011b2dda6fcd5584aaf4a8c7cdca74.jpg
And also:
- a way to pin events to the top of the feed of any of the relays.
- bug fixes.
#nevent1q…64vs

2026-01-15 01:08:41 UTC

- reply

I ended up implementing a way to remove a user but keeping them in the tree, that should probably be the default action when deleting someone.
I had to because I realized that @npub1jk9…lqz7 is flooding wss://pyramid.fiatjaf.com/ with a million notes with ugly images but he has invited a ton of people that I don't want to remove.

2026-01-15 01:06:25 UTC

- reply

Unfortunately I haven't seen your DMs, but Nostr wasn't really designed to do DMs, so I have that excuse.
Now, it can be adapted to do DMs reliably I think, but the implementation is lacking. With an extra 10% implementation effort we could have reached reliable DMs a long time ago, but probably haven't yet because the "let's all agree to use the same hardcoded relays" mentality took hold too much on most developers' minds.
In your analogy the shift required is turning UDP into TCP or something like that (but I don't know if I like the analogy).

2026-01-15 01:00:47 UTC

- reply

What do you mean? Is there another way to take a shower?

2026-01-14 20:36:24 UTC

- reply

I didn't read the full discussion, but npubs are definitely insufficient.
That's why we have nprofiles, nip05s or at least the dependency on a client having access to a greater ecosystem of recent events from other people that may provide updated hints to the relays of a bare npub, ideally through some form of local database. It's all on a best-effort basis, no mathematical guarantees, but I think that suits the messy world we live in very well.

2026-01-14 13:12:51 UTC

- reply

This was about the server side setup, unrelated to your VPN. How are you hosting it?

2026-01-14 11:59:19 UTC

- reply

Thank you for spotting that.

2026-01-14 11:58:49 UTC

- reply

That is the idea. I don't know what exactly you mean by a themed relay, but you can invite like-minded people and get them to do manual themed curation using the /favorites sub-relay.
Or you can get them to talk to each other either in groups or in the /internal relay about some theme, like a community.
Or you can get them to curate public posts submitted to the /moderated relay, ensuring those meet some standards.
If you are aiming for a curated experience you must be more selective with who you invite, as anyone invited will have the power to curate decide what goes in the thing. But if you're aiming for a community to live inside the relay you'll want to be more liberal and get more people.
Let me know if you need some crazy feature that isn't implemented yet.

2026-01-14 04:29:23 UTC

- reply

@npub1m3x…a5sf if you're compiling from source this should be working already: allowkind/listallowedkinds/disallowkind methods. It's not in the UI yet, and not released, but hopefully will be.

2026-01-14 04:20:12 UTC

- reply

That is a surprisingly hard problem, unless I make you adopt those users automatically when deleting their parent. Or what do I do instead?
If you're inclined to adoption you can do that manually already, by the way. You can have the same user be invited by multiple users currently, so if it is deleted in one branch of the tree it's still alive on the other.
On https://basspistol.org there is an example of that happening.

2026-01-14 04:16:23 UTC

- reply

Yes, I can make it configurable, but I never hit that limit myself, so please ensure that if you are running behind a reverse proxy you are passing the X-Forwarded-For header to the relay.

2026-01-13 18:40:05 UTC

Some #pyramid relays I know of that are up and running very profitably:
- https://spatia-arcana.com
- https://inner.sebastix.social
- https://basspistol.org
- https://nestr.nedao.ch
- https://garden.zap.cooking
- https://pyramid.fiatjaf.com
Please let me know if you're running one and how it's going. And send me your most adventurous feature requests such that they can be slowly worked on.
#nevent1q…psrs

2026-01-13 11:02:07 UTC

- reply

The links should be fixed in the latest commit. Thank you.

2026-01-13 10:19:35 UTC

- reply

Try this: https://viewsource.win/fiatjaf.com/nostrlib

2026-01-12 17:36:59 UTC

- reply

On vnak it will authenticate to any relay provided that you have set a main secret key at the top.
On nak you have to pass --sec and --auth. There is also --force-pre-auth for when the relay doesn't send an "auth-required:" message but you want to auth anyway.

2026-01-12 14:51:09 UTC

#grimoire is my chat client now:
https://image.nostr.build/f78e0689bb8ba7db8bdcbe0da245714b156075899b095385ed19067e74824af4.jpg
type `req -k 10009 -a $me` and it will show you a list of your groups, then you can just click on them.
https://grimoire.rocks/

2026-01-12 02:10:58 UTC

@npub1zh8…5aww is back (after a short 2-month interruption following the downfall of the Satellite CDN), now publishing exclusively to wss://basspistol.org because why not.
#nevent1q…af89

2026-01-11 17:00:07 UTC

https://viewsource.win/ now supports:
- displaying information about the latest commit
- a page with commit history
- fetching directly from github.com, codeberg.org, gitlab.com
- displaying README files automatically when no filepath is provided
- browsing the repository at any commit, branch or tag
https://image.nostr.build/35f2e5780e485595c14ef83b40a9d6cc5d6f848b2e5f9d803c2cd7062446ffa0.jpg
All without having to clone the repository or download a ton of stuff (if the server supports filters, which most should and eventually will).
Some examples:
- https://viewsource.win/npub1useke4f9maul5nf67dj0m9sq6jcsmnjzzk4ycvldwl4qss35fvgqjdk5ks/gitplaza should download directly from Codeberg
- https://viewsource.win/460c25e682fda7832b52d1f22d3d22b3176d972f60dcdc3212ed8c92ef85065c/amethyst should download directly from GitHub
- https://viewsource.win/npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6/gitview should download directly from wss://pyramid.fiatjaf.com or wss://basspistol.org
There were also some bugfixes. But let me know if more things are broken.
#nevent1q…c6gk

2026-01-11 13:38:31 UTC

- reply

"load profile" is for loading preexisting profiles created with "nak bunker --profile". That isn't well-developed yet, I should have probably hidden that button for now.

2026-01-11 11:36:25 UTC

- reply

It was released yesterday!
Do you have feature requests?

2026-01-11 00:36:22 UTC

vnak v0.2.0 is now widely available:
https://image.nostr.build/6b7ffdcef8861607eb3300d59b8b6796330b31c2cc85d3636dc62e3a48bf4149.jpg
https://image.nostr.build/c2094aa40330cd8e06e7276137a8825cb1d133266befbdbad8139224bdde55de.jpg
https://image.nostr.build/fb1b412c6b003fb4aec27413a0d037f053f3a970fe7bdcbee5edea4751268723.jpg
https://image.nostr.build/beb32dab7576d71e1e44481ffe5bd1e5449f0b4c10e150015c3847e002b17758.jpg
https://image.nostr.build/05af2b3f2edc24711500b1d3b0b7c894a0977d3ae1144feb99362ca9a159b67b.jpg
https://github.com/fiatjaf/vnak/releases/tag/v0.2.0

2026-01-08 13:12:42 UTC

new features in the latest last #pyramid release:
- primitive blossom support with a simple dashboard and upload button
- if you just publish an event with a date in the future it will be scheduled and automatically released at that date
- an option to automatically opentimestamp every note published to the relay
- an option to set the maximum size of events allowed
- many fixes
https://image.nostr.build/0e4f0ea0390ded77dd00c409a0c5bfb0d083e2fbdf6114528d17623264f3462e.jpg
#nevent1q…edd6

2025-12-31 23:49:42 UTC

- reply

We should run a prediction market for this.

2025-12-30 22:09:48 UTC

- reply

I don't get what is happening, are you running something or is it all in the browser? How do I test this?
Regardless of that we should be doing nevent1-sites, immutable, not npub1-sites.

2025-11-27 02:44:29 UTC

- reply

I think redundant coordinators will work, you just have to have more than one in your bunker URI. Also I don't want to run this coordinator forever, this was just a prototype until someone made one in Rust.
But for now I think the signers are the biggest problem. There may be something wrong with nostrlib that causes it to not reconnect to the relay after a connectivity glitch, so they stop reacting to messages. Or at least that is something that seems to be happening with `nak bunker` too and it uses the same approach, but I can't figure out what the exact cause is. If someone can help find the bug that will be much appreciated. Or rewrite it in Rust.

2025-11-21 12:03:44 UTC

- reply

Well, it's not just coding ability, there is also the ideas behind each software. When software took more work to create that work also served to filter out people who weren't very interested in it or hadn't thought enough about it and were still stuck in the first bad idea they had, also the laborious process of working served to educate these same people about many aspects of the thing they were making that would have an influence on the final product (or lack of, if they gave up or pivoted in the middle of the work).

2025-11-21 11:05:34 UTC

Vibecoding turned open-source software projects into a spam attack vector.
Before that you could expect high quality projects from a huge percentage of all apps, so acting on a message like "hey, I made this app, please try" was often a good use of your time.
These days there are at least 10x more of such apps, so assuming rationality the default reaction to "hey, I made this app, please try" will eventually become to not even look, or people will only look at apps from well-known creators.
Considering this it might be that rather than "democratizing" the creation of software vibecoding made it harder for a newcomer without reputation or reach to get their app in front of the eyes of others.

2025-11-14 10:09:33 UTC

- reply

Still happens every single time. I have no idea of what to do. It wasn't like that before, then it started happening without me changing anything. It also happens on other ngit/grasp codebases. Yesterday I updated ngit to the latest master to see if something was fixed, but the problem remained.

2025-11-14 00:16:21 UTC

- reply

Yes.

2025-11-13 23:19:27 UTC

- reply

Something is wrong. They continue to not work for me:
✔ wss://relay.damus.io no new events
✔ wss://nos.lol no new events
✔ wss://pyramid.fiatjaf.com/ new events: 1 user profile
✔ wss://relay.nostr.band no new events
✘ wss://gitnostr.com timeout after 7s timeout
✔ wss://relay.westernbtc.com/ no new events
✘ wss://relay.ngit.dev timeout after 7s timeout
I've tried this a million times in the last hours. My git clone URLs only have stuff up to some old commit and gitworkshop says it's out of sync or something, but `ngit sync` doesn't work either, I keep getting this:
updates: 1 user profile
fetching relay.ngit.dev/npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6/nostrlib.git ref list over https (unauthenticated)...
list: connecting...
list: failed: failed to connect to relay.ngit.dev: Operation timed out; class=Net (12)
fetching gitnostr.com/npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6/nostrlib.git ref list over https (unauthenticated)...
list: connecting...
list: failed: failed to connect to gitnostr.com: Operation timed out; class=Net (12)
(By the way, I keep getting that "1 user profile" every time, I don't know why but it feels wrong.)

2025-11-13 15:02:47 UTC

- reply

Something like this maybe: #nevent1q…jvaq

2025-11-13 14:59:50 UTC

- reply

This may be correct.

2025-11-13 14:57:03 UTC

- reply

Nothing is "needed", I'm just stating what is better.
Blossom is better because it's more efficient, it will work every time and not hit any barriers, it is much simpler to implement from the publisher side and from the reader side, it's also nicer to use other people's infrastructures the way they were designed to be used.
If we want this Napp model to become a standard (and we need that if we are to solve the problems I'm pointing) then these are all very valuable properties.

2025-11-13 14:49:57 UTC

There are a lot of people with such a wrong understanding of Nostr that it's hard to imagine these people like Nostr at all, given that the idea they have of what Nostr is is completely stupid.

2025-11-13 14:48:12 UTC

- reply

Some people are incapable of thinking about secondary or tertiary consequences of their actions.

2025-11-13 14:33:29 UTC

- reply

You should just make each Napp be an event with a list of filenames with hashes, then whoever wants to fetch that does it from the publisher's Blossom server. Once it's downloaded it's served from the local cache. There is no need or anything to gain from using this base93 madness.

2025-11-13 12:12:47 UTC

Once you see an ugly thing you can't unsee it for the rest of your life.
It will stay forever in your imagination, mixing itself with all your thoughts and memories and aspirations, infecting your dreams and your intelligence.

2025-11-13 11:34:22 UTC

- reply

(Of course none of what would be necessary if we all used native clients instead of "web".)

2025-11-13 11:16:26 UTC

This is a very good idea:
#nevent1q…x8f7
Not because of the way @npub1l3c…6vs2 frames it, but because it fixes the problem of web clients being what the user wants them to be and truly representing the user and not some new malware everyday loaded from a URL controlled by someone else.

2025-11-11 14:38:10 UTC

- reply

It was good while it lasted.

2025-11-11 12:28:05 UTC

- reply

@npub1jlr…ynqn I was going to invite @npub1syj…f6wl here but then I realized anyone else could snatch the invite code if I pasted it here.
Do we need a way to restrict invite codes to specific pubkeys?
Or maybe a process to just add someone without having to share a invite code?

2025-11-11 09:57:43 UTC

- reply

How can I generate an invite code from inside Jumble?

2025-11-11 09:56:16 UTC

- reply

How would that work? Do you have to provide a list of what you have somehow in order for it to work, right? How could that be done?
Currently there is `nak req --ids-only` which uses negentropy underneath but with an empty local set so you get all the ids from the server.

2025-11-04 12:23:17 UTC

- reply

Should be fixed on the latest master, or commit 210c0aa2820536f3bb947faed93a334c107961c9.
Please install it with `go install github.com/fiatjaf/nak@210c0aa2820536f3bb947faed93a334c107961c9` and let me know if it's still broken.

2025-11-04 12:12:18 UTC

If you love the terminal and you have a server and nothing else to do please try to run the totally new and much more fantastic version of the pyramid relay: https://github.com/fiatjaf/pyramid/issues/4
There will be a nicer installer later.

2025-11-04 11:14:28 UTC

- reply

Weird, neither nak nor the client library impose any timeouts.
After writing the line above I figured out that there is indeed a timeout of 10s in a place I didn't remember because I'm using a weird HTTP library only for the blossom stuff.
I'll increase that and make it configurable.

2025-11-04 02:32:17 UTC

Why is Android so stupid that if you open an app B from a link inside app A then app B is somehow opened inside app A, not as an independent process?
Now you can't alternate between the two apps because there is only a single app B opened. In order to go back to app A you have to press "back" on app B many times and hope it closes and control is returned to app A.
Even if you try to reopen a new instance of app A it will just bring you back to app B that is running inside app A. It's impossible to do anything.

2025-11-03 16:23:31 UTC

- reply

(By which I mean they are capable of ingesting a lot of content then they can spit out a basically infinite number of sentences that are grammatically very correct but empty of content.)

2025-11-03 16:17:34 UTC

Are journalists the precursors of LLMs?

2025-11-03 16:13:02 UTC

- reply

I think of litter bins as free storage devices too, but I've come to the conclusion that litter bins are a horrible protocol because every time I put things there they either come back dirty or I can never get them back at all.
That's why I always advise people against using litter bins.

2025-11-03 09:39:33 UTC

- reply

What do you have against NIP-34 and GRASP?

2025-11-01 16:09:33 UTC

- reply

Yes, it should be a local thing, shouldn't be hard, store it temporarily or something. People have tried to do it with events in the past, but that is an overbloated complication.
Still, it would be nice if well-behaved clients stopped mentioning people after a while, it helps no one.

2025-11-01 16:07:20 UTC

- reply

Clients should nudge users towards using such relays for DMs, and that is happening right now in every client.
Relays were never meant to be dumb in the sense that they wouldn't have any internal logic, that was an unfortunate choice of words and a huge misunderstanding that damaged Nostr a lot, it's a surprise to me that people could have thought a relay would be really that dumb, do no filtering, no access control etc and Nostr would still work? Nostr would never work if it was like that.

2025-11-01 16:02:25 UTC

Would be nice if clients dropped "p" tags from replies if the referenced public key hadn't participated in the discussion for some number of notes.
Someone mentioned me in a thread once, then 20 notes later I'm still having to read @npub1gcx…nj5z and @npub1226…grkj insulting each other about encryption.
It makes me think encryption is bad by itself. Encryption and satellites.

2025-11-01 11:11:42 UTC

- reply

You do have a way of knowing: if your DMs are going to a semi-trusted relay and you have to AUTH to it in order to read them (as you should under NIP-17) that relay may inform you about who has read them, when and from what IP address.
If we implement https://github.com/nostr-protocol/nips/pull/1647 that will also kill the attacker ability's to read anything in the first place unless he wants to reveal himself by registering his bogus device or publishing new encryption keys.

2025-11-01 02:19:50 UTC

- reply

There are a dozen of issues with this scammy paper.
The most important is that it only worked on a couple of clients that didn't check signatures. These clients only connected to a static set of semi-trusted relays and changing the relays they connected to would require a manual typing operation from the user.
For the attack to work it required victims to manually type the URL of the attacker relay, which makes it completely absurd.
It's like telling someone to visit "verysecretnotscammywebsite.com" and type all their secrets there, then read their secrets because the website leaked them and write a paper claiming that the web is broken.

2025-10-30 19:26:47 UTC

- reply

Aren't you also against inscriptions? Shouldn't knots also filter inscriptions?
Maybe it does that already, I don't know.

2025-10-30 14:57:58 UTC

@npub1jjn…vknm when are you going to host a nice reliable durable cheap GRASP-powered git server for the 4 of us that use ngit?

2025-10-30 14:22:36 UTC

- reply

Now it's much harder to play, every move I try to make it complains about something bogus.

2025-10-29 18:42:39 UTC

- reply

I think it's a bug, it should have matched by id only. But your report will be ignored because I've changed these database implementations multiple times (uncovering many bugs that lead to duplicated copies being stored) and in the end I've decided that I don't want to support Badger anymore.
I'm surprised you even managed to export all the notes given that the version you're using would try to load everything into memory before printing and if you had too many that would cause a crash. On the other hand there was a hardcoded limit of a million notes per query when using the eventstore CLI that you couldn't bypass, I guess that prevents the crash.

2025-10-29 18:38:33 UTC

- reply

@npub1000…vwqk when this note is showed in the grouped view in https://fevela.me it doesn't include the URL, but also doesn't include the preview box, so it's weird.
I kinda wish the preview boxes could be disabled.

2025-10-29 16:47:56 UTC

- reply

This was a great solution, although I have no idea how it could be implemented.
Same for the code that decides who won. I was almost amazed at it when it committed a mistake:
https://image.nostr.build/d56d6da2af7dfae2e3de8dcc79cecdd24273d9fb9ea587f60b7202ef1d06f730.jpg
By the way, if you get back to it, I think we need more space around the squares still (and smaller squares).

2025-10-29 11:25:45 UTC

- reply

Classic, everybody always tries to do this.

2025-10-29 11:23:39 UTC

- reply

It's almost as if kind:2 is back.

2025-10-29 02:33:44 UTC

@npub1jlr…ynqn what is the difference between https://github.com/coracle-social/zooid and https://github.com/coracle-social/frith?

2025-10-29 01:46:00 UTC

- reply

Looking pretty good, now just some more space around and between the squares (maybe shrink the squares a little?), disallow going through the square and make the computer recognize a win condition and that should be it.

2025-10-29 01:43:14 UTC

- reply

I'm not sure it's worth playing seriously on the computer, but I think having something visual is good so people can learn and then play on paper when they're waiting for their food at a restaurant or something like that.

2025-10-29 01:41:41 UTC

- reply

I've attached a picture now.

2025-10-29 01:39:30 UTC

- reply

This is pretty good, but yes, without some pathfinding around the squares and avoiding existing lines it doesn't work: #nevent1q…h2ca

2025-10-29 01:36:31 UTC

- reply

I forgot to attach an example game to the explanation: https://image.nostr.build/4e564b953116ae68259614dd2b6ad870ffd1a0a4f6a2f2b0691c6316a13b8881.png

2025-10-28 22:21:56 UTC

If someone is a real vibecoder and wants to show it please make this game in a computerized version: #nevent1q…ha5l

2025-10-28 19:29:03 UTC

- reply

Yes, I would never merge huge PRs myself either. And more clients is always better, even (or specially) if they are very similar but with subtle differences.
The problem is that forking and keeping track of upstream is an impossible inglorious task.

2025-10-28 11:44:43 UTC

- reply

Time to join forces with @npub1m4n…c2jl and launch a dissident Jumble fork with a different, name, design and feature set.
And remove translations.

2025-10-27 10:33:21 UTC

- reply

After doing all of https://spotify-scrobbler.fiatjaf.com/ I learned that Spotify doesn't allow new integrations to connect to data of external users.
In order to enable that you must sign a contract with Spotify and they will only look at you if you have 250k monthly active users.

2025-10-27 09:47:11 UTC

- reply

ahahah

2025-10-27 09:46:39 UTC

- reply

Yes, I did. I once said this was like a 2D version of Nim, but no one understood because no one else had ever heard of Nim.
Wait, now that I'm thinking, when I was a kid my grandfather once played a game with matches with me, he put a bunch of matches on the table then we took turns removing either 1, 2 or 3 matches until one couldn't.

2025-10-27 03:24:25 UTC

- reply

Can you make a computer implementation of this game? #nevent1q…ha5l

2025-10-27 03:23:22 UTC

Joseph's Squares is a game to be played on pen and paper. It goes like this:
1. draw 2 squares;
2. player A draws a line between one side of a square to a side of the other square;
3. once a side of a square has been connected it can't be used again.
4. lines can't be crossed ever.
5. player B draws another line again connecting one unused side of a square to an unused side of the other square.
6. and so on until a player can't draw any lines, that player loses.
In a game with 2 squares the second player can always easily win, but once you get that you can start playing with 3 squares, 4 squares and so on.
It isn't meant to be the best game of all times, just better than tic-tac-toe.
I haven't tried to play with triangles or other shapes but could be interesting too.
You have to be a reasonable human and not try to draw confusing lines. These confusing lines cannot change the game in theory but they can confuse everything and ruin the game for both players (it's ridiculous that I have to say that but many people to whom I tried to teach the game had this stupid idea, so I'm saying here).