Niel Liesmons

And in the case of a community forum, what would you do then?
Just have a community indexer or sth? And all the actual posts don't even actually live on the community server?

2026-06-07 17:43:03 CEST

- reply

In the case of users having their own pds:
Where does a user comment on another user's post live exactly?
And how do apps know where to even look for all comments? And how are they moderated?

2026-06-07 17:37:06 CEST

- reply

👀

2026-06-07 17:25:28 CEST

- reply

What's that first claim about?

2026-06-07 15:20:32 CEST

- reply

Ooooh 👀

2026-06-06 18:21:15 CEST

- reply

I like lexicons.

2026-06-06 18:16:59 CEST

- reply

This is awesome so far 👀.

2026-06-06 16:31:19 CEST

- reply

If we both have homeservers, is there any real cost in sending messages to each other? If not, then why/when stamps?

2026-06-06 15:34:44 CEST

- reply

How do you do access control with this? Me not get.
I need Proton drive, on open standard, on sovereign domain. With granular access control (per file/folder) for specific other profiles/domains/...
I don't need hashes and whatnot. If I have the actual domain, why would I make that compromise?

2026-06-06 14:53:54 CEST

- reply

No, I don't think the Nostr premises lead to any good solutions for drives, nor for key rotation.
Sth like Mlkut names + cryptree https://github.com/Peergos/Peergos/blob/master/papers/wuala-cryptree.pdf
= worth considering

2026-06-06 14:04:33 CEST

- reply

👌

2026-06-06 13:18:10 CEST

- reply

Fun fact:
I literally could sign you up and rotate the keys to you after.

2026-06-06 13:12:32 CEST

- reply

:wisp_salute:

2026-06-06 13:11:24 CEST

Family and community homeservers are a pretty straightforward foundation for #homeschooling parents that don't want to bring up their kids full-Amish style neither.

2026-06-06 13:08:15 CEST

• Access control on a file / folder basis
• On your encrypted drive
• Hosted on a sovereign, human-readable domain
What can you not build with that?

2026-06-05 21:46:11 CEST

Cool thing about mulberries is that they just start pumping out berries from a very young age, without it seeming to inhibit their growth much.
Transplanted this one in the #forestgarden just a few months ago:
https://i.nostr.build/IO5kmPsXaRv6Xy3k.jpg

2026-06-04 15:19:06 CEST

- reply

You copied that meme, didn't you? 😉

2026-06-04 15:01:50 CEST

- reply

Agree. And that'll be my family and online frens first. I can get them excited with something like this, no worries.

2026-06-04 14:19:27 CEST

- reply

Ideally you also don't need people to change internally for your app to be useful to you 😉

2026-06-04 14:18:01 CEST

- reply

Yesterday I was thinking: "Nostr isn't the right tech for this app."

2026-06-04 13:03:33 CEST

- reply

My 3D printer (when I was doing physical product design and engineering) was used for:
- custom mold making, to then make XXX time the same component with it
- pieces tailored to very "individual" individuals (children with cerebral palsy)

2026-06-04 12:54:21 CEST

- reply

🎯

2026-06-04 12:43:06 CEST

- reply

Yup.
That and the fact that :90percent: of what I want to read and write has the exact same format, filter, etc.. every day of the week.
I don't want to custom prompt things that aren't custom all the time. Let alone the unhabitual custom uis that come with it.

2026-06-04 11:48:14 CEST

This is still the super-app I need.
https://i.nostr.build/aiCfiyZgGQs0kukv.png

2026-06-03 14:49:39 CEST

- reply

And another one for the main courses? 😉

2026-06-03 14:30:10 CEST

- reply

Number one for me is that I can't build much on it.
But besides that point, Proton is sus in general and often open source in name only, with "source code" pointing to nowhere etc...

2026-06-03 14:12:28 CEST

- reply

Knowing what you need is the hard part.

2026-06-03 14:04:54 CEST

- reply

#shesaid

2026-06-03 11:11:44 CEST

- reply

Yup, it's communities with owners.
Moderation part isn't clear to me.

2026-06-03 10:56:12 CEST

- reply

I checked it out before, yes.
I like:
- the attention to style (must be Dutch folks, am I right?)
- them seeing the limits of Nip-72 and Nip-29 and just doing their own server-side thing.
- where the "repost to communities" thing eventually leads to
I dislike:
- the name
- the confusion regarding the posts not actually been targeted in communities by the authors themselves
- the comment sections not having transparent moderation (and seemingly relying on some form of outbox or hardcoded popular relays?), etc...

2026-06-03 10:51:03 CEST

When I go to a party I don't want to hear different music than everyone else, adapted to my interest and WoT.

2026-06-03 10:48:11 CEST

- reply

Wdym with hashtag moderation?
The hashtag has its own server and access controls? 😉

2026-06-03 10:41:15 CEST

I need communities to be like #tags.
- I can have multiple ones for one publication
- crowds behind each of them come together in the comment section
But unlike #tags, communities come with moderation, hosting, etc... And I as a creator can outsource those things to the communities I select (which can include my own, if wanted).
And it comes with a kind of global state around my publication, without needing a Platform to provide that.

2026-06-02 20:55:27 CEST

The wife you're looking for is more likely on a regenerative farm somewhere. Not in the city.
Playing your probabilities right isn't to go fish in the bigger pond. It's to go where she is literally fishing.

2026-06-02 15:51:18 CEST

- reply

You're not alone 😉.
Nostr has serious limits for this sort of thing though.
https://i.nostr.build/enPHLPHw5k5mJNkE.png
https://i.nostr.build/pHE0K79bclQjTWDh.png

2026-06-02 15:19:48 CEST

- reply

This open-ended set up is why barely anyone gets it at first glance, I think.
Nostr enforces signing JSON and sending/getting from relays. Which is in comparison, obviously, easier to grasp.
But being easy to understand, doesn't mean it auto-magically fits every web use case. The "simple" upfront constraints of websockets, JSON and needing signatures for everything require quite the set of complex of work around for 90% of things here, lol.

2026-06-02 15:00:31 CEST

- reply

I'm not really interested in solving single-person use cases. Also for that reason.

2026-06-02 14:43:06 CEST

- reply

If so, why would that be the case in that market and not in other FOSS markets?
What makes 3D software **that** different?
It's more expensive to build than, say, a note taking app, sure, but still. Curious.
It would explain why it is basically all proprietary software tho. So you might be right, just don't know why.

2026-06-02 14:38:06 CEST

- reply

I'm just saying there's a market for a FOSS 3D modelling product that is at the very least less crappy than FreeCAD and has the UX of SketchUp, while havingt the actual 3D constructs of something like Fusion360 behind it (not just surface modelling).
If it's FOSS, the you're selling hosting and compute 90% of the time, but in this case at least you're in a market where you can have some margin on it and still remain competitive.
I'm not talking about their more high end stuff like Revit and Autodesk etc... where you have tons of norms, regulations and need to 100% guarantee #nobugs, and where the cost of developing the apps is relatively way higher.

2026-06-02 13:54:43 CEST

- reply

1. I need sth open-source for 3d modelling product and architecture that isn't as slow, clumsy and buggy as FreeCAD
2. They charge high pricing for their products, that are in the end 'just' hosting, compute and access control.

2026-06-02 13:17:15 CEST

Linear, Notion, Figma, Autodesk, Proton Drive... are all 10x more worth building alternatives for than any big tech social platform.
They're productivity tools.
You can build transparent, honest, small businesses on things like that.

2026-06-01 22:11:16 CEST

- reply

https://i.delta.chat/#6FE626E39960C7D870329E8D6AD39C89D483BAC2&v=3&i=WCmXfoBqKr0TwgMP5rN_IHaJ&s=hiut2FhqyP3Qof3ae5rDjDKg&a=in9bgyspy%40nine.testrun.org&n=Niel+Liesmons

2026-06-01 22:05:38 CEST

- reply

If you get everyone together in one place anyway, call it a public community and enable ways to to set up and organically discover other ones.
Keep communities intentionally relatively small.
(less drama, more signal and better monetization strategies)
Enable publishing certain content types in multiple communities at once.
(blind-spot opportunity that most builders on open protocols keep missing and keeps them trapped in a false "public-square VS niche without reach" dichotomy)

2026-06-01 21:37:31 CEST

- reply

Hehe, yes, own app 🔜
I'd loved to actually test delta chat with someone tbh 😅

2026-06-01 20:52:56 CEST

- reply

Good ACL.
Good PKI.
Open content types.
Signatures where needed.
Pay with anything.
For individuals and small groups.

2026-06-01 20:48:37 CEST

- reply

Yup, I think initially focusing on private contacts, chat groups and small public communities is the better way to bootstrap in any case.
Why do so many of these projects want to bootstrap a public square?
It's freaking hard and in the case you succeed it'd be only to realize, once again, that every public square ever trends towards corrupt centralization, high school drama and whoredom.

2026-06-01 20:22:08 CEST

- reply

If I'm not wrong, that's kinda how pubky does it, right? But in public folders, on the homeservers.
#noob trying to understand it all here.
But I just want a root ID that points to my drive (homeserver), then have data on there, part of that data being profile metadata that I can choose to have discoverable by (some or all) others.

2026-06-01 20:17:47 CEST

- reply

Yup :110percent:

2026-06-01 20:17:25 CEST

- reply

The latter seems realistic and aligns with the number one business model I can see around all this: hosting + convenient security

2026-06-01 20:13:30 CEST

- reply

Ideally you have access control on that stuff.
This what people loved about Facebook (until they messed that all up), the ability to choose woo cash see what.
It'd be a bit like an auto-fill UX, that you confirm our not, when you want to post in some forum or join a chat.

2026-06-01 20:06:06 CEST

- reply

Lol, champagne problem, never heard that one before. Accurate 😅

2026-06-01 19:29:38 CEST

- reply

Many ways out of that, no?
(especially if you sponsor registrations for onboarding users for a big part)

2026-06-01 19:21:43 CEST

- reply

Ton of value in interop profiles tho.

2026-06-01 18:34:57 CEST

- reply

Yes please 🙂

2026-06-01 17:39:25 CEST

- reply

#smallworldtech

2026-06-01 14:56:48 CEST

- reply

I like homeservers. And there's many use cases where, in that case, signed events aren't really needed.
For us at Zapstore, I think it's valuable to sign releases in any case, even it was built on (or would embrace) a more homeserver-type protocol.
I like that we're building on key-based communities because that at least leaves some doors open to protocol improvements, without losing it all.
(because keys are IDs with agency and they can say stuff about themselves and point to whatever)

2026-06-01 14:31:25 CEST

- reply

Had the same question. 👍

2026-06-01 14:29:09 CEST

- reply

Before my daughter drew all over it, this little sketch was going somewhere 😅
https://i.nostr.build/WdDt3gK4nDTwaMlp.jpg

2026-06-01 11:05:56 CEST

Best way for me to organically discover great music and books is to just check out the publication houses where stuff I like was published.
Same for niche content, I find the most surprisingly useful stuff by browsing forums or communities I previously found good info on. (duh :padme: )
All the fancy algos and complex personalized filtereing protocols still pale compared to just having actual topical "places" where people can target stuff.
→ Just give me open, interoperable "Publication Houses"

2026-06-01 10:45:40 CEST

Community/group homeservers.
- It's how people organize by default in the real world
- It's by far the best level to get a centralized state for social data and moderation right (and no, people do NOT all want to see their own version of things)
- It combines events and media
- On an open protocol, users can target one publication to more than just one of these (avoiding the false choice between reach and niche most platforms push creators into)
Only thing individual users should (be sponsored to) host while onboarding:
- profile pic and metadata
#nevent1q…vc0p

2026-06-01 10:09:38 CEST

- reply

1. Build on your go/chess/dam inspired sigil instead
2. Make the O shape in it more prominent
3. Find a better encompassing brand name that start with an O
Something like "Outpost" or so, with some geeky chess meaning behind it too. Don't know.
My 2 cents.

2026-06-01 09:31:46 CEST

- reply

Not easy, that's for sure :PepeLaugh:

2026-06-01 09:26:40 CEST

- reply

I've always kinda seen it in the abstract like that:
Most comms = email
→ one-time convos, commercial follow up comms, ...
Anything more permanent (that often needs more content types in there too btw) = group/community server

2026-06-01 09:06:57 CEST

- reply

Same. For mostly text-based chats for work, that is.
For family chats that have basically a full drive of pics in there, different story, and I kinda would want all that on my own sovereign drive, automatically.

2026-05-30 14:18:37 CEST

- reply

Just checked again, and yup names are nice 👍

2026-05-30 13:58:04 CEST

- reply

Gotta make the names (and visuals for them) sexy enough 😜

2026-05-30 13:57:14 CEST

- reply

Mmm, yea. But is RNS sovereign enough over just the usual domain names etc..., to justify any move/adoption from orgs at all?
If the value prop is "sovereign", you go all the way, no?

2026-05-30 13:31:36 CEST

- reply

Yeah. Anything more serious beyond "shit posting", like what we're trying, at the very least leads to good reality checks.
You see pretty fast how much lower the value prop is without bringing in external content.
You see the flaws of nsecs when you have to build customer service for devs signing releases. Like what do we even tell teams that keak their key? "F you, start over, we told you to be careful"
Etc...

2026-05-30 12:09:04 CEST

- reply

Yup. So far Zapstore seems to be the only project that brings in external media like that, with analytics, comments etc... all gathered in one place (for outreach like you say...).
The Mostr bridge in some way too I guess, but no clear picture to send to the external creators.

2026-05-29 14:43:24 CEST

- reply

nielliesmons.16 (signal)
nielliesmons (telegram)
[email protected]

2026-05-29 13:15:33 CEST

- reply

Declarative.
Requires asking apps to ALL play nice.
Extra UI space in several apps for no extra relevant info.
Vague privacy zone for newbies/normies.
....

2026-05-29 11:18:18 CEST

Client tags are noise

2026-05-28 22:10:13 CEST

- reply

I'm getting my most valuable info from:
- forums who's UI lives in the 1990s
- group chats
- nostr critics here 😅

2026-05-28 21:16:35 CEST

- reply

Dead is different then having obvious flaws that are (as observed) not solvable within it's scope.
It means it can be superceded by something that doesn't have those flaws, for those interested. Especially if the alternative starts with a use case that requires zero network effect.
Just talking for me here. I want the best, simplest and most future proof solution fire my needs and wasn't too distance myself more from the Nostr culty vibes then be part of it.
Google drive has many obvious flaws too. It ain't dead. So we'll see in the broader sense of what you mean.

2026-05-28 21:10:58 CEST

- reply

With bare key pairs it's discussions that (clearly) go nowhere, no viable solution.
Might as well get over it, acknowledge just keypairs don't suffice (even if mostly cold imo), look into what is promising instead and build on Nostr in a way that is open-ended enough for it to be morphed into something better.
I'm still wrapping my head around it but I think a combo like this has promise:
- sovereign domain names, à la Mlkut
- honeservers hosted u der those names, for profiles and communities
- different kinds of events, signed or unsigned, encrypted or not, with good ACL, on those servers (next to the use case of just being a badass Google Drive replacement already).

2026-05-28 20:39:21 CEST

- reply

Lol, that typo went total Yoda on you 😅

2026-05-28 20:28:55 CEST

- reply

Until you want to build a business on it.

2026-05-28 20:27:32 CEST

- reply

Yeah, so just noting the logic here for myself:
1. You need key rotation if you're serious about sovereign web solutions
2. Ideally with a readable and capped set of names that is still huge enough
3. For that you only have two options: central registry VS blockchain
4. If you go with a blockchain, you go for the most reliable and assumed to be longest lasting one (Bitcoin)
5. If you want scalability and cheapness, you go with the most efficient and trustworthy chain tied to Bitcoin, not Bitcoin itself

2026-05-28 20:20:24 CEST

- reply

Thanks 🙏
Gonna check FIDs, but I think I agree with your conclusion.

2026-05-28 20:12:06 CEST

Lol, #noob me realized this 2 years ago.
Only to then underappreciate its relevance for 20+ months after.
I should read my own notes more. #nevent1q…59u7

2026-05-28 20:05:15 CEST

1. Don't have key rotation
2. Attract only the reckless early adopter crowd
3. Be unattractive to the rest of the crowd

2026-05-28 19:55:46 CEST

- reply

What's the alternatives of mlkut?
There must be other "crypto" projects attempting sovereign (readable) domain names with key rotation, no?
Is your main value prop that mlkut is tied to the most reliable, biggest blockchain, but in a cheap and efficient way?

2026-05-28 19:49:52 CEST

- reply

Lol, yes.
Year after, sovereign domains.

2026-05-28 19:17:31 CEST

- reply

Yup, on web.
zapstore.dev/studio (logged in)

2026-05-28 16:38:36 CEST

- reply

Agree.
And for Personal feeds I think Instagram story UX is the way (for recent ones). Where users can spotlight selected publications publicly (or even to restricted sets of contacts).
This opposed to having to scroll a feed of ALL the stuff of all the people you follow.
Feeds works well within the context is visiting 1 profile or 1 group/community. Trending feeds also work way better in the latter.

2026-05-28 16:29:32 CEST

- reply

❌ Relays
❌ Feeds
✅ Interop Communities

2026-05-28 09:51:39 CEST

- reply

Wait, do you mean this is a problem or good thing? Confused 😉

2026-05-27 20:58:46 CEST

- reply

Yeah :zapstore:

2026-05-27 20:57:02 CEST

- reply

What version of Zapstore?

2026-05-27 20:03:53 CEST

- reply

If you have sovereign domain names, than why even have the hash? It's a valid question, no?
Noob here.