people interested in nostr wiki:
thoughts on allowing collaboration on an entry?
Discussion
say I create a document about "Stalin" and I add ["p", "<nostr:npub1mygerccwqpzyh9pvp6pv44rskv40zutkfs38t0hqhkvnwlhagp6s3psn5p >", "editor" ]
this would mean that the most recent version of my version of Stalin is whatever comes back from the REQ { "#d": ["stalin"], authors: [
G could also add me as an editor, or not.
A side-benefit of this is that if I don't care too much about an entry and I'm just making a minor correction but don't care too much I can go UDP style and make the edit and add the person's entry I'm forking as an editor, that way I wouldn't have to go back to check if the person accepted my changes to defer my entry to them.
cc nostr:npub1ye5ptcxfyyxl5vjvdjar2ua3f0hynkjzpx552mu5snj3qmx5pzjscpknpr nostr:npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6 nostr:npub1m4ny6hjqzepn4rxknuq94c2gpqzr29ufkkw7ttcxyak7v43n6vvsajc2jl nostr:npub1r0rs5q2gk0e3dk3nlc7gnu378ec6cnlenqp8a3cjhyzu6f8k5sgs4sq9ac
+1
Can you even see this convo?
i can
He has us muted, tho. He seems to miss most of the convos I'm in or his comments are oddly lacking in awareness of the wider discussion.
ah yeah, lol... i am pretty sure my mute list has been empty almost a month now, acquired several layers of skin in the time on here
this is something that i think everyone gets after a 3-5 monhts on #nostr
it's character building, just wait until the actual asshats show up, we are gonna be Wolverine to their limp pale baby flesh
also, in the last year, or two maybe, i have had several times it happens where some scammer is in my DMs and by the end of it i have them insulting me like a bitch, and i laugh my ass off because they couldn't touch me
also, i'm looking forward to the nostr troll killing squad, it will be fun i will do more than a few outings
Who has you muted?
Karnage.
he also mutes me, it's hilarious
He floats into technical discussions and repeats things that had already been discussed and it's like having a half-deaf man wandering around the room, spouting non sequitors, because the battery in his hearing aid died.
well, you should be used to it by now
nobody is herding us here, and he clearly has blinkers on, it's quite an interesting scenario that is more or less unprecedented since the spooks tried to close this gate outside of the matrix
Oh, I am. Especially funny is that he has these groupies that also have us muted and cheer him when he shows up a day late and a dollar short and it be like
Sounds good, I don’t see a catch here, are there any downsides you are aware about?
It removes accountability because nobody is making editing decisions.
I like that, but maybe someone might not want to be listed as an editor.
yeah, that's a good point, I mean, you can't prevent anyone from p-tagging you, but it would have to be clear that editors don't necessarily agree.
perhaps a good balance is that they are not displayed anywhere in the UI unless the most recent version of the document was signed by them
Maybe you should just defer at the same time you're pushing your edit? If you trust the person enough to give them "write access" to your wiki you probably already trust their article enough and/or trust that they will merge your edit.
I don't like this approach because you have to specify who can edit ahead of time which seems like an anti-pattern.
Simplest solution is probably to just point your entry to someone else's more complete/better entry.
This might result in hill climbing towards very high signal to noise entries.
You don’t have to specify ahead of time; you can add people as you go, and anyone can permissionlessly edit whatever they want, this is just a scheme to say “whatever these pubkeys say is as good as if I had said it”
Ghost-writing wiki pages seems influencery and intransparent and precludes debate. The hurdle to remove a change you've merged is higher than just ignoring it or discussing before you merge.
Changed my mind. Don't like.
Too consensus-y/commie.
But I can just ignore this feature, if you add it, so.... 🤷♀️
fuck consensus, let's have the whole shebang
If you add this feature, then I can market myself as someone who doesn't use it. 😂
Laeserin, edgy and real.
I said. What I said.
These words are all mine.
The buck stops here.
Loving the feature, already. 😂
It would still be more effective to point to specific entries rather than to trusted pubkeys.
Then you follow the chain of pointers to get to the entry that everyone in the chain has helped refine and agrees is the best
that's what is already defined in the NIP under "defer"
I guess this is definitive proof that I'm a bit retarded
🤔 
It works for me
what browser are you using?
I think this could work. although I'm not sure how necessary this would be yet. Id like to see a lot more forking / diff tools / chaos in the wiki pages before we try to make things easier
I like the competitive/debate aspect and the ambiguity around controversial topics.
yes, this
I think the right approach to this is embracing biases full-throttle
my entry about alejandro muyshondt is fully from my perspective, not about when he was born and stupid trivia like that
I think we need comments first. Id like to see some heated discussions on pages (or event specific parts of the page)
Highlighter, to mark and comment on particular sections.
we would need some kind of history. but this would be awesome
Highlights should probably only apply to sections that have not been altered since the highlight was created.
Tricky, but yeah... awesome.
highlights reference the specific revision you had and the text you've highlighted is save on the highlight event, so it's very much preserved 😀
Is there a kind for replies to highlights? if so this really makes me want to implement highlights in noStrudel
How crazy would it be if we could have a conversation about a specific part of a wiki page or event a kind 1 😀
In damus we just have kind1 replies to highlights. I guess this would look bad in other clients that don’t support this, but maybe they should have a fallback note in those cases.
Ah, Damus has highlights implemented? Sweet.
I'm not sure how I feel about using kind 1 as generic replies. We do have nip-89 so it is good for discoverability but it still feels like creating a bunch of noise
Yeah i was originally against it but having a custom reply kind for every note variant is also kind of annoying
Q1: Is it possible: Similar to medium long form “most highlighted” segments, on nostr wikifree there can be “most controversial” or “least agreement” highlights.
Q2: might git log delta greentext, and redtext be useful somehow? This implies comparing article X version A, and article X version B
Q2: yes, usefull, nostr:npub1m4ny6hjqzepn4rxknuq94c2gpqzr29ufkkw7ttcxyak7v43n6vvsajc2jl asked for it and I'll do it (probably after Prague)
Q1: yes, very much so; this is not trivial and it spans a bunch of different use cases and data comparisons; I plan at some point writing this as a DVM because exploring controversial and disagreeing positions is of extreme value, particularly where we're going, it's going to be increasingly important -- integration of opposites, etc.
not noise, a tree
it's not.
the internet siloed around use cases because each platform wanted to own the data.
if I follow you and you are having a discussion around a book's quote, I of course want to see that and be able to follow along and participate in the conversation.
This didn't happen before due to a technological and financial limitation, but the UX of being able to discover a new book in my twitter feed because you are having a discussion on Goodreads, or Amazon, would have been kickass.
Siloing kinds reduces the network effects of nostr, and means that each use case pretty much must bootstrap a large part of it's own network effects.
NIP-31 + NIP-89 make this discoverable, functional and contextual at an O(1) cost for developers.
furthermore, this is one of these cases where nostr can do something platforms CAN'T. It might feel a bit counterintuitive to allow data to flow unconstrained from the use cases.
But aren't we building stuff that goes beyond the previous capabilities of the internet?
For example, today I found out that a couple of days ago I was participating in a git issue from my kind:1 client. I didn't realize this was happening within a github-like context, but all the needed context was there for me to participate and (hopefully) add something of value to the conversation.
It just magically worked and a "github" issue that would have had maybe one or two comments ended up with a fruitful discussion that added a bunch of color and nuance.
💗💗💗💗💗💗
I keep coming back to some type of kind1 inheritance. Either new kinds extend kind1 so clients can fall back to only showing what they know what to deal with. Or some sort of trait that would allow multikind notes where it is kind1 and enhanced by kindN where clients that know to do special things with the new kind.
I can't disagree with you on that, and I do like the idea of comments being used for discoverability. However it feels wrong using kind 1s as replies because most clients expect them to be social text notes, it already takes enough work filtering out what is a reply and what is a root note
It feels like we should have another kind for generic replies
Otherwise, it's not that different from Wikepedia, where there's agreement about content.
The novelty is in two or more people disagreeing and just leaving it like that because we allow for the possibility that neither is completely right or completely wrong, so read both and decide for your self.
Or read This Popular Opinion and then the 8 dissenters
Also, there's accountability in people looking at diffs before merging. Like a review.
A composite view would be cool. Where you could see which parts have differing entries, within one pseudo-article.
Should be easy to build. 🙈
Like, with a right-click or something.
Or maybe that's too distracting.
maybe better a button on the top right of the article
It'd be like those pro/contra newspaper editorials, but with each side expanding/honing their argument over time.
Make dialectic great again.
It wouldn't be a direct debate, you see, where one side can just say the other person is a doodoo head and win. Their article has to be convincing and readable alone.
yeah that would be sweet
Ha ha. Add a friggin heat map option, where you can see which parts are most controversial by the background getting darker. From ⚪ to 🟠.
with a fleshed out wiki ecosystem I'd love to analyze and extract the most divisive content - basically and intersection in content but a schizm in interpretation. Where are ideas initially published to and then broadcasted? How do groups curate their narrative? How robust are groups to external community brigading?
Being such a small community of early nostr users we don't need to worry about narrative takeover as much. However when nostr gets popular, connecting to the Damus relay or any other free relay for a wiki can quickly turn into quite the shitshow. Which is why I think nostr:npub1zuuajd7u3sx8xu92yav9jwxpr839cs0kc3q6t56vd5u9q033xmhsk6c2uc's nostr.how is such a great opportunity for a nostr knowledge base. A relay of white listed users creating content which can exist alongside the static content. Write articles about nostr at whatever level they care or as specified by whatever community guidelines for posting
Another great example is nostr:npub1stemstrls4f5plqeqkeq43gtjhtycuqd9w25v5r5z5ygaq2n2sjsd6mul5 or nostr:npub1yfg0d955c2jrj2080ew7pa4xrtj7x7s7umt28wh0zurwmxgpyj9shwv6vg having a community knowledgebase for its own users - not only for how to navigate their own app, but knowledge about anything music creation and colaboration. Beyond the service is the creation of some community with shared values, which encourages continual use of the service.
You can also just use WoT, or filter down for a particular topic within a particular app.
WoT can be gamed and gets more difficult with more users, especially if you're not caching it. As tools develop it will get more difficult to game, but that just means that those with intentions for to infiltrated are that much stronger. It is the use of both top-down (relay level- owner, leaders of communities) and bottup-up (user level- WOT with tweakable parameters) to help best curate signal at multiple scales.
That's true. Walled gardens tend to be high-signal.
10% of them, and the other 90% are dejavu
Well, each group gets the garden they deserve.
My Slack group and Simplex chat are deafeningly high signal.
that is because you are administrator and you have discrimination
Which is the whole point, a gradient of controls for both the users and relay operators. Users decide how wide open their feed is from external users and they also subscribe to relay operators that align with their curation ideals.
the point of nostr focusing on delivery and not filtering is that people can put filters in the client, in the relay, wherever they want
curation then becomes an opinion and a feed itself, instead of a gate thorugh which nobody can breach
100000%
you design your feed, fully voluntary
keeping the layers of a stack simple is something that takes experience
> that is because you are administrator and you have discrimination
Meant that is the who point of the silos. I'm arguing that we don't actually need to fight censorship anymore unless you are actually talking about nostr at the protocol level.
Sitting on top of the protocol level is that curation. Of course it is voluntary, but interesting things will happen there, observable and broadcastable by anyone else.
the main difference is that nobody can control the trafic on nostr, it's just a protocol, it's not a corporate intranet
Feel like I spend too much time on here discussing whether girls are stupid and should have rocks thrown at them, with men who couldn't reason their way out of a box.
Trivial midwittery abundant.
Distraction. Noise.
Precisely. Curate what you want to see, delegate mass curation to who controls your relays. If you want to see everything, personal and free relays are where its at and you'll broadcast whatever you want there as well.
> WoT can be gamed
how?
Goodharts law: "When a measure becomes a target, it ceases to be a good measure". Clients, open source or not will have their WOT metrics known or approximated which are now targets to optimize.
Whatever the system that is constructed to keep bad actors away is, its its still a proxy for human trust which is fallable.
Sure I trust myself for most of my curation, moreso with a WOT. However I would like to be signaled and not encounter known bad actors encountered by my community/s.
Considering the long game strategies and the long history of inciting intra group division and then letting them destroy themselves from the inside out. Can we solve it completely? No, but we can build robust tools to help guard against it.
I think this is a bit of a handwave; some things are brutally asymmetrically expensive to fake. WoT is, much like cryptography, P!=NP.
Interesting. I'll have to come back to this.
It could also be a difference in definitions where I'm actually not sure if I'm using the same one others are. I couldn't work out how wikifreedia does WOT from the code, could you mention that here? If you know any resources about it off your head I'd also be interested in reading them too.
Would only make sense within one section, though, and between two versions. Otherwise, it could end up 🍝.
We need to embrace biases. You've mentioned there being no global state. In many cases there is no global truth, but smaller local truths emerge from groups with a set of cohesive values. By embracing biases we will encourage polarized opinions, but this is perfectly fine because we don't need to fight censorship here. At some point I'd imagine a group with overlapping and consistent ideals would want to curate their narrative and content. This is also perfectly fine because they can close down their relay to their trusted members, but keep it read only to everyone else. Nostr members external to that community can find value, broadcast it to their exclusive relays and further develop on those ideas.
You don't get the best meal by throwing every ingredient and spice into it. Communities will find the right amount of mixing when they are provided the right tools for it.
For collaboration stuff, I like the idea of creating shared access to a Pubkey managed by a nsecbunker. That way the owner can easily add or remove access
yeah, but that has more overhead, this would be an in-protocol solution
the nsecbunker approach makes more sense imo for teams and stuff like that where the collaboration is more structured/formalized
More overhead for people writing the articles, much less overhead for people reading and much less complexity in-protocol.
The users main key would act as a nip46 ephemeral signer
Use a combined key with MuSig to publish.
Or just create a new key that all collaborators know, publish from it, then add it to your list of good wiki people and/or defer from your articles to their articles.
Yeah, like our project npub. Multiple people hold that key and we're going to use it for technical documentation.
But that is clearly a group of people, corporation, bot, etc.
Weird is when one npub seems human and establishes trust in Space A and then lets others write for them in Space B. Sorta creepy.
Let the entry with the longest history become the 'primary' entry for the topic, leaving the others to be found in the history. I'm thinking in terms of how the Bitcoin blockchain operates, where ~ every 10 minutes a hash function is confirmed and the next block is established. Maybe every hour the wiki system runs a hash function and locks down the content on each page, and the link to the version of the page with the longest history of hash confirmations becomes the destination for any links looking for that page. If no change, no hash function operations are needed so the system saves resources by not running it. More active pages might accelerate the frequency to prevent too much branching.
I'm just brainstorming. While collaboration is critically important, what matters is whether a page and its content are secured and not subject to branching and manipulation. The communications related to trying to get a page or article content agreed by all parties would need to be a "talk page", similar to how Wikipedia does it, but more formalized and in the open. I would love it if someone would put together an instance of Apache Wave for this that works on Nostr. Loved that system.
Let’s see if I can help on some subjects.
Yes.
Add a d-tag with a uuid. That uuid corresponds to the subject of the document.
Then create a new private key to sign the event.
Anyone you share the private key with can continue editing the document by creating new event with the same uuid and same key.
Everyone else can “fork” the document. They do it by creating a new event and using the same uuid in the d tag.
People don’t have to use the same uuid when creating a new document on the same topic. But they have incentive to do so. If your document about Blue Whales already got traction, I would want to use the same uuid to suggest my document as an alternative document for the same topic.
When creating a new version, one can also add his personal pubkey in a p-tag so people know who created this version.