Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Seedling – A Decentralised Medium (seedling-d.app)
56 points by dnpotter1 10 months ago | hide | past | favorite | 44 comments
Initial version of my decentralised publishing platform for articles and blog posts - like a Web3 Medium. It's open source and [almost] fully decentralised. Runs on the Polygon network.

Current version lets authors publish markdown content from their GitHub account and lets users tip articles. No wallet or GitHub account is needed to browse content.

Authors link their GitHub account or organisation to their wallet address via oauth and an on-chain user registry, after which they can submit article URLs directly to the on-chain content registry. The content registry ensures only registered users can publish, and only from their registered GitHub account.

By adding different hosting options and expanding the feature set, the ultimate goal is to create a platform for independent journalists to publish without fear of censorship or repression, and to fund their journalism through direct tipping, subscriptions and profit sharing models.




If there is requirement to link a github account, then there isn't any kind of decentralization, as github is then the one centralized point.

And well, the original idea of blogs was completely decentralized (“there is this thing accessible through HTTP”) with bunch of ad-hoc API surfaces for metadata and “pingbacks” (ie. back references). Then in the last few years there is Medium.

Bottom line: You are solving solved problem and have some kind of path-dependent view of that problem space that does not really make sense.


I agree, if you have to use github they have pages.

A better thing to use to be decentralised is ipfs. Which worse comes to worst if your file gets unhosted then do your own node.

Or Nostr would be good and might avoid needing a blockchain (and thus the expense and complexity)


IPFS and Nostr are both brilliant but they have their limitations. The main issue I see for an app like seedling is that by submitting your content you lose control over it. You cannot edit or delete.

Ideally many different hosting options should be supported.


You could create a Nostr protocol with a delete message, but the problem there is of course it would be a "soft delete", since anyone scanning the messages under the hood can see the original. In some sense a regular site is like that (you could see it on Internet Archive, but maybe that is less likely to be done, and it is coming from a secondary source anyway).


It's an interesting project. I think you should mention that your contract takes a 10% cut of the tips. Also, why not let any wallet publish? Why the dependency on GitHub?


It was an easy way to get the app up and running without needing to worry about hosting, editing features and user profiles. I.e. release minimal functionality then expand if people like the app. Intention is to add other hosting options, like IPFS, at which point it will remove the dependency on GitHub and decentralise hosting.


From my understanding, they're relying on GitHub for the hosting of the blog posts themselves.


So the opposite of decentralized then. Why do crypto people love inventing new middlemen to take a cut? Oh right, they’re the middleman. That’s why.


> Why do people love inventing new middlemen to take a cut?

FTFY.

Remember many on this site are partially responsible for the ad-ridden web and a proliferation of all sorts of spyware.

OTOH, we wouldn't skewer someone for charging for yet another toy SaaS and posting a Show HN, so... let's be a little more balanced.


I don't think it's unbalanced to call BS on a service claiming to be decentralized that in fact relies on a single centralized service they do not control, twice over. (Once for auth, second for content).


Yeah, I guess it’s more accurate to call it an [almost] fully decentralised discovery service rn. Decentralised hosting should come later. The almost is for the oauth server, which can be improved by adding other independent trusted servers in a permissioned network.


Totally fair! Sorry if my initial response was perhaps a tad harsh... my bullshit-0-meter goes to a million for anything even vaguely crypto/blockchain related.


But why specifically GitHub, instead of just git? Every git hosting provider could work with this model


That’s true but I am trying to discourage people publishing other people’s work. Hence users need to link their account through oauth. Other hosts can be added.


> I am trying to discourage people publishing...

A common motivating concept for decentralized is no authoritarian oversight to deplatform you.

You're seeking editorial influence (control) from the start.

Setting aside what users want versus what you want, as a platform technology you need to be careful here.

You way want to brush up on the safe harbor laws for Internet infrastructure/pipes/hosting firms and technologies versus the lack of safe harbor for platforms with desire or ability to exercise editorial influence or control on a user or content basis.

You can navigate this, by having reasons such as "I'm not editing, but by associating users with a non-anonymous identity they will self moderate", as you've noted. You can, with such reasons, still have a bit of cake after eating it. But be careful not to veer into actual editorial control unless you're prepared for full liability.


Useful. Thanks


I'd rather use Nostr and a client focussing on long posts :) Like https://habla.news/


Oh, that's really cool! I wasn't aware of Habla - thanks. I'll check it out.


Is that habla as in idiot (feminine) in Arabic?


I’d bet is Spanish.


Quick note that it might be better for this audience to link to the code: https://github.com/Bubble-Protocol/seedling


Thanks.


What does the chain part provide that's not available right now through any page with an interledger link in meta and maybe an .eth link if you want to be fancy with anticensorship?


Lets him skim 10%.


Lol! It’s not my intention for this to be a get rich quick scheme but there are running costs (user registration and TheGraph network payments). If the platform is to be self sustaining it needs to cover its costs.

Is there a reasonable long term platform fee that you would find acceptable? Say 1% or 0.1%?


So it sounds like you can DoS the service through user registration and graph usage?


Yes, that I need to solve


So the decentralised system relies on you as a SPOF dealing with abuse and paying the bills?


Only in the short term. TheGraph is migrating from a centralised service to a decentralised network and the community can run alternative oauth servers to help mitigate attacks. I'm no cybersecurity specialist though.

Also, the data is always available on-chain and the protocol is open so other people can write an alternative ui with their own blockchain query capabilities. Like Bress.xyz does with Mirror.xyz.

Payments can be automated on-chain.


It provides a discovery service


How is this different from Mirror.xyz?


Other than no funding and a team size of 1? ;-)

Slightly different vision and approach but both are publishing platforms. There’s plenty of room for competition in this space, I think.

Monetisation: Seedling uses direct tipping instead of NFTs. Plans for paid subscriptions to private content. Crowdsourcing will offer share of tips for investment.

Architecture: Seedling is primarily a discovery service and will allow content to be hosted on different platforms. Seedling has no backend server to deliver content.


Interesting, thanks for sharing.

Any plan on integrating other chains? This would be great on Solana


I considered Solana. Lower fees allows for micro tipping. Polygon is pretty cheap though, less that 0.3 cents at current prices. Its the sort of app that really needs to stick to one chain to save confusion and complexity, I think.


The problem is Polygon's general unreliability and lack of end users. Almost no one I know ever has any cash on Polygon. It's either Solana, ETH, or something like Arbitrum


That's interesting. Sites like DappRadar show far more Polygon apps than Solana apps (https://dappradar.com/rankings/chains, https://www.alchemy.com/dapps). I don't know how representative that is though.

I've noticed occasional issues with Polygon, which is frustrating, and some intermittent viem/Metamask errors that I only seem to get with Polygon. I've put it down to the immaturity of the web3 stack for now but it needs to get better.


What happens if Github doesn't like the content or gets leaned on by a government agency?


Not OP but generally in this kind of system GitHub is just an convenient example. The payload is pointed to by on-chain state and that pointer can be both multi-valued and accommodate multiple protocols. So you can for example point to a self-hosted Gitea or similar, IPFS CID and so on, besides the LCD of Github.


Where are the actual content payloads stored?


Currently, on GitHub but could be anywhere.

Need to balance the freedom to publish anonymously with the risk of people claiming credit for other people’s work.

Perhaps the platform should allow any url to be published and restrict tipping to only authenticated users’ content.


Do you have a link to the contracts?



[flagged]


A plain URL could be one of the future publishing options. Hopefully a platform like Seedling could add value through discovery and content monetisation.


wow, that's cool




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: