Hacker News new | comments | ask | show | jobs | submit login
Akasha: A Social Media Network Powered by Ethereum and IPFS (akasha.world)
120 points by alexandrerond on May 12, 2016 | hide | past | web | favorite | 48 comments

A service like this already exists - zapchain.com (bitcoin instead of ether). I have earned billions of satoshi on it but well, it is failing. Because identity could not be solved with partial blockchain, same mistake Akasha is doing.

An identity is a proxy for past work done, which the identifier further wishes to use as a proxy for future working. A blockchain bypasses this by directly asking a proof of current work, irrelevant of past work done.

When a system does not use an exclusive blockchain, the identity is unable to proxy itself through PoW. Spam rules zapchain now.

To make this work they need a PoW Akashacoin.

We are aware of zapchain and their issues. I'd like to point that zapchain is not completely decentralized - it is rather a hybrid application using servers for content distribution and blockchain for payments.

Regarding security/spam - Proof of Work is not the only way you can secure networks; one of the interesting potential solutions we explored is a sort of Proof of Stake mechanism.

Basically this would allow a sort of "deposit" scheme where you would deposit an X amount of tokens for an Y amount of time and depending on the feedback received such as votes for example you would either get the deposit back or lose it. That would make it expensive to spam while not affecting the people generating good content as they would get back their tokens.

Things get a bit tricky when we're talking about controversial topics and how that would be treated but that's a different thing.

That being said, this is an open and complex discussion that probably will take a while to figure out.

In the meantime we're happy with proving that it is possible to have a completely decentralized publishing system that doesn't need any servers, creating in the process a product that empowers people with unfettered freedom of expression. One step at a time :)

Ooh, good solution and good problem to point out on controversial stuff! Why not, while allowing upvoting, also allow independent assessment of risk/controversy/value, and these somehow weight things to reward such gambits, as long as a substantial minority of users vote them up or something? Maybe a sliding percentage, so that if even people who downvote it mark it as controversial or whatnot, the more people thing it's controversial/provocative, the smaller the percentage who have to agree with it/up vote it? Just a quick reaction, I'm sure with some iteration and/or thought you could develop a better similar (or different) idea with some models to support it as optimal, or pareto optimal given some constraints.

Firstly, I really want such project to exist and will start working on it next year if nothing successful exists by then.

I am not advocating PoW for security (yes sure PoS & other federated schemes can do it) but for spam. Your solution is a pre-deposit. Nothing is impossible but good luck having creators to pay to blog. [Try brainstorming a PoW akashacoin & you will see the secret sauce that solves this pre-deposit problem]

Being decentralized certainly makes the geek in us happy but is only relevant to user if it brings some advantage, here it being censorship-resistant which takes you to the controversial topic problem.

Good luck with this, I'm really excited to give it a go.

You say that Akasha "doesn't need any servers". If a user is to run a "node" on their PC, how will the network react if a friend/follower is offline when data is requested from them. Also, will a node be runnable on a home server system, allowing for higher-availability and easier backups of personal data?

First of all thanks for the encouraging words!

> If a user is to run a "node" on their PC, how will the network react if a friend/follower is offline when data is requested from them.

There are some things currently in development such as IPFS clusters that can tackle this issue. There's also the IPFS filecoin token (not yet live) which would provide an incentive for people to keep the hash "alive". This is similar to how Ethereum's Swarm will work.

In the meantime the gateways mentioned in a comment bellow would act as a sort of "seed" even if the OP goes offline.

> will a node be runnable on a home server system, allowing for higher-availability and easier backups of personal data?

I don't see why not. Remains to be tested though :)

@mr_stud the functionality you are asking about is addressed by IPFS. I worked on the project briefly about a year ago -- exciting stuff. IPFS's content distribution is similar to that of a torrent. So the content gets stored across the network, with certain concentrations where people can pin it. It leaves room for giving an economic incentive for pinning it. Cost of storage shifts over to the people who care about the content rather than the people who care about it's distribution.

There still would be a case where a friend's content is unavailable because it is offline. If the friend is offline too much, one way is to have an intermediary willing to pin the content when that friend comes briefly online. It may be possible if the friend is within a large enough group, there will always be some sort of coverage.

Check out the IPFS website sometime. There is a white paper and demo videos.

IPFS is largely the reason for me battering F5, on this thread, this afternoon.

I've only briefly looked into it, on its website, upon originally reading about it. I will definitely give it more attention, have a deeper read, and have a go at running an instance to see what can be done.

The blockchain is only half of it. These guys are using IPFS for the other half.

They are using blockchain for decentralized payment system and IPFS for decentralized file hosting but they have rejected the hashcash algo which is the basis of invention of blockchain, by not using PoW.

They solve this by providing a paid-membership. So basically, a content creator is paying this fees for availing the network affect of akasha.world instead of publishing it for free on their own blog and putting in an address.

It mystifies me that people seem to think users will suddenly start caring about decentralization and privacy. If either of those things is your social network's USP, you're sunk.

Decentralization is more powerful than you think. If I open a social network that is just like Facebook, it will go under. But if it is also properly decentralized (=federated), meaning you can have account at my site but interact with every Facebook user as if you had an account there, Facebook immediately looses it's network effect. And if I offer an additional silly feature, say let users customize the color of their home page, my decentralized net is at an advantage.

Unfortunately, Facebook explicitly forbids that kind of interaction, of course.

I see two options how decentralized nets could win:

1) Legislation that forces large sites like Facebook to federate (makes sense if you argue that FB has become infrastructure) or without that:

2) Lots of people try to build federated sites, and fail... but one day, Facebook goes the way of Yahoo and MySpace, and one of the decentralized contenders happens to be the new Facebook by chance.

> Legislation that forces large sites like Facebook to federate

This could happen faster than you might think. The Telecommunications Act of 1996 ([1]) just needs some updating to current technological standards.


> Interconnectedness. Since communications services exhibit network effects and positive externalities, new entrants would face barriers to entry if they could not interconnect their networks with those of the incumbent carriers. Thus, another key provision of the 1996 Act sets obligations for incumbent carriers and new entrants to interconnect their networks with one another, imposing additional requirements on the incumbents because they might desire to restrict competitive entry by denying such interconnection or by setting terms, conditions, and rates that could undermine the ability of the new entrants to compete.

[1] https://en.wikipedia.org/wiki/Telecommunications_Act_of_1996

Centralised SNs is probably a boon for NSA etc, so I doubt the US will be eager to make new laws. FB is prism for free.

I'm not so optimistic here. To my knowledge if it hasn't been changed to force interconnectedness of email, I doubt it will be changed only to force interconnectedness of social networks like FB. But i guess we'll see.

Why would it need to be changed to force interconnectedness of email? Email has always been federated.

> Facebook goes the way of Yahoo and MySpace

That is already happening. The majority is there only because of the messaging aspect.

A whole plethora of messaging services are gaining traffic. Next thing that follow in them is "smart AI" for content with "bots" and then we'll reinvent what Asian messengers already offer for years.

But convincing an user of $chat with decentralization and federation is laughable, until :

1) he gets a good reason for current $chat to not be as good,

2) his network flocks away to something else.

Don't want to be pissing on anyone's parade here, but this Akasha Project is currently putting a cart before the horse.

2) "One of the decentralized contenders happens to be the new Facebook by chance" is not going to happen.

Interesting reading material: http://benkler.org/Pub.html .

I agree that not everyone cares about these sort of things. However there are still people out there that do care and now those people will have an alternative.

For those that do not care about the decentralized/freedom of speech/privacy aspects embedded in our app we think they might find it interesting thanks to the way good content and thoughtful commentary can generate "real money" rather than "points".

For those that don't care about the money either and just want a nice user experience without being bombarded with spam and ads, we think that they might find it attractive thanks to the clean and intuitive UX/UI we designed. We don't like ads either and we're set on experimenting with alternative business models made possible by cryptocurrencies in general in the pursuit for a sustainable social media network model that doesn't rely on ads.

Then, there's also the people that actually create and do things such as writers, artists and activists. They might appreciate AKASHA for different reasons as it allows them to connect more directly with their fan base/followers by-passing in the process some of the blocks that come in various forms on Web 2.0 platforms.

Hello everyone,

AKASHA founder here.

Looking forward to hearing your thoughts on our project!

Ambitious project. Good luck :-)

- Will Akasha leverage existing social graphs (fbook, twitter, email, phone)?

- Does Akasha plan to interoperate with any of them?

- Why is a blockchain needed? What aspect of decentralized publishing does this solve? Verifiability of content? Pub/sub relationships?

Many thanks!

> Will Akasha leverage existing social graphs (fbook, twitter, email, phone)

This is actually something we were discussing yesterday with Dr. Gavin Wood, one of the founders of ethereum and ethcore. It is a bit tricky to do without compromising the privacy of the users (associating an insecure form of identification to a key/address) due to the nature of blockchains, but there are a few interesting ideas on how to solve this issue. We’re looking into it but we need to experiment and see which one is the best before deploying something like this inside our application.

> Does Akasha plan to interoperate with any of them?

Social web 2.0 applications such as Twitter and Facebook are not necessary the friendliest when it comes to interoperability but as we will approach the beta we will look into making an API allowing interesting things to be done in an IFTTT fashion.

> Why is a blockchain needed? What aspect of decentralized publishing does this solve? Verifiability of content? Pub/sub relationships?

The chain removes completely servers from the picture and enables us to create a truly decentralized publishing mechanism by adding IPFS to the mix - since it is expensive to store content directly on chain, we're storing IPFS hashes on chain.

Besides verifiability and pub/sub relationships, identity/profile management and micropayments are also a big pluses that come bundled with the ethereum chain.

Hope this answers your questions!

I'm really hopeful that this - or for that matter any reasonable platform - deploys and gets used. I'm a big fan of decentralization and hope these kinds of efforts pan out! Good luck to the Akasha folks!

Thank you for the kind words and support!

Here is a simmilar project in Golang https://github.com/vole/vole

It's a shame that it's using BitTorrent Sync. Use of something open source, with an API, like Syncthing, could allow for a smoother experience in finding users to follow, rather than manually copy-pasting shared secrets.

I have had my eye on https://ind.ie/, who forked Syncthing in order to build a social platform upon.

All I can see is a blank white page — doesn't fill me with confidence.


akasha.world is at the moment just a presentation website hosted via GitHub Pages. We needed javascript to make it pretty :)

If you want to find out more about our project with javascript disabled, our blog should be readable http://blog.akasha.world/

> We needed javascript to make it pretty

You can make a page pretty just with CSS; and even if you add behaviour via JavaScript, you can still degrade gracefully.

I see nothing but a blank page in both lynx and eww.

I would suggest not browsing the Internet with a broken browser, then. (Hint: NoScript.)

Websites that are javascript only are broken.

That's odd. It's not broken for me. Are you sure it's not your browser? Maybe you have a broken javascript interpreter?

A side note, but couldn't help notice ".world". Wonder why they didn't buy akasha.space which is more suitable because akasha means "space" or "sky" in Indian languages. (500 USD to buy the domain name though. May be affordable?)

I was hoping to read something about the protocols. For example, how do they handle access control (e.g. so that only friends can see my status, photos, or so that groups can be managed)? How are updates broadcasted (without polling)?

Hi there!

In the first release/MVP stage AKASHA will resemble a lot to Twitter/Medium in this regard, in the sense that your published entries will be public and visible by everyone.

That being said, access control is on our roadmap but it is not a priority at the moment.

The updates are broadcasted via the Ethereum network through a smart contract that records and dispatches events associated with profiles such as new entries.

This way, the people that follow you on AKASHA will receive in their stream a notification letting them know that a new entry has been published and can be accessed via IPFS at hash X.

Can people join this network without downloading an application - will it be available in the browser? The reason I ask is most decentralized applications up to this point require a download to start your own node (zeronet for example). Availability from the browser I believe is important for decentralized applications because it lowers the barrier to entry for consumers.

PS., not sure if you noticed, but some of your comments on this page are marked dead. I have no idea why this has been done to you.


In the first phase an application is needed as interacting with the Ethereum/IPFS network currently requires a local Ethereum/IPFS node.

IPFS for example is currently working on a javascript implementation that would make it usable from normal browsers and basically "invisible" for average users. Ethereum also has some efforts in this direction but most likely we will see these things coming together somewhere around Q3/Q4 2016.

When these options will be available we will definitely make use of them as we understand the potential adoption gains we can have by removing the need to install an app.

In the meantime we were thinking to setup a few IPFS gateways that would serve a static version of the content in the form of a normal website (http://akasha.world for example).

This way people can at least access the content/read/share it over traditional Web 2.0 mediums. On the other hand, if they want to interact with the network (publish/comment/vote) then they would need to install the app. Not the best solution, we know, but seems like a decent compromise while the tech "gets there".

> PS., not sure if you noticed, but some of your comments on this page are marked dead. I have no idea why this has been done to you.

I have no idea why this was done either...Did I broke any HN posting rule without knowing?

Here is a similar project writedown.co and will be probably transitioned to work with lisk.io. Disclaimer I am the founder

Wasn't clear from the page. What can I use it for? Why does it exist?

>What can I do with AKASHA?

>You can publish, share and vote for entries, similar to Medium and other modern publishing platforms, with the difference that your content is actually published over a decentralized network rather than on our servers. Moreover, the votes are bundled with ETH micro transactions so if your content is good you’ll make ETH from it – in a way, mining with your mind.

if your content is good you’ll make ETH from it

Does it have to be my content? Is there anything to stop me writing a bot that looks for popular Medium/Reddit/HN/whatever articles and post them to AKASHA under my name to harvest ETH?

That's exactly what people will do if they believe there's money to be made.

Nothing, I also guess it is a good chance to make money from cat pictures :)

Sounds good except for only using microtransactions for voting (it is also easy to inflate your own votes).

I'd prefer to see something closer to a federated system for voting / popularity tracking (however unlikely it may be). To detect spam bots and similar manipulation methods you need trusted humans to filter out the bad votes by spotting the trends.

And being able to see the differences in average votes between various voting clusters (different sets of users) would probably be an improvement (through seeing the varying rankings across servers with different userbases).

We are using a quadratic voting system for the microtransaction side with weights between 1 to 10. Basically this enables people to easily support their authors/artists/cause via votes/likes with amounts from 1 cent to 1 dollar.

Quadratic voting is also quite interesting as it allows people to express "how much they like/dislike" a particular thing - which is not possible with the traditional 1 like/vote per 1 person mechanisms.

We are aware of the potential problems that might emerge from this model and this is why we are very interested in exploring various solutions such as Proof of Individuality and reputation systems that can bring both local and global rankings into the mix.

These areas are still in the early days of pioneering but we are determined to experiment as much as possible in the search for the best solution.

Anyone that is interested in this sort of things is more than welcome to join the discussions on our slack at http://akasha-slack.herokuapp.com/

That sounds.....like fun....I'm serious. Very interesting.

Will this use Swarm instead of IPFS once it's out? Also, using contracts costs gas. So how will that be paid for? What's the business model?

We will definitely experiment with Swarm once it's out and see how well it fits with our needs. This document covers some of the differences between swarm & IPFS for those interested: https://github.com/ethersphere/go-ethereum/wiki/IPFS-&-SWARM

For the MVP we will gift a bit of ETH to each user that installs AKASHA and creates a profile. We are talking about enough ETH to enable an average user to publish, comment and vote for a few good days. The "perfect" amount will most likely be determined after observing how real users interact with the application during the alpha. If along the way the user will generate content or comments that will in turn generate upvotes he will refill/increase his ETH balance.

It's important to note that users will be able to read entries and comments without needing any ETH so access to information is free in this regard. ETH is required only when you want to interact with the dapp living on the Ethereum network.

In any case, from my rough calculations, the costs involved in publishing an entry with 1 tag would be recovered after receiving 1 upvote. All the upvotes beyond that point would be profits going to the author.

We are also open to experimenting with things that might enable us to take over some of the interaction costs at the smart contract architecture level, but that is outside the scope of the MVP.

As for the business model we don't like ads in general so we want to experiment as much as possible on this side as well to see if we can co-create a sustainable social media network model that does not rely on ads to survive.

We don't know yet if that model means a custom token, a micro-fees scheme or something completely different that we can't even imagine at this point so we chose focus first on building the dapp and learn from the actual usage what direction we should take.

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