Holochain Forum

Cardano: A Holochain Perspective

This has been a good year so far for Cardano. It successfully forked to become a multi-asset blockchain, enabling users to create tokens that run on Cardano natively, similar to how tokens operate on Ethereum. (Full smart-contract functionality remains on their roadmap.) Its token, ADA, became tradeable on Coinbase in mid-March. And investors lined up: the price of ADA rose 560% over Q1, the most among the CoinDesk 20, a core group of cryptocurrencies tracked by the news site. Then in April, Cardano announced a major blockchain-identity initiative in Africa, beginning with a partnership with the Ethiopian government to improve the quality of education.


Read Article: https://blog.holochain.org/cardano-a-holochain-perspective/
1 Like

I believe this article is making some unrealistic claims about holochain which weakens the article’s credibility.
There is no such thing as “instant” latency and its quite incompatible with the notion of eventual consistency. “Seconds” would have been more realistic and would have still made the case against Cardano.
I also don’t see how holochain could ever achieve live video streaming. If it ever did, it would be because it would heavely rely on actual streaming protocols like RTSP.
Handling of large-media applications is also quite questionable as you can’t rely on holochain alone to do this.

6 Likes

Since Damien chimed in I think I will too. Personally I don’t really like this article. The other major issue I see with it is that it is dishonest/misleading in it’s criticism about Cardano. To be clear, I am not particularly interested in Cardano, nor do I hold ADA, so I have no stake in this. But my understanding is that transactions are relatively quick on Cardano, and reports of slow transactions are the fault of centralized exchanges, not the Cardano design.

Personally I think this article either needs major corrections, or should be retracted, and perhaps an acknowledgement/apology or explanation of some kind. Because it seems problematic and looks bad to me.

Edit to add: Consider the reactions you see in this post on Reddit

5 Likes

This article is nothing but BS. How long has Holochain been in existence and where is it now? By the time Holochain can do all the things they said they can do, it is too late. It is easy to put in words what the capabilities are and how it is better but to get it done… you need to compare what is working vs what is not, not imaginary capabilities.

1 Like

Ugh, apples and oranges.

Hey everyone – Thanks for the responses to my article. It took a while for this thread to work its way over to me, but here I am… and I’ll make sure to monitor the Forum more closely in the future.

Responding here to the a few of the comments above…

First to @brooks:

I have done my best to present a fair analysis; it’s not my intention to “badmouth” another project at all, let alone unfairly. At the same time, I do believe it’s important that people have the opportunity to understand how Holochain is different from blockchain projects. I don’t believe it’s possible to educate about Holochain thoroughly without pointing to some of the limitations of blockchain projects. This is why we are embarking on this #AHolochainPerspective series, which is scheduled to continue for at least another few articles.

That being said, I do think an edit is warranted in the article in order to clarify the matter about transaction wait times on Cardano. It’s true that “users have recently reported transaction wait times on Cardano ranging from 1 to 18 hours”, as the discussion behind the link shows. But it’s also true that there may have been reasons for these wait times unrelated to the network itself. Part of the process of writing this article was to integrate input from various Holochain team members. One of my takeaways from receiving some critical feedback on this piece is that I need to more thoroughly consider the suggestions I’m receiving before finalizing their inclusion, or at least their languaging, in the article.

The only other specific critique I see in the Reddit thread you linked to is about use of the word “substantial” to describe Cardano’s gas fees. As far as I know it’s unclear exactly what Cardano’s gas fees will be once their network is up and running at scale. “Substantial” was a word designed to cover a broad range of possible costs, basically anything that adds up to a meaningful expense if anyone is doing a lot of transactions. One user on the Reddit thread said that 25 cents is not substantial, but if that kind of fee applied to every post, comment, like, share, etc in a social media platform, that absolutely would be substantial. Even for financial transacting that amount sounds quite substantial to me, especially if we want crypto to be versatile enough to use for small payments around the world.

By the way, I actually do like Cardano, and I hold ADA. I believe it’s a better blockchain protocol in a lot of ways. And I also think it’s not a stretch to say that the potential of blockchain, while substantial, is basically limited to low-throughput applications like financial services, with live-collaboration and media-intensive applications likely to remain always out of reach (unless decentralization is substantially compromised). Holochain does not have such limitations, or at least is far less limited by orders of magnitudes. That’s a story that needs to be told… especially because, to the unfamiliar, even the word ‘Holochain’ sounds like another blockchain project.

I am curious if there’s anything else about the article you “don’t like”. If it’s just the fact that we’re pointing to the limitations of another project, I don’t really see us changing that approach any time soon. But if there’s something further about the way I’ve done so, I’m interested to know more.

Now to @ddd-mtl:

Re “instant” latency: we considered a number of word choices here, and I do stand by this one as a fitting word. That’s because you actually do write to your own source chain immediately, and that’s actually what matters because your source chain is the definitive source of information on the matter. There is nothing else that happens, really, that could be considered something you are “waiting” for before transacting again based on your new account balance. The transaction does not need to be validated by the network. As soon as you write your transaction to your source chain, your balance is updated such that someone else may transact with you (by verifying your account balance themselves).

Re live video streaming: this is something I’ve spoken about with @artbrock on numerous occasions, including in my research for this article. He reports having seen live video streaming with latency of around one minute, which is suitable, and that was on Holochain’s prototype, which was thousands of times slower than the current version. (Note that we do not say “instant” here… it’s reasonable to expect a bit of latency in video streaming, in both centralized and decentralized environments.)

Re large media applications: again, my understanding is that, actually, you can. I’ve spoken with Art extensively about this as well. What suggests to you that “you cannot rely on Holochain alone” for large-media applications?

In conclusion:

It’s essential to me that Holochain communicate with rigorous integrity to the absolute best of its ability, and it’s doubly essential to me that I do. I appreciate the checks and balances showing up here in the community, and it’s causing me to look at how I can improve my processes. At the same time, drawing a comparison between Holochain and blockchain projects – including pointing out the limitations of the latter – does not itself constitute unrigorous communication, in my view.

10 Likes

Hi @joshmzemel

I’ll just clarify on the last point concerning large-media apps.
Art’s latest medium post said one year of Twitter would required 20 GB of storage space so imagine how much one year of Youtube would require. It would make more sense to offload the file storage to a dedicated solution (S3, IPFS, bittorrent…) and bridge with it. This is what I meant by not relying on Holochain alone.
I believe I’m not stating anything new here.
Back in the Holochain proto days, there was an explicit list of things not suitable to do on Holochain. As you can see “large files” is on that list, and I don’t see why it shouldn’t be since the reasoning was based on Holochain’s architecture which hasn’t changed since.

1 Like

@ddd-mtl What about “chunking” large 720p video files (as a Holochain entry) into many 1-minute long 720p video files (many tiny Holochain entries)? Yeah, forget adaptive bitrates and runtime video-transcoding (basically all the things a media server does); they’d have a very hard time modeling a media server as a Holochain role that an agent with a powerful machine might take (though theoretically even this should be possible). However, even forgetting real-time video transcoding, seeing that high speed 5G connection speeds should soon be the norm everywhere, I think simply storing “normal” high-res video files in small chunks as entries with enough redundancy so as to allow millions of leechers to simultaneously watch a video by ‘get’ting those little 1-minute chunks every 1 minute, saving them as video files, and simultaneously deleting them from their device after watching, should be enough to make HoloTube a possibility… Also, nothing should stop people from running Holochain nodes on huge data centers (or on the cloud); a HoloTube with a total capacity of an exabyte is not that far-fetched! The only thing Holochain fundamentally lacks so as to be able to realize such a scenario is an out-of-the-box incentive system, as in: why should I store so much of others’ data when I myself produce negligible data from my side… [Kinda like a negative externality…] Ethereum for example has gas fees, and Ether as their compensation money. But I personally think it is another of Holochain’s plus-point, in that they provide the bare minimum with which to realize almost any use-case (apart from Machine Learning, AI, recommendation systems, live video-calls, etc, which do warrant being offloaded to other systems). All HoloTube will need apart from the bare minimums that Holochain framework provides is a financial layer on top of it so as to be able to compensate those who actually hold and host those video chunks. Instead of the naive ‘get’ requests to fetch the video-chunk, a leecher would first have to get the hashes of the chunks that make up the video he wants to watch, then do another ‘get’ that retrieves him a list of nodes who claim to be in possession of the video-chunk along with the monetary amount that they’d charge him for the chunk’s delivery, then do a remote-call with the cheapest provider to actually fetch the video-chunk, only to finally pay them the said amount for the chunk’s successful delivery. Or perhaps you pay the provider first, then get the opportunity to remote-call with them (taking along with you the proof of payment); the providers being players of repeat dealings, thus it being in their best interest to play nicely to earn positive reviews (which a leecher should be able to leave if they have a proof of payment; thus, if a provider takes my money then goes offline, I should be able to downvote it with my proof-of-payment as my validation package)… So yeah, without advertisements, HoloTube probably wouldn’t be free (unless you yourself host video-chunks as well, thus canceling your expense incurred by binge-watching with your hosting income), but it definitely can be a reality, all without ever having to resort to Ethereum Swarm, AMZ S3, LivePeer, commercial CDNs, etc…

TLDR: Holochain Yolochain; haha.

1 Like

Perhaps @artbrock might like to weigh in here?

1 Like

Hey Damien ( @ddd-mtl ),

Yeah… When I wrote those guidelines in 2017, I was trying to discourage people from thinking of Holochain as just file storage or global token state management, and trying to get them to think in terms of data driven apps.

But you can certainly use Holochain for file storage as long as people know that’s what they’re getting into. If’ I’m gonna run the media-server DNA for a YouTube like apps, then I know I’m gonna be storing and streaming a lot of data. You’ll have to break the videos into no larger than 16GB chunks because of the size limitations on entries (which happens to match size on websocket transfers without additional stitching).

You’ll not cryptocurrencies are on that list and we’re also rolling out HoloFuel… just not as a token-centric currency implementation, but as agent-centric accounting.

-art

2 Likes

Hey @artbrock :slight_smile: I think you meant 16MB instead of 16GB, just pointing it out because this thread could be visited by lots of people.

2 Likes