Topic: Technology AND marketing of decentralized crypto (Read 15890 times)

---

---

My discussion has been moved for the time being to a Synereo thread:

https://bitcointalksearch.org/topic/synereo-earn-money-using-social-media-1344997

http://techcrunch.com/2016/01/24/guy-kawasaki-on-startups-entrepreneurship-and-the-state-of-social-media/

I think I have perhaps found an economic opportunity for microtransactions that demands the control be decentralized:

https://developers.soundcloud.com/blog/limits#comment-2086111174
http://techcrunch.com/2016/01/24/why-soundcloud-will-be-worth-more-than-spotify/
https://on.soundcloud.com/premier

This is an example of research that can possibly drive a marketing strategy.

This looks to be the precise sort opportunity I was looking for, but now I need to do a lot of deep research and analysis.

The user submitting the files to a decentralized file system could create multiple copies of the same file, each encrypted with a different private key and then only one storage provider would be paid for storing each copy. Even if Sybil attacked. the attacker would need to store each variant, because attacker doesn't know the private keys. But this still doesn't resolve the issues entirely because at least two reasons. First, asymmetric decryption (e.g. public key) is very computationally intensive and normally public key cryptography is only used to interactively exchange a symmetric encryption/decryption key, but in this case the submitting user can't interact with every user that wants to decrypt the file. Second and more importantly, there is no way for a block chain to verify the Proof-of-Storage challenge wasn't replied to. The miner winning the block could simply ignore the reply to the challenge. Only the full nodes which are online at that time can verify the miner ignored the reply (if they've also received the reply), but they can't prove that miner winning the block didn't receive the reply so they can't penalize the miner by ignoring his winning block (for if they could then any one could revert the winning block by producing a late reply and then revealing it). This is analogous to the discussions that monsterer, smooth, and I had about why propagation can't be proved.

It is necessary to quote this longish post from another thread to bring the context into this thread.


Page 14 of the Storj white paper makes it clear they are relying on a propagation full node security model, which we have entirely discredited in public forum discussions I have had with monsterer and smooth over the past month or so. Storj, Maidsafe, and Sia are incompletely specified systems, and frankly Storj, Maidsafe, Sia, and Permacoin look like fantasies trying to sell speculators a bridge to no where.

Before exploring the potential markets for decentralized file storage, I would first like to analyze more deeply the technological viability.

Availability

First if we are not employing ISP user clients as storage providers, thus not stealing upload bandwidth from ISP user accounts (as I had explained in the prior thread where the quote above originates), and thus not able to use the unused storage space of user computers (i.e. we are using then decentralized/duplicated hosted server storage providers), then any redundancy of data storage will incur the increased cost of duplicated storage. However, note that duplicated storage would be required for a centralized storage provider as well (since it would require redundancy for durability of data) and with decentralized copies of data, each provider need not build in internal redundancy as it can require the data (after loss, e.g. due to storage drive failure) from the other providers in the decentralized system. Also with bandwidth costing roughly $1 - $2 per TB (e.g. hivelocity.net) and data center HDDs roughly $0.03 per GB ($30 per TB), a file accessed more than say 100 times is going to be dominated by bandwidth (and CPU) costs. Btw, this is why Google has created a pseudo-offline storage class for infrequently accessed files which reduces costs to $0.01 per GB probably by using recently invented higher latency drive technologies.

Durability

The goal is to insure a copy of the data is always stored and thus the data is never lost. A centralized storage provider maintains redundancy of storage to provide very low probability of loss, and might even back up files in cold (tape or optical disk) storage.

Storj, Maidsafe, Sia, and Permacoin are proposing to use Proof-of-storage to insure there are multiple copies of data stored, but I explained already that this is fundamentally flawed and can't be sound:


For example, the storage nodes (providers) can be Sybil attacked such that all the attackers nodes share the same data store. Certainly this will be the outcome since the attacker can earn more coins for each Sybil node added to the system. IP addresses aren't that expensive. This is yet another reason I think Storj, Maidsafe, Sia, and Permacoin are bullshit and probably scams.

So now that I've explained why paying each node (for the copy of the data it holds) is not immune to Sybil attacks and thus not economically viable, I assert that the only ways to insure durability in the decentralized provider context are either of:

• keep a backup copy in a trusted centralized provider (this can be encrypted if not providing public access so that the centralized provider can't censor based on data content)
• trust a statistical incentive that at least one decentralized provider will be incentivized to store your data

In the first case, the decentralized copies serve the high availability function and the backup copy the durability guarantee.

The second case would be perhaps the incentive providers have to retain the file if it is frequently accessed (assuming they are being paid each time they serve a file) and then setting up a cron job to frequently request and pay for a serving of your files when they have not been otherwise served to the public interim. In this case the decentralized copies serve both the high availability function and the backup copy the durability guarantee.

Depending on the market use case, one of the other might be more economic.

Legality

With hosted content, the host provider holds the person on the account responsible for violating the Terms of Service which includes hosting illegal content. Hosts will likely ban a person for habitually violating the Terms of Service and for not having a streamlined policy for illegal content complaints and legal actions.

In the decentralized providers context there is suppose to be no centralized party to blame or send a legal action to. Storage providers can't know who to believe about whether a file must be removed. The legal authorities can't go serve action on all the decentralized providers, because the storage providers come and go (and are too numerous and distributed). So the only action that can be taken is to require the hosting providers do not allow the system to be run from their servers. Thus the decentralized system becomes essentially banned every where.

The only solution I can think of is to require each file submitted to the system to be signed by public key authorized by a trusted authority that verifies the identity of each signature authority it authorizes. Then the authorities can serve the legal action directly to the trusted authority to revoke the public keys (and all decentralized storage providers comply by removing those files).

So in conclusion, decentralized file systems could potentially provide some higher availability and scaling, but durability is more problemmatic (but probably can be acceptable) and authority over which files can be stored will remain centralized.

The higher availability and scaling is interesting. Not sure if it is compelling though. Users will still have to pay file storage, there is no way to reduce the costs.

And since they don't offer any centralized mechanism for legal actions, Storj, MaidSafe, Sia, and Permacoin will end up entirely illegal or run only from ISP user clients in which case they will be slow (low availability and/or high latency), blocked by many ISPs thus not reliable for users, in addition to the other fatal flaw I mentioned above about Sybil attacks.

---

---

Here is the information about the flaw I had found in Bittorrent in 2008:

I am interested in analyzing the marketing and technology required to have significant quantity of actual users. Most of the Top 20 crypto currencies other than Bitcoin don't have any usership (zilch!), other than the speculators who are trading the coin.

Let's try to stay on topic. If you want to refer to a P&D marketed to speculators as a marketing strategy, then yes that is true but it is doesn't require any discussion. It is obvious that is what most altcoins are. I am interested in discussing real market strategies and respective technology for significant user adoption.

I still however think your point is valid even if we are not referring to the Top 20 altcoins, in that marketing strategy is absolutely essential (otherwise there is no market) and the technology is an implementation detail. Nevetheless technological flaws can render marketing a failure (but often not when the only marketing strategy was a short-term P&D because the technology never gets stress tested as there isn't any usership).

Marketing to speculators is easy. Perhaps I could have pocketed a $million doing it had I been so inclined in 2014. Simply make up a bunch of convincing bullshit and sell an ICO (and sneak in a premine after the fact as well). It would violate my ethics, but others are not so inhibited. Having a well respected forum member as one of the developers can also help the success. But this is getting somewhat more difficult now, because some speculators are becoming more astute and some of the mania from 2013 has worn off.

I would even say, that unfortunately marketing makes a part of around 70% and tech just 30% if I check out the Top 20 crypto currencies. There is a lot of hype done by media and forum attention and the tech is often just the second thing.

That is not an entirely invalid point, but I think we can also discuss marketing strategies that not just in terms of gender or ethnic demographics. We can also consider genres of applications.

Maybe you should market your coin to women?  Or Arabs? 

These are the 2 untapped markets of crypto.  Women and Arabs.

I agree technology and marketing are both required; and moreover the marketing research needs to be done at the same time as the technological design, because they impact each other (unless one is simply doing academic research with no requirement to necessarily tap a market). Note I am referring to more than just promotion. I am referring to the strategy about who will use the technology, why, and relative advantages and disadvantages compared to centralized options for the same markets. So I want to continue the analysis of various potential markets for decentralized crypto and the technologies required. And to analyze if those technologies have flaws. I will continue the direction I was on in my next post. And if anyone else raises any other prospective markets to discuss or investigate, then I may research those as well. Others should feel free to also discuss their research/analysis on these matters on any markets (and respective technologies) applicable to decentralized crypto.

There are some great products which doesn't reached the market due to its failed marketing. Its all really about marketing and in fact companies just want to use the word "SUPER" even when their products aren't really super just because of marketing. but because of marketing, it worked and sold millions.

technology may be great but may also be useless without people using it.

Note this discussion began in the "Technology vs. Marketing. Which is more important?" thread. Please read that thread first before read this one. At least up until my last post in that thread as follows:


I am starting a new thread because I want to discuss more in depth the technology and marketing of decentralized crypto, because I think they are inseparable and not one more important than the other. And so I want to continue the analysis of which markets and technologies are valid for decentralized crypto.

This thread is self-moderated because I will delete any off-topic, time wasting trolling at my sole discretion. And that includes thus who attempt to obscure trolling in nonsense posts that don't many any sense whatsoever (which includes replying to a troll post). I won't waste time making any explanations as to why I deleted a post. Nevertheless, no one who is sincerely on topic will be censored, even if I disagree with their opinion or appraisal of the facts. I have been very tolerant in my other self-moderated thread, allowing endless trolling. This thread I will not allow any personal attacks on anyone and that applies to myself as well. Stay on topic please. If we reach a situation where a poster continues to repeat a myopia that I think should have been understood already, then after many attempts at trying to clear up a difference of viewpoint, I will ask that we stop discussing that viewpoint and that poster can create another thread and link to it. Just because we don't need to go on and on and on and on, as that won't help anyone understand. But normally we can resolve such multi-page (long-winded) misunderstandings such as this example.

I will also delete posts that quote a very long prior post. Please quote the relevant snippet of a long post instead.