Zero-fee = Fee-free = Feeless = Bytemaster's original idea -> Not suitable name for abit's implementation.

Balance based fees -> Also misleading, because fees are not going to change based on balance. Fees will be same for everybody, but depending on how much your balance is, you have credits that you can use to pay the fees.

Come on people, this shouldn't be so hard. Only scammers and pump'n'dumpers will market their products with vague and misleading names and descriptions.

This is what happens if we choose misleading name for feature:
"Oh, cool, you have zerofee transactions! That's best thing ever! I'm going to start using Bitshares and tell all my friends about it!"
"Uhh... Wait a minute... There are lots of fees here... This is not a zerofee system..."
"So you lied to me! Bitshares must be a scam project. Fuck you guys, I'm going to tell everybody that Bitshares sucks and it should be avoided like a plague."

Never ever lie to your customers. It will backfire every time.

just call it what it is:  free transactions

if you need more explaining call it:  Free transactions for shareholders

You can call it what you like but people will still call it free transactions so we should to.

Yeah, I like this too.

Samupaha makes a good point from the newbie perspective: their transactions are not free unless they acquire a sufficient stake. What amount of stake they need to acquire to do X is not necessarily obvious. To enable micro-transactions for example for apps like IoT, how much stake is necessary for what period of time?

I still like abit's approach, I think it has some advantages over BM's, primarily that it achieves most of the desired affect with a huge reduction of effort to implement, as well as preserving existing fee structures for those that may rely on them. I do see it more as a "fee subsidy pool based on stake" than rate limitation. There is nothing that limits the transaction rate, not a thing. It's just a matter of whether the cost of those transactions will be higher or lower, which is based on transaction rate and stake size.

Abit's implementation is great because it doesn't change the functionality of current Bitshares model that much. We can implement it without any troubles. It's really fine compromise between our current system and Bytemaster's idea of feeless blockchain.

But the idea of feeless blockchain is still pretty damn attractive. Graphene-based feeless blockchain would be just perfect for IoT-projects. Truely state of the art what comes to performance. It would be very valuable as a sidechain for Bitshares (as I wrote on the Possible sidechains projects thread).

More I think about rate-limited blockchains, more I like them. We could really create an awesome ecosystem with them.

The gambling chain was my first idea. But anything else? Everyone, put your thinking hats on!

IoT projects maybe. They need microtransactions, and if they can make them without paying any fees at all, that's pure win. Fast and free transactions are exactly what IoT needs. We can offer that for them.

Bytemaster's idea of mutual aid societies might be also suitable for it's own blockchain, although I'm not sure if it will need a rate-limited one.

But then there is the essential question: "What is the role of Bitshares blockchain in this ecosystem?"

Answer: Exchange, bank and gateway. In the other words, the financial platform. Other chains can focus on doing their job, whatever that is they want to achieve, and leave handling of finances for Bitshares. Investing, fiat-gateway, crowfunding, exchange for traders, etc. They can also enjoy positive network effects by hiring our witnesses to be witnesses for their sidechains.

@bytemaster talked also in the latest hangout about this article: Why your Ethereum project will most likely fail

...but he remembered it a little bit wrong: he talked about "why ethereum will fail" but that wasn't the point. The point was that ethereum dapps will fail. Most projects are not going to be valuable enough when compared to their centralized equivalents. People might try them but in the end they will choose centralized one because it will just work better overall.

Ethereum will probably do just fine. They are the one who is selling shovels for golddiggers. There will be always ideas what to build on Ethereum platform, so it's not going to fail anytime soon. There is huge demand.

We could see Bitshares like that, too. We will be the financial platform that servers all the sidechain projects. Many of them will fail because they can't attract enough customers because they just don't have a good enough product. But there will be lots of them and Bitshares can make shitload of money by serving them while they exists. Preferably they all should be Graphene-based, so that sidechaining is easy to implement and it will also make money for Cryptonomex.

Offtopic: Thom, Oren Klaff's book Pitch Anything is great stuff.

Really interesting stuff!

Economics of Blockchain

Sinclair Davidson
RMIT University - School of Economics, Finance and Marketing

Primavera De Filippi
Université Paris II - Panthéon-Assas

Jason Potts
RMIT University

March 8, 2016

Abstract:     
Claims blockchain is more than just ICT innovation, but facilitates new types of economic organization and governance. Suggests two approaches to economics of blockchain: innovation-centred and governance-centred. Argues that the governance approach — based in new institutional economics and public choice economics — is most promising, because it models blockchain as a new technology for creating spontaneous organizations, i.e. new types of economies. Illustrates this with a case study of the Ethereum-based infrastructure protocol and platform Backfeed.

Number of Pages in PDF File: 23

In the latest hangout Bytemaster talked about what kind of sidechains would be worth of doing. Those have to be radically different from any other blockchain that exists today, little tweaks like parameter changes aren't big enough foundation for a new chain.

I don't generally support gambling but if people are going to do it anyway in the blockchain, they should do it in the Bitshares or in a Bitshares' sidechain. Gambling sidechain could be a good alternative – if it's a sidechain, there is no regulatory risk for Bitshares (or at least it's considerably smaller).

Gambling chain could be a good opportunity to test rate limited free transactions in a way that Bytemaster discussed in his blogpost How to build a Decentralized Application without Fees.

I've heard that many gambling business operators have stayed out of blockchain because they need fast and reliable transactions (Bitcoin blockchain is out of question). If the sidechain is Graphene-based, it will be suitable for their needs. If it's truely rate-limited for transactions, operators can be also sure that they have a certain quota of capacity for their games, based on how much core asset they own. If they need more capacity, they will buy more core asset.

Of course the gambling chain would make a good sidechain for Bitshares. That way users will have several different ways to acquire assets that can be used for games, including fiat-gateways. Gambling chain operators don't have to worry how their customers will get funds in and out, they can just focus on building more games and optimizing the blockchain. It would be win-win situation for both, Bitshares and gambling chain.

Does anybody else have any good ideas for possible sidechain projects?

Ok, maybe I've understood something wrong. Can you explain what exactly is rate-limiting in abit's implementation? I'm guessing that's the way we are going, it has already gained so much support.

In Bytemaster's original post the rate limitation is a way of preventing spam. Users can make transactions depending how much they own the core token. When they have used their quota, that's it. Transactions are limited for them until they have waited for some time, then they can make transactions again.

In abit's implementation we are still using fees to prevent spam, not rate-limitations. Users can make transactions as much as they like, they have to pay either with feecredits or with BTS or asset they are transacting.

Talking abour @abit's implementation, if shareholders earn a tradable token for fees .. we could literally establish a 'fee-market' .. people can buy fees or become a BTS shareholder

I don't think that feecredits are meant to be tradable. That would change the system in a very different way and I don't see any immediate advantage with that. It just makes things more complicated than they need to be.

In Bytemaster's original blog post the rate limiting of transactions was important piece of the system, but abit's implementation is more like "subsidized transactions" kind of system, where users are given credits which they can use to pay for transactions.

While the original idea and the abit's implementation accomplish pretty much the same outcome, mechanics are still very different.

After the credits are used, there are no limits for use of Bitshares – users just have to pay normal fees for transactions just like they do now. Talking about "rate limitations" will be very confusing for newbies because in reality there is nothing limiting in abit's implementation.

From a marketing perspective the name "rate limited free transactions" is really bad because it's so confusing. We need to come up with a better name.

Imagine that you are total newbie, you don't know anything about Bitshares. You go to the website. What there should be written so that you understand immediately what this feature is about and how it works?