poll for the percent based transfer fee

[kenCode]

We really should be going this route with the %-based fees worker proposal. @abit, are you able to do this?

1. Nothing is impossible.
2. All my implementation is based on BSIP (feature definition). If you want me to change the behavior of my implementation, you need to change the BSIP first.
3. Your proposal will increase complexity, thus will increase time and cost for implementing/testing, and probably increase possibilities of introducing bugs into the system.
4. Manually adjusting the parameters on a monthly basis is acceptable for me.

1.
2. @tbone can you knock that out for us with abit's guidance?
3. You stated $10K before, so will this raise your price to 12K or even 14K??
4. Not me. I'd rather let the code handle these intricacies. Don't sign yourself up for that job if you don't have to. I could see that becoming a real pain in yer arse after awhile. Just let the code manage it.

[CoinHoarder]

The fees should remain as low as possible until the DEX has some resemblance of liquidity. Once there is sufficient liquidity, we can revisit the fee structure.

[abit]

We really should be going this route with the %-based fees worker proposal. @abit, are you able to do this?

1. Nothing is impossible.
2. All my implementation is based on BSIP (feature definition). If you want me to change the behavior of my implementation, you need to change the BSIP first.
3. Your proposal will increase complexity, thus will increase time and cost for implementing/testing, and probably increase possibilities of introducing bugs into the system.
4. Manually adjusting the parameters on a monthly basis is acceptable for me.

[tbone]

@kenCode and @clayop are exactly right, the minimum fee should be used only to prevent spam.  In addition to macro-transactions and messaging, that may enable many other use cases. 

As for the % fee, which generates the income, it seems most here think it should be in the .05-.2% range.  Let's keep in mind that any % in this range that works for average to large size transactions, also works fantastically for micro-transactions.  So let's just concentrate on setting a % fee that generates enough revenue without discouraging general use.

Regarding the cap, someone mentioned that Yunbi charges the equivalent of  $.40 at current BTC price (i.e. more if BTC rises) and no one seems to care.  Keep in mind that $.40 translates to a HUGE % for smaller transactions, whereas with this proposal the % fee is flat and only larger transactions would pay the maximum fee.  So I think it's safe to set the maximum above $.40, especially if it's a STABLE value (i.e. won't go up when BTS goes up). 

Speaking of using stable values for the fee limits, I will continue to contend that this is necessary.  Although @kenCode, to accomplish this I really don't think it's necessary to rely on anything more than the price feeds.  So for example, let's say we go with a minimum fee of $.005.  Based on the price feeds, that minimum fee can be calculated in terms of BTS, CNY or anything else.  Those calculations can be performed as often as price feeds are published, although I think daily would be more than sufficient.  So now you have the minimum fee in terms of any asset you need, ready to be applied to any transaction.  Same goes for the maximum fee ($.50-ish?).

We really should be going this route with the %-based fees worker proposal. @abit, are you able to do this?

[kenCode]

Micro-transactions include almost 0 value transactions not for money transfer purpose but for information transfer purpose, such as messaging. Most of them may send 0.0001 BTS with the minimum fee. In terms of this, 1 vs 3 BTS makes very big difference to their costs; 300% increase. Can you imagine how many demands will decrease due to the 300% more cost? The minimum fee is not a source of income but a tool for preventing spam. Currently, the smallest fee is 1 BTS (for account update and call order update, excepting publish feed), and it seems to work in terms of spam prevention.

 
 

explain to me why we're trying to set the minimum fee in terms of BTS?  It doesn't make sense to me since 1 BTS may work for micro-transactions now, but what happens when the BTS price doubles, triples, etc? 

So what I'm saying is, even if fees must be collected in BTS, it seems the minimum fee should be set in USD terms since a) it's stable and b) the minimum spam deterrent was stated (by BM) in terms of USD.  So it seems to me we should set the minimum fee in terms of USD, and then the BTS equivalent can be calculated hourly or daily based on the USD price feed.  Does that not make sense?

 
 
1. Maybe we use statistical data for the calculation, run it every 30 seconds.
 
2. Use a basket of Smartcoins instead of trying to choose the right "one". Similar to BRICS, but use Smartcoins instead, like bitUSD, bitCNY, bitEUR, bitCHF, bitGOLD and bitSILVER. Call the basket something like BTSBASKET and use that for the network fees.
 
3. Our mobile wallets will have real-time encrypted chat too, similar to whatsapp. If you are chatting with somebody, just top off your wallet with some BTSBASKET (or just BTS as a backup), say $2 worth and you're good to go for about a month of chatting. In chat, you might say hello, well if a hello is 5 bytes, that could equate to .00000005 BTS. bye is .00000003 BTS and so on.
 
As @clayop said "The minimum fee is not a source of income but a tool for preventing spam."

[btswolf]

fixed fees.
We already got the tools to define the right fees.
increase fees on demand by the committee (no matter if spam or naturally demand)
decrease fees when demand is low.
This way we will always end up with a balance between spam protection and the the lowest fees.
so let it be 0.001 BTS or whatever the majority votes vor.

BitShares is providing a ledger for all kind of transactions not only an exchange.
No need to charge differently for transactions as they are basically all the same, just a transaction limited in its size.

edit: can`t change my vote to "other"

[jakub]

I don't see how we can justify a max fee other than the standard flat fee (currently 30 BTS).

The two transfer types will coexist, so anyone paying more than the current fee will be cheated when doing so. Unless I missed something about standard transfers being disabled for assets adopting percentage based fees?

@svk

If we set the max to the current 30 BTS, we will effectively lower the transfer fees by a significant amount and thus lose a lot of revenue.
Instead, we should set the min, max & percentage in such a way that in the long run on average the transfer fee income is roughly the same as today's income with flat fees.

So the current 30 BTS should be treated as some type of statistical medium value - this way nobody will feel cheated.
During a longer period (e.g. a month) on some transfers (the bigger ones) you'll pay more and on other (the smaller ones) you'll pay less so on average you should feel that it's a fair deal for you.

Sure, but if the option remains to pay 30 BTS instead of 300, why would anyone choose the latter?

The transfer fee mode (flat or percentage) will be decided by the issuer for all transfers on their asset, not by the user making a particular transfer.
So it's up to the issuer to decide what makes more sense for their asset.

[svk]

I don't see how we can justify a max fee other than the standard flat fee (currently 30 BTS).

The two transfer types will coexist, so anyone paying more than the current fee will be cheated when doing so. Unless I missed something about standard transfers being disabled for assets adopting percentage based fees?

@svk

If we set the max to the current 30 BTS, we will effectively lower the transfer fees by a significant amount and thus lose a lot of revenue.
Instead, we should set the min, max & percentage in such a way that in the long run on average the transfer fee income is roughly the same as today's income with flat fees.

So the current 30 BTS should be treated as some type of statistical medium value - this way nobody will feel cheated.
During a longer period (e.g. a month) on some transfers (the bigger ones) you'll pay more and on other (the smaller ones) you'll pay less so on average you should feel that it's a fair deal for you.

Sure, but if the option remains to pay 30 BTS instead of 300, why would anyone choose the latter?

[jakub]

but the referral program destroy all these, now new users need either pay high fee or pay the LTM fee, in my view paying some network fee is OK, users can understand and accept, but why need to pay so much to the so called referrers?  for months the daily transfer volume kept under 200, nothing can prove that the referral program will succeed.

I'll repeat what I've said many times before: how do you expect the referral program to work if there is no product to sell?
Until the fiat gateways work smoothly and there is liquidity on the smart-coin markets, there is no way I could go to an online merchant and sell them BitShares as a payment system.

1BTS fee do not necessarily link to complexity and high fee gateway and 100 BTS fee does not necessarily link to cheap&easy gateway, and here the point is that gateway always provide  independent service, so gateway should charge independently and the gateway fee should not be added to the transfer fee.

1 BTS fee does link to complexity and high fee gateway because the gateway needs to have a financial incentive to exist.
If you lower the transfer fees to much, you kill the referral program and with it you kill the incentive to run a gateway business.

At least, this is how I understand it - I assume OpenLedger would not go into the trouble of setting up a gateway in Europe (and operate it without any commissions) if there was no revenue from the referral program on the horizon. But maybe @ccedk needs to confirm this.

@bitcrab , I've asked you this question earlier on but got no answer : what is the business model of your CNY gateway, if you do not plan to rely on the referral income?

[jakub]

I don't see how we can justify a max fee other than the standard flat fee (currently 30 BTS).

The two transfer types will coexist, so anyone paying more than the current fee will be cheated when doing so. Unless I missed something about standard transfers being disabled for assets adopting percentage based fees?

@svk

If we set the max to the current 30 BTS, we will effectively lower the transfer fees by a significant amount and thus lose a lot of revenue.
Instead, we should set the min, max & percentage in such a way that in the long run on average the transfer fee income is roughly the same as today's income with flat fees.

So the current 30 BTS should be treated as some type of statistical medium value - this way nobody will feel cheated.
During a longer period (e.g. a month) on some transfers (the bigger ones) you'll pay more and on other (the smaller ones) you'll pay less so on average you should feel that it's a fair deal for you.

[jakub]

@clayop, @jakub: can you please explain to me why we're trying to set the minimum fee in terms of BTS?  It doesn't make sense to me since 1 BTS may work for micro-transactions now, but what happens when the BTS price doubles, triples, etc? 

So what I'm saying is, even if fees must be collected in BTS, it seems the minimum fee should be set in USD terms since a) it's stable and b) the minimum spam deterrent was stated (by BM) in terms of USD.  So it seems to me we should set the minimum fee in terms of USD, and then the BTS equivalent can be calculated hourly or daily based on the USD price feed.  Does that not make sense?

As for the maximum fee, again, I don't see how we can have a meaningful conversation about possible maximum fees set in terms of BTS since it will be a moving target as the price of BTS fluctuates over time, sometimes wildly.  What am I missing here?

The issuer (or the committee, in case of smart-coins) have to take care of the problem by periodically adjusting CER. 
It's not perfect but this is how it currently works anyway so percentage-based fees do not make this problem any worse (or any better).

The think CER does not need to follow the market price very closely, I think it's good enough when its within a 15% range.

[bitcrab]

I think 0.2% is too high for micro-transactions. 0.05% would be better IMO (because given 0.05%, under $6 will be considered as "micro"transaction).
Regarding maximum, it can be debatable. I honestly agree with both sides, and am doubtful to both sides as well. Sometimes people pay relatively high fees in transferring BTC as well. For instance, Yunbi charges 0.001 BTC ($0.4 or 130 BTS) but people seem not care. How much people are willing to pay given 0.05% fee? That's beyond my capacity.

Bitshares is a platform.
Bitshares is a platform.
Bitshares is a platform.
repeated 3 times to emphasize the important thing.

yunbi charge 0.001 BTC for withdraw, it is not comparable to common transfer in Bitshares, it is more like a gateway withdraw service here.

I hope Bitshares can give users most freedom as a platform, can be most user/invester friendly as a platform.

BTS2.0 provide various ways for entrepreneurs to make profits - market fee of privatized smartcoins and UIA, FBA,  one can build his own business with these tools without disturbing other areas. this is wonderful scene.

but the referral program destroy all these, now new users need either pay high fee or pay the LTM fee, in my view paying some network fee is OK, users can understand and accept, but why need to pay so much to the so called referrers?  for months the daily transfer volume kept under 200, nothing can prove that the referral program will succeed.

this is Internet times, this is blockchain times, information passes quickly, seldom people rely on "referrers" to get useful information. and no new users will be happy when he know he not only need to pay to the network, but also need to pay the "referrers".

to me it is ok if the rate fall in 0.05% to 0.2% area, in my mind the max cap make more sense.

some said users which always transfer big does not care 30 BTS fee or 100 BTS fee, yes, some may do not care, but some really care, bitshaes need to attract as more people as possible.

What would you prefer in case of transferring 100k BTS?
- pay 1 BTS for the transfer fee and have all the complexity and high fees on the gateways
- pay 100 BTS for the transfer fee and have cheap & easy gateways

1BTS fee do not necessarily link to complexity and high fee gateway and 100 BTS fee does not necessarily link to cheap&easy gateway, and here the point is that gateway always provide  independent service, so gateway should charge independently and the gateway fee should not be added to the transfer fee.

I have a dream, I dream that one day each user can enjoy the cheap basic fee,  and if they like they can buy add-on service provided by entrepreneurs. and   can entrepreneurs can build their businesses freely via various tools on an attractive platform. 

 

[svk]

I don't see how we can justify a max fee other than the standard flat fee (currently 30 BTS).

The two transfer types will coexist, so anyone paying more than the current fee will be cheated when doing so. Unless I missed something about standard transfers being disabled for assets adopting percentage based fees?

[tbone]

@clayop, @jakub: can you please explain to me why we're trying to set the minimum fee in terms of BTS?  It doesn't make sense to me since 1 BTS may work for micro-transactions now, but what happens when the BTS price doubles, triples, etc? 

So what I'm saying is, even if fees must be collected in BTS, it seems the minimum fee should be set in USD terms since a) it's stable and b) the minimum spam deterrent was stated (by BM) in terms of USD.  So it seems to me we should set the minimum fee in terms of USD, and then the BTS equivalent can be calculated hourly or daily based on the USD price feed.  Does that not make sense?

As for the maximum fee, again, I don't see how we can have a meaningful conversation about possible maximum fees set in terms of BTS since it will be a moving target as the price of BTS fluctuates over time, sometimes wildly.  What am I missing here?

[clayop]

We first have to define the range of micro-transaction IMO.

Under $15? $6? $3?
Depending on the numbers, we can set the percentage. I personally like $15 which in turn has percentage fee of 0.02% (1/5000).
Given 0.02%, we can take max 300 BTS that gives equal fee amount to 0.2%/1/50 (but maximum 30 always gives less).

So I would suggest three options that generate about 300k BTS fee (the original value is about 480k)

1. 0.02% / min. 1 BTS / max. 300 BTS
2. 0.1% / min. 1 BTS / max. 60 BTS
3. 0.2% / min. 1 BTS / max. 50 BTS

If we can sacrifice about more fees, we can choose among

1. 0.01% / 1 / 300
2. 0.02% / 1/ 100
3. 0.2% / 1 / 30

These options generate about 200k BTS fee.

my suggestion: 0.2%/1/20
high friction hurt economy, now it's not time to try to make higher profit, it's time to make higher attraction!

I think 0.2% is too high for micro-transactions. 0.05% would be better IMO (because given 0.05%, under $6 will be considered as "micro"transaction).
Regarding maximum, it can be debatable. I honestly agree with both sides, and am doubtful to both sides as well. Sometimes people pay relatively high fees in transferring BTC as well. For instance, Yunbi charges 0.001 BTC ($0.4 or 130 BTS) but people seem not care. How much people are willing to pay given 0.05% fee? That's beyond my capacity.