I already mentioned that bandwith should be part of the farming reward (CPR actually, because that is what it boils down to).
Aernoud is correct in stating that calculations remain the same if we put all the dependent factors in the CPR number. It is easier to adapt the calculation of the CPR number later on, than changing the whole calculation.
The only difference between 2B or 4B is the number of existing tokens migrated… Other than that, it doesn’t change anything…
So I would certainly plan to incorporate bandwidth, because it IS important for the functioning of the grid, that is obvious…
Grid usage could be used to slow down farming rewards, but on the other hand, it is not something the farmer can remedy, he can only provide the best possible hardware and bandwidth, but he cannot drag users on to his machines of course…
We just need to make very sure there are no ethically questionable rules or actions because that would quickly end up in situations like ripple, but Ripple is an established project by now and will not stop because of this, but our project would not survive a similar public attack…
And besides, marketing of course, having a simple setup and usage demo, because currently not even I was able to set up the SDK…
ThreeFold Tokenomics Discussion
I think we are largely in agreement then. However, I do think Farmers have a role to play in getting users to the Grid. Nobody can ‘drag’ users to the Grid, but promotion by Farmers can help. They know exactly what they are talking about.
The update of the CPR can wait. Moving 2B or 4B over indeed doesn’t make a technical difference.
So, no reason to not get full steam ahead then. IPv4, On-click SDK Install, Liquid Listing, Marketing…
Indeed, what are we waiting for actually?
If this is the responsibility of Sam I would suggest he kicks in a higher gear here, because I ain’t seen nothing yet!
@gosam
100% agree.
It’s go time with the marketing!
ok guys, I think we have it figured out,
see https://github.com/threefoldfoundation/info_threefold/blob/development/src/grid/farming_logic_python.md
We took as much of your recommendations as I could and came up with this.
Hope you like it.
Please before commenting or getting more into discussions, read the doc well, and please ping me to ask for improvements where you believe we should. We don’t have much time left to make changes.
Anyhow indeed its time to go.
Kristof
Sorry, did you provide the correct link? This leads to a piece of code.
Looking forward to this coming week of marketing!!!
May 4-8, start of marketing campaign
5 Zoom calls which will be recorded and made available, we will announce them everywhere.
Monday 4 May: the launch of our TF Grid 2.0 and short demo
Tuesday 5 May: the values of our TFT & link to simulator
Wednesday 6 May: instructions on how to use 3Bot connect, wallet & the Stellar exchange basically explain how tokens can be bought
Thursday 7 May: TFGrid SDK install & usage
Friday 8 May: detailed demo session how to deploy and use Kubernetes on the TF Grid
LET THE SHOW FINALLY BEGIN!
kristof,
I don’t understand the calculation, you will never get to 0 tokens farmed if you divide by a number, that can never go to 0…
I would suggest changing to this, it also leaves out the starting amount of tokens (which is irrelevant actually)
def difficulty_level_get(self, month):
nr_of_tft_ever_farmed = int(self.simulation.current_tft_total) #look at current amount of TFT in existence
p = nr_of_tft_ever_farmed / 40000000000
if p > 0.999999:
perc = 0
else:
perc = 1-p
return perc
and in the calculation of the tokens farmed, change to:
def tft_farm(self, month):
tft_farmed = self.node_config.cpr * self.farming_cpr_tft(month) * self.difficulty_level_get(month)
(multiply instead of divide)
Also,
The utilization rate is now calculated for the farmer node, but I would take into account the utilization of the complete grid for these calculations (I would do that indeed based on a prior date to not have to calculate that each time), but the idea is that a farmer has no control over the usage of his farm, he can only provide the best possible hardware and connectivity.
If he does that, the CPR rate will be high and he will earn more tokens. it his CPR rate is low, he will earn less, and also be less attractive for customers, but he can set a lower price for his hardware.
The idea for the utilization factor is to prevent farmers to farm lots of tokens at the start of the grid (because nr of existing tokens is low so rewards high) without the grid actually being used…
But again, don’t let these details refrain from shouting out the news of what we do have and let’s open the gates for marketing!!!
Yes… it must be on grid-level. Utilisation on node level is prone to fraud also I believe.
Tokenomics aside this projects success comes down to the technology and marketing.
Time to tell and show the world what’s been cooking for so long. Why the world needs it and how it improves the planet.
Any preview of today’s marketing @gosam
The tokenomics discussion stopped at some point. Was consensus eventually achieve on this topic?
Sorry to ask as I’m sure the info is posted somewhere, but could someone point me to where I can find;
- The FINAL tokenomics, particularly as a legacy farmer as I understand that my old deal is no longer valid.
- How to sell TFTA. I think they can only be sold through a Bot at a price set by TF but I’m not sure how to do that.
Thanks in advance
Hi Colin,
Here is the links where you can read more:
https://wiki.threefold.io/#/tokenomics_updated
https://wiki.threefold.io/#/token_overview
https://wiki.threefold.io/#/token_limitedsupply
https://wiki.threefold.io/#/threefold_marketmaker_bot
https://wiki.threefold.io/#/tfta_vs_tft
Here you can make a projection on your farming income: https://wiki.threefold.io/#/simulator_configure. In short, all the chances were supposed to make economics even more interesting for farmers or at least leave it as good as it was before.
Regarding marketmaker bot its not yet released, we hope to have it within 1 month.
For everyone who is curious and not able to read the provided python code, here are the details of the new tokenomics/minting specification. Please let me know if anything is unclear of if you spot any mistakes.
Certification
Certified farmers receive certain perks that will factor into calculations below. Details on which hardware can be certified and what the process looks like are available in the wiki.
Capability Zone
This metric rewards farmers who bring nodes online in areas that are currently underserved by internet infrastructure. For Africa and South America, this value is 1, while for the rest of the world it is 10.
Bandwidth Check
Nodes need to be connected to adequate bandwidth to be able to utilize their storage and compute resources. This metric penalizes farmers who provide less bandwidth measured in download speed, as follows:
Certified
Bandwidth has been certified, so always 100%
Uncertified, Capability zone 1
100% for 2mb
50% for 1-2mb
0% for less than 1mb
Uncertified, Capability zone 10
100% for 20mb+
50% for 10-20mb
0% for less than 10mb
Utilization
Uncertified nodes are penalized if they do not maintain a certain level of utilization. There is a one year grace period where this check is not applied, which I assume began whenever the new tokenomics took effect in either May or June.
100% for greater than 50% utilization
50% for 25-50% utilization
0% for less than 25% utilization
Uptime
Certified nodes have agreed to 99.8% uptime (1.44 hours downtime max per month), while uncertified nodes are expected to maintain 99% uptime (7.2 hours downtime max per month). Nodes that do not meet the uptime requirement will not farm any tokens.
Difficulty Level
As the token supply grows, the tokens minted in a given month diminish. This both rewards early adopters and also ensures that the supply is limited to 4 billion tokens.
Difficult level = 1 - (number of existing tokens / 4 billion)
So for example, when 2 billion tokens have been minted, the difficulty level will be
1 - (2 billion / 4 billion) = .5
And when 3 billion tokens have been minted
1 - (3 billion / 4 billion) = .25
Once token supply reaches 4 billion (actually just a bit lower by the current logic) no new tokens are farmed. Assuming 800 million TFTs in circulation currently, the difficulty level is:
1 - (800 million / 4 billion) = .8
Cloud Production Rate in TFTs
Cloud production rate is subject to change, and in fact there are two (slightly) different formulas provided, one in the wiki and another in the code used to display the CPR in ThreeBot’s admin dashboard for farmers. The cloud unit calculations are a bit complicated, so I won’t repeat them here.
We assign a price to purchase the hardware corresponding to one CPR of capacity, which is quoted as $40 USD at the beginning of 2020. Then we assume that this hardware should pay for itself over six months. Finally, the value is converted from dollars to TFTs, based on an average TFT price from the last month. Right now, this looks something like:
CPR TFTs = CPR investment cost / 6 / TFT price = $40 / 6 / $.15 = 44.44 TFT/CPR
TFTs Farmed
Using all of the values derived above, we can make the final calculation for how many TFTs a given node will farm per month. Note that percentage values given for bandwidth, etc will need to be converted to decimal by dividing by 100 (50% = .5), and uptime is 1 or 0 depending on whether applicable uptime requirement was met.
TFTs farmed = Node CPR * CPR TFTs * Difficulty Level * Uptime * Utilization * Bandwidth
For certified nodes, this can be simplified to:
TFTs farmed = Node CPR * CPR TFTs * Difficulty Level * Uptime
Conclusion
There are a number of variables here that are under the Foundation’s control which will affect the ultimate number of TFTs farmed by a node, but this is the best information we presently have on what the new tokenomics look like in practice. Again, questions and corrections are welcome. Thanks!
thank you, this is amazing.
can you tell us where you found an issue on the wiki?
If you don’t mind you can leave all requests, remarks, improvement ideas on https://github.com/threefoldfoundation/info_foundation/issues
This allows us and the community to improve our docs.
many thanks
Sure thing, Kristof, and thank you for your reply. It’s really good to know that’s where the work is being managed on wiki updates, as I do have a number of ideas, many of which seem to be already in progress.
I can log this over on Github too, but to keep this thread complete, here’s what I’m seeing. Cloud unit calculations differ between the wiki spec and the ThreeBot admin interface.
In the table on this wiki page, we see:
In ThreeBot’s code, we see:
cu : Math.round(Math.min(r.mru/4*0.95, r.cru * 2) * 1000) / 1000,
su : Math.round(( r.hru / 1093 + r.sru / 91) * 1000) / 1000,
Hi all,
We’ve been receiving questions lately regarding tokenomics and here are some clarification to the topic on the current state:
1. Is there any locked TFT or TFTA tokens ?
- Yes, we have significant amount of tokens locked. Here you can see dashboard that shows locked tokens live stats: https://tokenstats.threefoldtoken.com/.
2. Is there any token burning in ThreeFold tokenomics?
- No, token burning was only a proposal idea discussed earlier 2020, but it never got accepted in production. So there is no token burning in the ThreeFold Grid.
3. The amount of TFT’s earned for farming is normalised against the US Dollar. Does this mean the amount of coins created is not at a fixed predictable rate but is instead adjusted based on the exchange rate? If so, does this change the speed of coin production / inflation?
- As explained here in the wiki: Until the end of year 2020 the following USD TFA price is used for minting: USD 0.15 plus 2% monthly starting May 2020.
Thank you everyone for support, let us know if anyone has other questions.