It’s cool, it’s all cool, we just need to keep moving forward, slow or fast, but forward. That’s the key. We don’t want to knee jerk and hurt folk, we also don’t want to knee jerk and then reverse (we have done that and it’s boring ) and we don’t want to do any, let’s adjust this knob and see what it does, type “fixes” as those really nag at me and affect my sleep a lot.
We do want considered, focussed and measurable improvements. A key is measurable, how do you measure an ant? You cannot, so this is where the community is vital, we need to measure you folks reactions and findings, not so much yer opinions (sorry), but the reality of the use of the network, good and bad and we can never be defensive. So we need to dig amongst the rubble of even the “where’s my money” and “why has he got more than me” messages to carefully seek out network indicators and possible issues.
So it’s all working in it’s chaotic way and I believe more and more each day that the chaos is actually part of the playing field and we cannot scream at the moon, we need to just listen carefully and select what we can when we can.
It’s all part of the game we must play and we have all decided to be on this pitch, and we are all in the same game.
We actually had client calls that were FIND_CLOSEST that timed out and returned nodes THAT WERE NOT CLOSEST and that is a definite issue. It’s horrible, there are reasons, excuses and complex use case defences for it, but it’s just wrong. That one will be fixed and hopefully improve things. @Anselme 's analyser app will also help checking close nodes are close nodes and so on with random sampling of these things now in our grasp.
As above there’s a couple of issues that are critical, but as I said internally the amazing thing is the network holds at all right now, given these issues. So yes I think the retry logic is great @happybeing but we are leaning on app devs to go way too far beyond where they have to and we will address it.
There’s a significant effort to refactor and simplify ant-networking (check out those mega functions in there, it’s horrible). A couple of the guys, who shall remain nameless (it was @Anselme and Victor ) did a wee test of a new client in 100 lines of code instead of the 12000 lines (I did not typo that ) So there’s lots to play with and for here and we will get it done.
Check this rubble…
My wee Hetzner box consistently gets more relevant action than boxes from home (pro router has never used >20% of its session tables and 500Mb fibre)
My other home boxes are down cos the man cave is being rearranged but most never see half the no of records the Hetzner does and none of them are actually “relevant”.
Such promises worry me a lot because they are not based on the available data.
There is currently 1.53 EIB of free space in theory, to upload 30 PB of data would cost over $20 million at the current price of eth just for the Arbitrum fees according to several community calculations.
How much ANT will it cost to upload this data?
I very much doubt that in the next 2 years and in general before the native token comes we will see the main revenue coming from data uploads instead of subsidies…
I’m serious, such statements bother me because they drive people away.
Just like on the site it is written that the network is for people with free resources and literally every day people come to Discord with 3-20 nodes to complain that they don’t earn anything.
This is not ok, it’s nice to write that this is the ideal that the network strives for, but right now without hundreds of nodes it takes an unrealistically long time to earn even 1 token. At least 10 people I’ve brought in over the last 6 months have already given up, and they were each with 200-600 nodes.
-:
Wow, how are you doing with the workstations, I finished when I mentioned it to you and I haven’t been on Discord.
Dimitar Autonomi Bulgaria:
Currently, the mining has increased because the whale with the millions of nodes hasn’t updated
5000 nodes make 10-15 tokens per day
-:
wow
Yes, I’ll skip it. 200 nodes is more or less an acceptable maximum for me, but it’s a hobby considering that this is on the computer I’m working from.
I got people to invest, but in the range of about 1k roughly. 1k$. Well, the important thing is that no one complains to me - I explained it well to them: “You’re throwing away some money you don’t need in a very promising project and you’ll thank me someday”
I’m glad you are serious, it bothers me too.
I regard some of our “marketing” as downright misleading and eventually self-destructive.
But Ive said my piece and now Im only boring the rest of you with continual repetition.
Who was it years ago threatened to eat his socks?
Something to do with the token price not going above some magic level, wasn’t it?
CBA searching archives for “socks”
But FWIW, with the hopeful cut to emissions and fixes in the pipeline, I believe David when he says “thoon”
I can prove it is not if you want. Some of the nodes being quoted are definitely not the closest. And doing a quote 4 times within minutes gives 20 different nodes at least for the exact same chunk. And prob more if I did more quotes.
If you need more evidence to convince you definitely that the “closest 5” group is not being returned most times then just do the quote multiple times one after the other and see the nodes for each chunk.
The way I understand it is that the older nodes will be getting routed to before (or more correctly more often than) newer ones are, so a closer node not as old will take longer to find, maybe one extra request, but optimisations are happy with close enough it seems. This is what it seems from observations and reading the updates about the optimisations.
I understand it is being worked on and could change a lot with bug fixes that will get to the closest 5 anyhow. Just I feel that older nodes will be favoured somewhat
Anyhow just observations and spitballing here. Looking forward to where it ends up after all the investigations and fixes. Have to get a look at the analyser, sound like a great tool.
Are you saying that soon node runners will be getting more ANT from storing new records to the network than from the emmisions? If so, there is something I don’t understand.
The payments for storing a new record at the moment are in the order of 0.00000000186275819. I can tell because I have had some. I’ve also had some at 0.000000005588689584 so maybe 0.00000000186275819 is the absolute minimum possible in an empty network.
Let’s say for the sake of argument that the higher value is what people can expect per record.
1 / 0.000000005588689584 = 178932822.259967
So at the current (generous) ballpark figure for uploading a record each ANT can pay for the storage of up to 178932822 up to 4MB records on the network.
This implies that in order to earn 1 ANT from storing new records a noderunner would have to store up to (because of the 4MB max record size) 4 x 178932822 = 178932822 MB. That’s 174739 GB.
174739 / 32 = 5460. So that’s 5460 nodes.
Even if they are all full to their 32GB record store size of new records. Which they won’t be. And they won’t ever get to 100% full because the network will incentivise more nodes to be started.
Even if the Reward for storing records increases by a factor of 1000 from 0.000000005588689584 to 0.000005588689584 it’s still looking like most noderunners will never earn a single ANT from storing records.
But as long as they are running around 100 nodes they will earn 1 ANT every day on average because from running a few I estimate that the Emissions are around 0.01 ANT per node per day at the moment. There is clearly lots of variability but I hazard it is in that order or magnitude on average.
I don’t see how what can be earned from storing records will ever get anywhere near what can be earned by just running a nodes.
And most nodes will continue to be very underprovisioned in terms of actual storage and network bandwidth which will make the network fragile and poorly placed to deal with shocks of sudden decreases in the number of nodes or sudden increases in uploads or downloads.
Quite the opposite in my opinion, storage space is not a big factor in stability when it is very unlikely that there will be even 1 PB of data uploaded to the EiB network due to Eth fees.
There’s lot’s of opinions As I say I am focussed on the technical fundamentals for. solid working network. The money side I am less caring about right now. Not that I think the network can survive without money, that’s crazy.
In terms of emissions/data costs etc. then it’s back to fundamentals of what are we building here, a network that pays folk some token just for providing little value to humanity or a network whose purpose is security, freedom and privacy to folk.
The chat about emissions / data costs etc. is making an awful lot of opinions look like designs and they are not. My view is we get the thing working, living and breathing and see what apps will be built and see how valuable they are and then ensure we are the most valuable proposition for people to live their digital life securely.
When the dust settles and the foundation is solid then we can tell a lot, but these money focussed discussions are too much too soon IMO. The number of opinions will reach infinity and they will all be correct from some angle.
My job, in my eyes right now, is a solid working foundation and then we can attend to the loud opinions and what should charge what and what emissions should be paid to in what manner in what type of digital currency that we believe folk won’t consider “crypto” and all the stuff around that.
It’s very very hard to keep a conversation on the technical fundamentals of where we are right now in terms of working upload/download/API and having a base we can know works and is solid that is not “tuned” to work in one setting, but one that will adapt to scale and scale both ways (up and down) etc.
I appreciate all the opinions and appreciate all the folk feeling hurt, but it would be really nice to find the basics here. It really does get lost in a whole Buch of unrelated conversations and every thread falls into hurt feelings and loud voices that honestly make it very difficult to work out what’s happening technically in any depth.
I think this is what you’re missing; the payment required to store records will go up by a much bigger factor than 1000 as the network fills up.
There’s a pricing curve that starts low & goes up exponentially as nodes fill.
Unfortunately ‘soon’ may be quite a while (My ill-informed guess is months not weeks), as 1) it depends on uploads working well (hopefully not too long), 2) due to emissions there is an over-supply of nodes, which will suppress the ANT cost of uploading vs if there were fewer nodes to fill, and 3) ETH fees create a floor price for the cost of uploading data to the network, which may be a multiple of any equilibrium ANT upload cost, meaning it will never be ‘cheap’ to fill the relatively empty network in the early days as the pricing curve had intended it to be.
Also, as the network grows in scale over time, emissions will be split between ever more nodes, but growing demand for uploads should see node earnings from uploads remain relatively steady.
Ah. That will make a big difference then. I still think it would take a large increase in the upload cost to make Rewards for storing new records the larger incentive than Emissions for running nodes.
Yes, I agree with all that. I think it will take a long time for anything to change.
Maybe not right now but storage space will be a big factor when uploads start and we find that a large proportion of nodes simply don’t have enough space. At the moment a node would require much less than 1GB. When uploads increase a lot of nodes are going to pop which will cause others to get full. We all know what this looks like from last year.
The best we can hope is that some have provisioned their nodes with 0.5GB real storage, some with 1GB, some with 2GB, etc. so that the cascading failure is somewhat staggered.
And network bandwidth usage increasing will be factor as well.
) So there’s lots to play with and for here and we will get it done.
