Specific questions:
The holding of each balance on-chain takes up an amount of resource in terms of storage, memory etc on nodes, it is preferable at a technical level to have only “real” accounts at the start of the chain to reduce chain bloat in the early days
Q: Wouldnt we only have real accounts if they are opted in?
Yes for those that are real opt ins. But it opens the door for malicious actors which has been seen before on the Votes module and Token spam for example, a specific example if we allowed 0.000001 XEM to opt in, the cost of the opt in Tx is 0, so it would be relatively simple and cheap to set up 10k of them and programmatically opt in.
Do you think that there is one large holder which controls a big part of the 114k accounts with <100 XEM? Asking because if someone wants to spam he has to have control over alot of accounts.
Setting a minimum threshold helps avoid spam of the post launch opt in process and makes it easier to ignore spam attemptsI think this only applies if there is indeed 1 or a few big holders which control a big part of the 114k accounts.
I have no idea or way to be certain but my gut tells me it is unlikely. The logic outlined I think holds for the existing 114k, but the number can number increase any time from now to snap shot if someone wanted to do so.
Setting a minimum threshold also makes it more expensive to initiate a spam disruption by placing a barrier on someone for example opting in 100,000 “dust” accounts with tiny balances
Again, this would only be possible if there is indeed 1 or a few holders which control the 114k accounts.
If you just create some new ones and register say 1m with 0.0001 XEM, now we have 1.1m account with less than 100XEM and 90% are controlled by a single person.
Just some assumptions… Lets say there is indeed 1 or a few holders which control 114k accounts, which waits to spam the symbol blockchain. What would stop them (besides from fees*) from sending everything to one account, optin, create 114k accounts on symbol and start spamming?
There is nothing to stop them opting in one large account and then splitting it up when they are on Symbol, they would be spamming the network. This approach doesn’t stop that, I haven’t said it would and it is a valid, albeit annoying, action on the Symbol mainnet (or NIS1 for that matter). There would be no spamming of the opt in process, but there would be spamming of the Symbol chain and it would bloat the number of accounts.
Fees as you note would be the disincentive to do this on the mainnet post launch.
I dont want to lose them just because some theoretical problems which seem to be mostly made up on the “spam problem”.
All security/dns/spam risks are theoretical unless they are utilised, if they aren’t used they don’t become an issue. We discussed this informally with the security review company and there is a risk there, how big, how likely etc is a judgement, I don’t pretend to know the answer and don’t think anyone else does either, it is a guess based on preference.
My opinion on it: I think most of the 114k accounts are “lost” and not controlled by anyone anymore. I do think there will be few active ones. I dont want to lose them just because some theoretical problems which seem to be mostly made up on the “spam problem”.
However, I see the problem with genesis block (limited amount of space) and why they cant be included from start.
I think this is the crux of your feedback and is probably valid, it looks to be shared by others on the chain as well.
Question
Are you suggesting then that a 0XEM threshold is preferable, or some kind of minimum threshold but not as high as 100XEM or something else?
