Dash Evo and the 10 Sins

GrandMasterDash

Well-known member
Masternode Owner/Operator
I've read the Dash Evolution paper and I have some serious concerns to this proposal. I'm sure there are parts I simply just don't understand well enough. Anyway, here goes...

Account Requirements

I'm massively concerned that email phishing attacks could completely destroy dash's reputation. The process of receiving emails and clicking to confirm / invite / register opens up a HUGE can of worms:

1. How does a user know for sure the email is genuine? - just supplying the username and email address isn't good enough.

2. Email servers are notoriously prone to misconfiguration and thus hugely vulnerable to phishing opportunities. Especially true if we're giving MN operators the option to run their own email servers.

3. Email may be the baseline for developed countries, but in many parts of the world, there are millions of users that have no email address. In asia, for example, mobile phones are the primary connection to the Internet and services like wechat are hugely successful simply because they use the phone's telephone number.

Technologies

4. DSQL? Is it really SQL or, hopefully, a more modern key/value store? SQL is so last decade and SQL table maintenance when things blow up can be a real PITA. I specifically want to know which technologies are being used / proposed.

5. "Archive mode requiring a few servers with terrabytes of storage" - no no, not only does this amount to centralisation of sorts - even if it is encrypted - but also a nice easy resource for data analysis. With encryption, the weakness is in tomorrow's technology. I don't see why we've got to put it all in one place for someone to more easily analyse.

Friends and Money

6. "Sending money without a friendship link will require retrieving an address from the network which was never encrypted." My concern here, and I need someone to clear this up for me, is that merchants might purposely withhold anonymous payments and require friendship links, which I may not want.

7. Will it be absolutely necessary to create wallet addresses through MNs, or can I still do it offline?

8. "By allowing reservation of usernames and email addresses, we can ensure when you search for the user, imposters will not show up." But I disagree because the email system to register really is the weak link. Once in, someone could register a different user with a valid email.

9. What are user ratings for? - service? - honesty? - price? One rating to fit all? - don't get it

10. If a user sells their domain name, how do they move their wallet / reset their ratings?
 
With my ear to the ground, here's what I'm hearing:'
many don't like email requirement so they're thinking about how it might now be needed. However, this requirement can be encrypted and not public, so nobody but yourself has access to it. I think it was primarily a requirement to prevent DDOS in creating accounts. (not sure that's the exact type of attack, or is it Sybil? ) emails aren't used, as far as I know, for sending funds though. Only user names.

DSQL, I think, is just a placeholder name. What is being used is sharded and encrypted information being stored on multiple nodes to reduce the amount of required hard drive space masternode will be required to provide while storing petabytes of information. Plus, a user's wallet is the only way to retrieve and reassemble the information. Therefore, having full backups on several servers just in case all, say 5 servers that are storing the same shard of information, go offline, one of these "backup" but expensive servers will be able to seamlessly step in and provide the information. Because the information would still be in shards as well as 256 bit encrypted, there should be no worry about centralization or spying. It's simply a matter of redundancy.

Friends and money: I've been trying to get an answer to why, to what purpose, exchanging 5 addresses with friends accomplishes. I still don't really get it. If you can send anonymously by seamlessly retrieving an address from someone on demand, why store a bunch of family member's addresses on your phone? I don't get it, honestly. does it make a big difference in speed?

#7, I'm not sure, and it might be nobody knows yet as the structure of how exactly things are being set up isn't written in stone yet. It could be that the MNs issue somehow your next account address, or it could be your wallet? If it's your wallet, then I'd suspect if you could find that number out, you could send to it??
 
#8 I would guess, you would need to confirm the user name is indeed the person you want by checking with them. But yes, this could be a problem just like how websites look almost the same in the address bar, etc.... Also, anyone can make as many accounts as they like.

The ratings are like ebay ratings. It's your reputation. If you have a good rating, people can feel more confident dealing with you. If you have a bad rating, and you want to start anew, you start a new account. But you also have to start from scratch, so people will likely be wary of you for a long while. This means it's naturally not beneficial to screw with people, supposedly.

To start a new account, you come up with a new user name and send your funds to it.

You don't need to collect ratings, you can keep your user name and account info hidden from public view.

Remember, this is what I understand from what I see discussed. it's also not written in stone, but is in the process of being written in clay to soon be fired :)
 
TanteStefana thanks for that, I'm glad the email thing is being reviewed, that's put my mind at rest quite a bit.

I think there are some basic things that really should be present:

1. Network neutral

The system should be network neutral; a system that encourages standards for basic functions e.g. usernames, retrieving payment details, network billing etc. If dash goes it alone, it might win in the short to medium term, but in the long term other alts would collaborate and win.

My basic idea is that a user can chooses and assign a network name that is unique across all networks (dash MNs or other). Maybe work with SuperNET / Ethereum? The user, for example, might choose the network name, "GrandMaster" and assign it to the dash network.

Once assigned to the dash network, the user can create their own namespace with key/value pairs that works in the way they want. If they want usernames with just numbers, or if they want duplicate usernames, that's all up to them. The API will simply query the namespace.

2. Network charges

I know there's this idea that the system should be free for 90% of users but I personally think everything should be chargeable, albeit at micro-micro levels. This wouldn't completely reduce system abuse but I bet it would help enormously

Charges would be for namespace storage and API queries, say, per one thousand results. Maybe even compensate the end users a little, like an automatic tip?

3. Why?

The basic idea is that the dash network would behave more like a service to service providers. It would be a service equally accessible to individuals or large companies alike. I might, for example, create a network name and manage a very small namespace; those charges would be minuscule. On the other hand, I might grow a user base to millions and I may, or may not, charge my users, that's entirely up to me. If I didn't like the dash MN network e.g. missing features, I could simply assign the network name to a different network and start over with the exact same API.
 
GrandMasterDash, I would definitely PM Evan with your ideas. It's all still being designed even so the basics are being coded. So now is the time to discuss these ideas. And Evan is the perfect person to talk with, directly because your ideas have a depth another person acting as a liaison can not express.
 
I was hoping others might join in with their ideas and thoughts. We all have a vested interest in how this goes forward.
 
I'm thoroughly confused as to why this social networking aspect is being tacked on. I don't see how privacy/anonymity can possibly be maintained. It seems like a huge, confusing mess with no function or purpose.

Explain how I am wrong. Please.
 
8. "By allowing reservation of usernames and email addresses, we can ensure when you search for the user, imposters will not show up." But I disagree because the email system to register really is the weak link. Once in, someone could register a different user with a valid email.
This is exactly the part that confuses me. Why tie everything to an email? What the hell does email have to do with being a currency? Who's going to reserve my username and then extort me for it, or impersonate me? The whole fucking point is to dis-associate the money from identity information. Why can't the money just be money? Why do we have to tie all this useless, counter-intuitive, counter-productive crap to it? It's being encumbered tot e point that it not only defeats privacy completely, but creates a huge pile of new problems that wouldn't exist if this whole "attach it to an identity" thing were simply not done... DASH will become unusable and even worse for privacy than publicly publishing your bank statements in the New York Times...

I've got a $20 bill in my pocket that doesn't cause me any of this "sign up for shit" grief. It'd be simpler, easier, cheaper, more reliable, and more private to simply stick to the fiat system.
 
Why tie everything to an email? What the hell does email have to do with being a currency? Who's going to reserve my username and then extort me for it, or impersonate me? The whole fucking point is to dis-associate the money from identity information.

I think it's because:

1) Email addresses are unique, due to how DNS works (You could just use a throwaway email address, I believe it's just to check for uniqueness/reserve a username.)
2) They're a lot easier to remember than a random cryptographically-generated string of alphanumeric characters.
3) Almost everyone has at least one email address these days. It's a lot easier for John Q. Public to use & understand (and John Q. Public also knows how to make "throwaway" email addresses too.)

4) This enables mapping of a unique identity (real or fake) to a BIP32 public key chain.

I might be wrong on all of that, but that's my best guess at this time.
 
A throw-away email address still comes back to a name and an IP.

I have no need to remember a random, cryptographically-generated string of alphanumeric characters in the first place, much less make it easier to remember.

Maybe I don't want to use BIP32? Maybe I don't want to use deterministic addresses? You know, because they're deterministic... And, can be determined... Sure, my gpg encrypted backups could be brute-forced, maybe... But, you'd actually have to have physical possession of them AND brute force them. With a BIP32, all you have to do is brute-force the password. The gatekeeper for which is in perpetual exposure. Can never physically secure it. I don't want my money in the cloud. Hell, that's what banks are...
 
Back
Top