Name resolution system

[ripper234]

Motivation / context.

The idea is to create a separate embedded system (optional to use) that maps a unique string name to a currency ID (or other entities in the system).

The use-case is just like domain names: It's nice to be able to type in something unique other than "Currency ID 1273015" to get to a currency. Instead think of "currency-cool-hoverboard-tokens-series-a".

The design for this should borrow from DNS, Namecoin, and Nxt's embedded "DNS-like" system.

TL;DR

1. Names are unique
2. Registering a name entry would burn X MSC.
3. Renewing a name entry would burn Y MSC
4. Mapping a name you control to an asset would burn Z MSC (we don't want people switching these pointers around all the time!)
5. If you don't renew an asset and it expires, everybody can register that.

Things we need to decide:

1. Is uniqueness case sensitive or insensitive? (advantage of insensitive - we don't really want two assets names that differ only by upper/lower case)
2. ...

[LOLLOLOOLOL]

Can I suggest this as an Omniwallet ("premium") feature to integrate Namecoin? Realistically, Namecoin already functions (and can be used, both as DNS and other functions), and there's no reason to redesign the wheel.

[petertodd]

Namecoin doesn't actually have any real functionality beyond what Bitcoin already does, and it's highly centralised because only a few big pools mine it. (it's incompatible with p2pool's varience reduction) I can't recommend that we use it.

On 5 July 2014 23:27:37 CEST, LOLLOLOOLOL [email protected] wrote:

Can I suggest this as an Omniwallet ("premium") feature to integrate
Namecoin? Realistically, Namecoin already functions (and can be used,
both as DNS and other functions), and there's no reason to redesign the
wheel.

---

Reply to this email directly or view it on GitHub:
#216 (comment)

[taariq]

Agreed with Peter on this one. No functionality beyond novelty.

[ripper234]

What about my original idea of including a Namecoin-like resolution system in Master Protocol itself?
Wouldn't this be more useful than Namecoin?

@petertodd @taariq ?

[dacoinminster]

This is interesting. Why limit it to name->currency matchups though? Why
not do real DNS too? In fact, why not just a decentalized key/value table?

For instance:

type: "DNS-IPV4" key: "mydomain" value: "1.2.3.4"
type: "MSC-SP" key: "mycurrency" value: "1234"
type: "mynewtype" key: "mykey" value:"myvalue"
. . . .

Rules would be

• First-to-register
• Type+key must be unique
• Updates to value only allowed from issuing address

On Sun, Jul 6, 2014 at 2:07 AM, Ron Gross [email protected] wrote:

What about my original idea of including a Namecoin-like resolution system
in Master Protocol itself?
Wouldn't this be more useful than Namecoin?

@petertodd https://github.com/petertodd @taariq
https://github.com/taariq ?

—
Reply to this email directly or view it on GitHub
#216 (comment).

[ripper234]

Looks great, we can make it generic.
I think Ethereum has something along these lines.

[dacoinminster]

I probably don't need to add this, but I will anyway: this is another
feature that should burn a bit of Mastercoin :)

On Tue, Jul 15, 2014 at 12:05 PM, Ron Gross [email protected]
wrote:

Looks great, we can make it generic.
I think Ethereum has something along these lines.

—
Reply to this email directly or view it on GitHub
#216 (comment).

[ripper234]

+1 for burning some MSC for registering a name.

While designing something like this ... we probablly should to incorporate:

• Registering only for a limited period of time (to limit the effect of squatting)
• Transferring an name to another owner
• Putting up ownership of a name up for sale on the DEx

The first two are blockers, without that all major names would be squatted up to an infinite period of time.

[marv-engine]

Is this a proposal to use the blockchain to store arbitrary key/value pairs? I thought that's frowned on. Why would the blockchain be the right place to store that data? See #215 and the earlier proposal that's referenced there.

[petertodd]

Security is the reason. Now I don't agree with promoting yet more namespaces - I think they are disincentives to using Mastercoin - but if you are going to do is the blockchain is the place to do it in. Secondly on a technical level there's some clever stuff that can be done with using the UTXO set for it by bruteforcing prefixes to make it easy to retrieve entries later.

Incidentally, my proof of concept for blockpop, certificate transparency via the blockchain, is progressing nicely. It's interesting how its turning how to be a very similar problem: you have some key, a software package, and you want to know all values associated with it, all PGP signed releases issued by the developers.

On 16 July 2014 11:03:32 GMT-04:00, Marv Schneider [email protected] wrote:

Is this a proposal to use the blockchain to store arbitrary key/value
pairs? I thought that's frowned on. Why would the blockchain be the
right place to store that data? See #215 and the earlier proposal
that's referenced there.

---

Reply to this email directly or view it on GitHub:
#216 (comment)