Parachains v1 registrar module. #1559
Conversation
montekki
added
A0-please_review
|
Also it's not quite clear if this should check the limits such as validation code length and head data length. |
montekki
added
A3-in_progress
drahnr
left a comment
drahnr
left a comment
There was a problem hiding this comment.
Looks very clean, the amount of test code required to verify this is not insignificant
| Ok(()) | ||
| } | ||
|
|
||
| /// Swap a parachain with another parachain or parathread. The origin must be a `Parachain`. |
There was a problem hiding this comment.
What's the reasoning for this limitation? Is it not enough that the types of ParaId x and y are not equivalent?
There was a problem hiding this comment.
This is about the origin of the Call - parachains can send messages to the relay chain that are executed as calls to modules. So this requirement means that the parachains themselves must initiate the swap
Yeah, they are not meant to be used together. This registrar is for permissionless deployments, whereas the sudo wrapper is for permissioned (testnet) deployments. |
runtime/parachains/src/lib.rs
Outdated
| #[derive(PartialEq, Eq, Clone, Encode, Decode)] | ||
| #[cfg_attr(feature = "std", derive(Debug))] | ||
| #[derive(Debug, PartialEq, Eq, Clone, Encode, Decode)] | ||
| // TODO: #[cfg_attr(feature = "std", derive(Debug))] |
There was a problem hiding this comment.
TODO needs to be resolved. Maybe derive(RuntimeDebug) is an option?
| /// | ||
| /// Must be sent from a Signed origin that is able to have `ParathreadDeposit` reserved. | ||
| /// `gensis_head` and `validation_code` are used to initalize the parathread's state. | ||
| #[weight = 0] |
There was a problem hiding this comment.
did the previous registrar code have 0 weight also?
| impl<T: Trait> Module<T> { | ||
| /// Register a parachain with given code. Must be called by root. | ||
| /// Fails if given ID is already used. | ||
| pub fn register_parachain( |
There was a problem hiding this comment.
I note this doesn't set Debtors. The previous logic used something called SwapAux because deposits for parachains are in the slots module. I would prefer to have all deposits in one place, and this includes XCMP/HRMP channel deposits. So at some point we should have a deposits module where we can implement swapping over. I'm fine with the code committed as is for now, though.
montekki
added
A0-please_review
* master: Make collation an optional return (#1787) XCM: Land xcm-handler and xcm-executor (#1771) v0.8.25 (#1785) add two node local net script (#1781) Adjust max nominators down to 128 (from 256) (#1782) Companion for substrate/pull/7215 (#1768) Remove Stale Upgrades (#1780) Update Polkadot Weights for Substrate 2.0 (#1761) Parachains v1 registrar module. (#1559) Derive `From` for `AllMessages` and simplify `send_msg` (#1774) implement remaining subsystem metrics (#1770) Companion for paritytech/substrate#7236 (#1773) WIP: remove deprecated only/except clauses, build is now manual on PRs (#1769) Increase Westend `spec_version` (#1766) move Metrics to utils (#1765)
4 participants
Mostly a wrapper around
parasmodule, howeverparas_sudo_wrapper? It seems that they may conflict.ParaIds? Atm it's done by hand and I am not sure that the approach existing in the previous registrar would work.parasmodules. For instance, it should not allow registration forids that are outgoing in theParasmodule. It's done now by looking intoparasmodule but probably I've overlooked something.This also has to build on #1562
Closes #1508