Discussion: Length of `moduleID` property

Implementation
1

I would like to discuss the length of the module ID property in this thread. This point was raised by @JesusTheHun in this thread: Define New Transaction Schema - #13 by JesusTheHun

In short, the question is: should we increase the length of module IDs, so that they can possibly be used to encode readable strings?

I summarized the pro and cons as follows:

pro: more user friendly module IDs
con: larger transactions and cross-chain messages
con: performance of the sparse Merkle tree. After the Merkle tree implementation is finalized we should benchmark it.

I opened this PR to update LIP 0040 for the type of module ID, I think we could also consider changing the length as part of this PR (and then update accordingly other LIPs): Update LIP0040 by ricott1 · Pull Request #151 · LiskHQ/lips · GitHub

1 Like
2

The discussion actually involves the moduleID but also the commandID.

1 Like
3

Can you clarify what should we consider about the commandID? Also increasing the byte size or something else?

4

We finally moved to strings for the module and command properties: https://github.com/LiskHQ/lips/pull/155

Similarly, we are updating other module LIPs to also use this new identification mechanism.