The current implementation of the ComputeSSID function can lead to inconsistencies in the length of the generated SSID. This issue arises because the id parameter can have a variable length, resulting in SSIDs of varying lengths. Consequently, this can cause the SSID and ZKSsid lengths to be inconsistent, which has been observed to cause failures in verifying the Pedersen parameter proof during the resharing process.
Additionally, during the Reshare phase, the id used to compute ssid is derived from dkgResult.Bks[selfId].String(), generated during the DKG phase.
Is there any special significance to the current implementation, or can I use padding to address this issue?