Too short principal? It isn't

I call my canister with my principal:
dfx deploy zon_pst --argument 'record { owner = principal "qobqg-7qjc5-twsoq-gpyl7-ouflv-ftbw4-dxzof-lhkgp-qayke-unhpz-rae"; subaccount = null; }'

Error: Failed while trying to deploy canisters.
Caused by: Failed while trying to deploy canisters.
  Failed while trying to install all canisters.
    Failed to install wasm module to canister 'zon_pst'.
      Failed during wasm installation call: The replica returned a replica error: Replica Error: reject code CanisterError, reject message Canister avqkn-guaaa-aaaaa-qaaea-cai trapped explicitly: RTS error: blob_of_principal: principal too short, error code None

Why does it count the principal too short? I copied principals to try from NNS and (now) from NFID and they should be correct. But both principals fail with “too long” error message.

dfx 0.14.4

This looks correct to me. Is it possible that there’s something different going wrong? The error message could be misleading I suppose… Do the number of arguments match?

Yes, just one argument:

shared ({ caller = _owner }) actor class PST(
    creator : {owner : Principal; subaccount : ?ICRC1Types.Subaccount},
) : async ICRC1.FullInterface {

I would suggest to again go through over the canister actor code, may be you find Principal library used somewhere incorrectly.

I found the error. My bad, it is not related to Candid, it is another Principal than one passed through Candid.