fix

src/ic-cdk/CHANGELOG.md

@@ -13,7 +13,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   * When defining canister endpoints, we add the following attributes: `#[update(decoding_quota = 10000, skipping_quota = 100, debug = true)]`
     - `skipping_quota` limits the amount of work allowed for skipping unneeded data on the wire. If this attributes is not present, we set a default quota of `10_000`. This affects ALL existing canisters, and is mainly used to improve canister throughput. See [docs on the Candid library](https://docs.rs/candid/latest/candid/de/struct.DecoderConfig.html#method.set_skipping_quota) to understand the skipping cost.
     - `decoding_quota` limits the total amount of work the deserializer can perform. See [docs on the Candid library](https://docs.rs/candid/latest/candid/de/struct.DecoderConfig.html#method.set_decoding_quota) to understand the cost model.
-    - `debug = true` prints the instruction count and the decoding/skipping cost to the replica log, after a successful deserialization. `decoding_quota` and `skipping_quota` is logged only when you have already set a quota in the attributes. The debug mode is useful to determine the right quotas above. Developers can send a few large payload to the debugging endpoint and know the actual decoding cost.
+    - `debug = true` prints the instruction count and the decoding/skipping cost to the replica log, after a successful deserialization. The decoding/skipping cost is logged only when you have already set a quota in the attributes. The debug mode is useful to determine the right quotas above. Developers can send a few large payloads to the debugging endpoint and know the actual decoding cost.
   * When making inter-canister calls, we have a new function `call_with_config` to config the same decoding quotas described above. It's strongly recommended to use `call_with_config` when calling third-party untrusted canisters.
 
 ## [0.12.1] - 2024-01-12