Could you optimize the make_preview() method? It's currently linearly walking the entire model, and since we have extensible schema now it should be easy to save a key which will allow sub-linear search (afterall there's an in-memory model sorted by timestamp). Also it's reading fields from the model that are already present in the ScopeResult.
Could you optimize the make_preview() method? It's currently linearly walking the entire model, and since we have extensible schema now it should be easy to save a key which will allow sub-linear search (afterall there's an in-memory model sorted by timestamp). Also it's reading fields from the model that are already present in the ScopeResult.