When processing by group there is the option to wait for all features to be receive before splitting up the groups and there is an option to process when group changes.
Both options are great, however, in the “Process when group changes” method the process doesn’t actually start until the group value changes. This makes working with live data streams difficult sometimes when processing in group - especially trying to capture the last feature, or group of features.
In some cases it’s possible to know the size of the group(s) - for example in some cases the size is just 1. If the site of the group was given then once that size was reached then it could head off to be processed before the next group arrives.
see this question for use case:
This method would also allow for processing of groups of a defined size where the content isn’t so important. At present one must create an artificial group ID to support this workflow.
