Why we can not remove data_top or data_start from _temp_event_data_type

[dachengx]

To Reproduce

If you remove data_top or data_start from _temp_event_data_type in straxen/plugins/events_nv/event_waveform_nv.py. Then run

pytest -xsv ./tests -k test_event_waveform_nv

you will see error like:

FAILED tests/plugins/test_plugins.py::PluginTest::test_event_waveform_nv - KeyError: "Field 'data_start' was not found in record with fields ('Start time since unix epoch [ns]', 'time', 'Length of the interval in samples', 'length', 'Width of one sample [ns]', 'dt', 'Waveform data in PE/sample (not PE/ns!)', 'data', 'Width (in ns) of the central 50% area of the peak', 'range_50p_area', 'Width (in ns) of the central 90% area of the peak', 'range_90p_area', 'Time between 10% and 50% area quantiles [ns]', 'rise_time', 'Dummy, total area of all hitlets in event [pe]', 'area', 'Dummy top waveform data in PE/sample (not PE/ns!)', 'data_top', 'Peak widths in range of central area fraction [ns]', 'width', 'Peak widths: time between nth and 5th area decile [ns]', 'area_decile_from_midpoint')"

This should not happen in principle, because this calling https://github.com/AxFoundation/strax/blob/a528e4b9b7d452c6c0a654692ed6fb688b4e3aae/strax/processing/peak_building.py#L243 should have used default arguments store_in_data_top=False, store_waveform_start=False.

There is an expedient to fix this but we'd better understand why this happens: #1447