Closed andreas-wuestefeld closed 2 years ago
format name is now *.miniDAS
My suggestion would be to avoid confusion by keeping .hdf5 as the extension. To determine the format type/version, the solution is to access the metadata.
If it is HDF5 I would argue for .hdf5
.
However it is to argue that our data is not hierarchical whatsoever and we may be better off defining a binary data format ourselves. After all the data is rather simple. HDF5 may be overkill.
I agree that .hdf5 or .h5 (more common in my experience) is better. Using HDF5 is nice because it is self-describing and can already be ready easily in many environments, so I'm happy with that as a format and don't think we need to devise another binary format. I agree with @jpmorten-asn that the file metadata should be used to tell that it is miniDAS and which version of the file format it is.
Although the data is in fact a HDF5 format, it would be beneficial to indicate data format by a distinctive extension
Does anyone has an opinion on the file extension ".das"? Is that already being used and causing conflicts? how about ".iris"?
other suggestions?