Open MikeStall opened 1 year ago
Also, we need to think how do we want to model this. One way is that the file is just the array of bytes. Another way is that the file is an array of bytes + metadata such as name, mime type.
Independently of the representation, we should have a way to:
For our needs, "Binary" format can be enough.
Do we need a file type?
This may be a purely SDK issue - from a language perspective, these would look like Records. Existing type checks should also ensure that we can't assign between real files and records.