Details
-
Suggestion
-
Status: Accepted
-
Trivial
-
Resolution: Unresolved
-
None
-
None
-
None
-
Minor
Description
The current implementation of DFU server uses
enum DFUfileformat { DFUff_unknown = -1, DFUff_fixed, DFUff_csv, DFUff_utf8, DFUff_utf8n, DFUff_utf16, DFUff_utf16le, DFUff_utf16be, DFUff_utf32, DFUff_utf32le, DFUff_utf32be, DFUff_variable, DFUff_recfmvb, DFUff_recfmv, DFUff_variablebigendian}; and typedef enum { FFTunknown, FFTfixed, FFTvariable, FFTblocked, FFTcsv, FFTutf, // any format, default to utf-8n FFTutf8, FFTutf8n, FFTutf16, FFTutf16be, FFTutf16le, FFTutf32, FFTutf32be, FFTutf32le, FFTrecfmvb, FFTrecfmv, FFTvariablebigendian, FFTlast } FileFormatType;
Both of them (and probably some other parts of the platform) are a mixture between 'kind' (like 'CSV', 'fixed') and encoding (like 'utf-8', 'utf-16', etc.) file attributes.
I think it would be nice and necessary to clean them up especially in cases where we want to allow spraying/copying and convert differently encoded CSV, XML, etc. files.
But the problem is that the whole DFU server (and perhaps a part of the ECLWatch) implementation is dependent on them.
richardkchapman, ghalliday, jakesmith, gordonsmith, miguelvazq what is you opinion?