I'm now realizing the problem of g_format_size_for_display() + "/s". I have a new approach to that.
What upstream can do: Use bytes internally and write/use some helper functions for inputting/displaying them:
* bandwidth_unit(): Return a string containing the bandwidth unit (e.g. "KB/s")
* input_bandwidth_to_byte(size): Converts the input value to bytes (e.g. size * 1024)
Then we can write a small patch for Ubuntu that modifies these two functions:
* bandwidth_unit(): Return "kB/s"
* input_bandwidth_to_byte(size): Return size * 1000
Anything that I forgot?
@Martin: If we drop the g_format_size_for_display() patch, we have to patch all applications that use this function to comply with our policy.
I'm now realizing the problem of g_format_ size_for_ display( ) + "/s". I have a new approach to that.
What upstream can do: Use bytes internally and write/use some helper functions for inputting/ displaying them: _to_byte( size): Converts the input value to bytes (e.g. size * 1024)
* bandwidth_unit(): Return a string containing the bandwidth unit (e.g. "KB/s")
* input_bandwidth
Then we can write a small patch for Ubuntu that modifies these two functions: _to_byte( size): Return size * 1000
* bandwidth_unit(): Return "kB/s"
* input_bandwidth
Anything that I forgot?
@Martin: If we drop the g_format_ size_for_ display( ) patch, we have to patch all applications that use this function to comply with our policy.