Hi Marcel & Samuel, On 18 February 2011 22:05, Samuel Ortiz <sa...@linux.intel.com> wrote: > Hi Marcel, > > On Fri, Feb 18, 2011 at 11:54:43AM -0800, Marcel Holtmann wrote: >> Hi Samuel, >> >> > > + int32 Privacy [readonly] >> > An int16 would have been just fine. >> > >> > >> > > + Enable or disable IPv6 privacy extension >> > > + that is described in RFC 4941. The value >> > > + has only meaning if Method is set to "auto". >> > > + >> > > + Value <= 0 means that privacy extension is >> > > + disabled and normal autoconf addresses are >> > > + used. >> > > + Value 1 means that privacy extension is >> > > + enabled and system prefers to use public >> > > + addresses over temporary addresses. >> > > + Value >= 2 means that privacy extension is >> > > + enabled and system prefers temporary addresses >> > > + over public addresses. >> > > + >> > > + Default value is 1. >> > Now that this is user configurable, the default value should just be 0. >> >> can we please do proper strings here and not some magic integer numbers >> that are related to some kernel internal implementation. > Yes, that's even better. So something like "disabled", "enabled" and > "default".
When creating the patch I tried to invent a list of string that would describe the number values in kernel, but I was not very successfull so I just used numbers in this version (the numbers are easier to map to the kernel anyway :) What kind of mapping would do in your opinion? One choice would be like this 0 = disabled 1 = default 2 = enabled but that has the problem that the default is 1 instead of 0. Also kernel documentation mentions that default is 0 (for most devices) and -1 (for point-to-point devices). Should we somehow take this into account in connman? Regards, Jukka _______________________________________________ connman mailing list connman@connman.net http://lists.connman.net/listinfo/connman
