[stub-resolver] [getdns-api] getdns-api review - no brainers sections 2.2, 2.3, 2.6, 2.7

[Willem Toorop]

For the meeting: Do we need to make a decision on this?

op 31-01-14 19:24, Goyal, Neel schreef:
> On 1/31/14, 5:48 AM, "Willem Toorop" <willem at nlnetlabs.nl> wrote:
>>
>> Personally I feel more comfortable when I can "see" the pointers being
>> passed around.  So me personally would rather see this:
> 
> I agree.  Perhaps in the implementation files we use the full declaration
> and in the public header only show the opaque handle types.
> 
>>
>> typedef struct getdns_context getdns_context;
>> typedef struct getdns_dict getdns_dict;
>> typedef struct getdns_list getdns_list;
> 
> Or as a compromise we don¹t use the _t variants and just go w/ a typedef
> that removes the use of Œstruct¹.  Then pointers are clearly passed around
> and our public header declarations match those in the implementation file.
> 

_______________________________________________
Stub-resolver mailing list
Stub-resolver at verisignlabs.com
https://lists.verisignlabs.com/mailman/listinfo/stub-resolver