On 24.06.2013 17:57, Michael StJohns wrote:
> Hi Stef -
>
> My take is the general model is that CK_XXX is used for typedefs and
> structs, while CKx_XXX is used for value enumerations. The fact that
> the OTP stuff did this differently is an interesting point of
> information, but not compelling - it actually has a few "type"ing
> problems - see below.
Even if we exclude the OTP stuff, CK_XXX has been used for more than
typedefs, albeit not yet for enums. For example:
CK_TRUE
CK_FALSE
CK_INVALID_HANDLE
CK_UNAVAILABLE_INFORMATION
CK_EFFECTIVELY_INFINITE
> CKV_xxx might not be the correct prefix, but there are a number of these
> small enumerations that need to go somewhere and ideally not at the top
> level, but grouped and defined under an enum-like type definition.
>
> How about CKCC_XXX for certificate categories and CKCM_XXX for midp
> security domains?
In my opinion, these start to be too finicky and cluttered. -1
So again ... I would be in favor of either CKV_ or CK_<ENUM>_ (where
ENUM is a name like OTP or CERTIFICATE_CATEGORY).
> I don't have strong feelings on this, but if we're going to throw away
> the enum model, we should do it with eyes wide open.
Fair enough, and that's why I posted this question last week. Nobody
responded to that email.
Cheers,
Stef
