I think more for past safety reasons. Leif did quite a bit of work around the 5.x time frame to make the release logic more forgiving.
On Wed, Jun 13, 2018 at 4:22 PM, Walt Karas <wka...@oath.com.invalid> wrote: > In TSHandleMLocRelease(), in the one case where this function is not a > do-nothing, the "parent" is validated, but does not seem to be > actually used. Is the parent relationship needed for some > future-safety reason? > > On Wed, Jun 13, 2018 at 2:36 PM, Alan Carroll > <solidwallofc...@oath.com.invalid> wrote: > > I think it's a type pun of one of a variety of different classes. Type > > punning makes it less likely the user of the API will pass the wrong type > > to an API call, although given the apparent variety of things passed back > > as TSMLoc, I'm not sure how good a job it's doing of that. Although > > definitely void* would be worse. > > > > On Wed, Jun 13, 2018 at 1:58 PM, Walt Karas <wka...@oath.com.invalid> > wrote: > > > >> In the TS API, what is the rational for the existence of the type > >> TSMloc? How does it add value? As opposed to simply using void * . > >> >
