FreeLists / openbeosstorage / [openbeosstorage] Re: SK_TEST_* and old BeBooks

[Ingo Weinhold <bonefish@xxxxxxxxxxxxxxx>]

On Sun, 12 May 2002, Tyler Dauwalder wrote:

> 
> > > That's great. I suppose we ought to be able to #define
> > > USE_OPENBEOS_NAMESPACE now as well, correct? Cool.
> > The other day I read on the IK Team list, that in order to prepare the
> > move to the new directory structure/build system the OpenBeOS name space
> > shall be removed. Wait a second... there it is:
> >   http://www.freelists.org/archives/interfacekit/04-2002/msg00007.html
> > Mmh???
> 
> I personally don't see that leaving it in there causes any harm whatsoever. 
> The powers that be don't want an OpenBeOS namespace used -- why? The best I 
> can tell is they don't see it as necessary, which is true from the 
> perspective that you'll only be using BeOS R5 *or* OpenBeOS R1 at any given 
> time, never both at once. 
> 
> But if we ever want to write up binary compatibility tests that *do* use 
> both at the same time (which I kind of think would be cool), the only way 
> to do it is to wrap one of the versions in a namespace. We could change it 
> to StorageKit or something else if they don't want end-users to be 
> potentially confused by the presence of "namespace OpenBeOS" wrapped up in 
> #IFDEFs, but other than that I don't see that leaving it in is any sort of 
> liability.

I agree in general. The only thing that annoys me is, that BeIDE's
function seeking tool (you know, the `{}' menu that allows you to jump
directly to a function) doesn't work when using a namespace. But that's
a problem of BeIDE, not of the namespace.

BTW, I like rather using the StorageKit namespace for SK private stuff, as
we already have done. The global optional one should be OpenBeOS -- or
rather the name that will be picked.

CU, Ingo