SCO port update - what now?


Sat Jun 11 09:33:43 PDT 2005


On Friday 10 June 2005 23:13, Daniel Stone wrote:
> On Fri, Jun 10, 2005 at 10:25:19AM -0700, Kean Johnston wrote:
> > > glxinfo is a C program, not a C++ program.  The only problem arises
> > > with broken linkers that do not properly link in libstdc++ to dependent
> > > libraries, but this is not glxinfo or imake's problem.
> >
> > Ah. Fair enough. But if using SimpleCplusplusTarget makes the compile
> > more robust in the presense of such broken linkers, is there a
> > downside to using it? Some unintended side-effect? I know this may be
> > a trade-off between "correct" and "robust" but ... I usually tend to
> > prefer the latter to the former :)
>
> I think it could have some unwanted side-effects, but can't remember
> which off the top of my head.  I assume the linker SCO's using requires
> this?

It requires that the program in question be written as a strict subset of C 
and C++ (remember that C++ adds keywords and thus there exist valid C 
programs that are not valid C++ programs).  And it also adds a dependency on 
the C++ ABI not breaking, which historically happens roughly 1.1 times per 
compiler revision.

I don't think any of those necessarily block this change, though it'd be nice 
to wrap it in a platform ifdef so we can identify which platforms have this 
broken linker behaviour.

- ajax
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.x.org/archives/xorg/attachments/20121004/ca4a6191/attachment.pgp>


More information about the xorg mailing list