Re: Announcing Binary Compatibility/Testing

[Kurt Garloff]

Hi Robert,

On Wed, Oct 13, 2004 at 07:24:15PM -0400, Robert Love wrote:
> I'd venture, in fact, to say that this effort is very important but does
> not affect the kernel at all.  Current "fault" lies in things e.g. like
> the C++ ABI, which is constantly fluctuating (rightly so, to fix bugs,
> but still).

These days, it's mostly libstdc++ which is still developing; the
compiler side (name mangling, layout of virt tables, ...) is quite
stable these days; it's these occasional bugs which are found from 
time to time in the most obscure scenarios, which are often overstated.

> Any other incompatibility lies in libraries, but we have library
> versioning.  There is nothing wrong with newer libs breaking
> compatibility so long as they have a different soname.  Vendors just
> need to ship compat libs and ISV's need to make sure they request the
> right lib and don't touch internals.

Except that we have deeps stacks of libraries these days and things 
break apart if you need different versions of one library at the same
time ... 

D needs libs C and B, but C needs A and B needs A'.
In reality, the picture is many more levels deep.

Thus breaking libs all the time brings nice challenges for integrators
and ISVs as well. Fortunately, most libs don't do that.

Regards,
-- 
Kurt Garloff, Director SUSE Labs, Novell Inc.

Attachment: pgp00000.pgp
Description: PGP signature