Comment 15 for bug 1024435

On 08/08/2012 07:29 PM, Jamie Strandboge wrote:

> MIR review:
> * Does not ship a symbols file or use an empty argument to dh_makeshlibs -V. This should be considered, but is not a blocker.

Consider it considered. I'm not aware of shipping a symbols file, so
it's something for me to learn about and see whether I can apply it to
my other packages. That said, I have a release checklist that includes
bumping the version in the shlibs file if any new APIs have been added,
so my shlibs files are likely to be as tight as needed and not tighter.
An automated method for keeping this right is better than a manual one
though, so I'll look into it. I'm not sure whether this being a C++
library complicates it as it complicates enumerating symbols in the
ldscript (if I understand correctly). Any pointers (especially stack
protected ones, oh wait, wrong kind of pointer) would be welcome.

Thanks for the detailed review.

--Jay