[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: RFS: tnat64 -- IPv4 to NAT64 redirector



Hi,

On Thu, 2011-06-30 at 14:26 +0300, Andrew O. Shadoura wrote:
> Hello,
> 
> On Thu, 30 Jun 2011 13:07:12 +0200
> "SZERB, Tamas" <toma@rulez.org> wrote:
> 
> > >> * Andrew O. Shadoura <bugzilla@tut.by>, 2011-06-30, 00:30:
> > >> >> 5. Lintian issues (lintian -iI --pedantic):
> > >> >>    W: tnat64: package-name-doesnt-match-sonames libtnat64-0.1
> > >> >>    W: tnat64: non-dev-pkg-with-shlib-symlink
> > >> >> usr/lib/libtnat64.so.0.1 usr/lib/libtnat64.so
> > >> >>    I: tnat64: no-symbols-control-file usr/lib/libtnat64.so.0.1
> 
> > >> >Isn't really relevant, as this is a LD_PRELOAD-able library.
> 
> > >> Then there's no reason to install the library into /usr/lib.
> > >> Install it to a private directory instead. And you don't need any
> > >> symlinks either.
> 
> > > Agree. In that case this is also relevant to tsocks package (Cc'ed).
> 
> > Could you provide the bug #number to see the whole?
> 
> There isn't a bug filed, but as my tnat64 is actually a fork of tsocks
> (he-he), this bug must present in tsocks too (as it
> has /usr/lib/tsocks.so).

this discussion about moving a shared lib into private namespace is
somewhat puzzling me. If the code it provides is only used internally -
why not link it in statically and don't ship it?

And if it can provide a useful ABI for others - why not make it a real
shared lib and ship it as such from the same source?

-- 
Best regards,
Kilian

Attachment: signature.asc
Description: This is a digitally signed message part


Reply to: