Re: Is there any reason to give our libraries "libXXX.{dll, lib}" names on Windows?

[Graham Bloice <graham.bloice () trihedral com>]

On 10 July 2014 07:30, Graham Bloice <graham.bloice () trihedral com> wrote:

> On 10 July 2014 02:06, Guy Harris <guy () alum mit edu> wrote:
>
> > On UN*X, the file names for libraries begin with "lib" - the linkers all
> > look for "libfoo.*" if you link with "-lfoo".
> >
> > On Windows, that's not the case.
> >
> > Is there any reason for us to put "lib" at the beginning of the file
> > names for the wsutil, wiretap, wireshark, etc. libraries?  I infer from
> >
> >         https://code.wireshark.org/review/2957
> >
> > that CMake might, by default, not put "lib" there, requiring us to, in
> > the CMake build, force it to be there on Windows by adding some stuff to
> > the CMakeLists.txt files.
>
> I can confirm that Windows CMake has an empty "prefix" value and authored
> that change only to make things the same as the nmake build.  I'm not
> fussed either way, but changing the library names may have consequences for
> the installer as well.  Currently I haven't got the windows CMake as far as
> building an installer yet.
>
> Note that on Windows nmake, filetap and wiretap don't get a "lib" prefix,
> instead they get a version suffix, e.g. "wiretap-1.99.0.dll"
>
> One other piece of info, without this change, CMake produces a
> wireshark.exe and a wireshark.dll and the .pdb file for the dll gets
> overwritten by the one for the exe.
Also, many of our 3rd party libraries follow the UN*X way, e.g.
"libglib-2.0-0.dll", so there is some form of consistency there.  In the
end I don't think it matters for Windows users, devs or the tools.


-- 
Graham Bloice

___________________________________________________________________________
Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
Archives:    http://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://wireshark.org/mailman/options/wireshark-dev
             mailto:wireshark-dev-request () wireshark org?subject=unsubscribe