Re: -no-undefined on Win32

[Evgeny Grin]

18.04.2014, 19:17, "Bob Friesenhahn" <address@hidden>:
> Why does it create more headache when porting to Win32?  Using this
> option indicates that the project has been constructed in a way which
> will work on systems which do not allow undefined symbols.  Many
> projects (particularly those targeting only GNU/Linux because it is a
> popular operating system) are not suitably constructed and require
> adaptation.

As it's required mostly for Win32, project creators don't use this option by 
default. Moreover, to ensure maximum portability, this option often used 
conditionally for win32 only.
I can't get the idea, why libtool needs information that author is aware of 
undefined symbols (or aware of something else). This is inconsistent with other 
GNU tools. For example, GCC don't need information that author is aware of 
C++11 features. You can instruct GCC to use C++11 dialect or instruct to use 
C++03 dialect (and not use C++11 features).
I suggest to use the same logic for libtool. Use "-no-undefined" options to 
instruct linker to not use undefined symbols where it's possible to use 
undefined symbols. If some particular platform always need some compiler/linker 
option for shared lib then substitute required option automatically. The main 
idea of libtool (as I understand it) is simplification creation of static and 
shared lib, but currently this will make compilation more difficult than 
without libtool.

> Libtool always defaults to successful compilation and link, to the
> maximum extent possible.

That's nice, leave it to compiler and linker. If something can be compiled and 
linked, it will be compiled and linked. If it can't be, then compiler or linker 
will fail. Why giving up before even try?

-- 
Best Wishes,
Evgeny Grin