Re: bignum branch

[Paul Eggert <eggert@cs.ucla.edu>]

Eli Zaretskii wrote:

> What are the
> arguments _for_ using "the standard libgmp that everybody uses"

It's what Emacs does for the other libraries that it uses, some of which are 
also essential for Lisp operation, and there is no good reason for treating 
libgmp differently. libgmp is a common dynamic library used by many other 
libraries and programs, Emacs is already dynamically linked to libgmp on 
GNU/Linux distributions with few if any reports of dynamic linking hell, so 
there is no good reason to change on these platforms.

> My main reason for linking GMP statically is that the way we integrate
> it into Emacs is different from all the other libraries we use in
> Emacs: the other libraries support optional features, whereas GMP
> supports a feature that is part of the Emacs Lisp language.

First, that's not true for Emacs. libc and libm are dynamically linked and are 
essential for Emacs Lisp.

Second, libgmp is essential for other GNU programs like GCC, and they work just 
fine dynamically linked. Emacs is not special here.

> I wasn't talking about GMP, I was talking about other libraries GDB
> links in.  libexpat, liblzma, even libintl -- all of those are linked
> statically here.  Maybe they, too, are linked dynamically on
> GNU/Linux

Yes, these libraries are linked dynamically on GNU/Linux.

> it isn't something outlandish.

Sorry, but static linking *is* outlandish on GNU/Linux. Again, maybe things are 
different on MS-Windows, but let's not let the tail wag the dog here. I'm not 
saying static linking never happens on GNU/Linux -- it does -- but it's 
definitely oddball nowadays.

Much of this discussion appears to have been based on a misunderstanding of how 
things commonly work on GNU/Linux. Static linking is no longer routinely used by 
applications on these platforms. Some GNU/Linux distributions don't even fully 
support static linking any more, for security reasons. It would be a mistake for 
Emacs to insist on it.