[omniORB] SSL support in v3.04 NT

[Dan Kegel]

Rob Riggs wrote:
> 
> Duncan Grisby wrote:
> 
> >On Thursday 13 December, Dan Kegel wrote:
> >
> >>Autoconf makes life soooooooo much easier.
> >>
> >
> >There are plenty of comments in the mailing list archives from people
> >who disagree with you.
> >
> >All things considered, I think autoconf probably will be the way
> >forward, but it won't be easy to make the change, and it will make
> >some people with strange platforms unhappy.
> >
> I'll toss my 2pf in and agree with Dan here. Autoconf solves far more
> problems than it creates. And I don't think that one can rightly hold
> autoconf responsible for the relative happiness of people with strange
> platforms.

I think we can have our cake and eat it too.  All autoconf
requires is bourne shell, really, and everyone has that,
even Windows.  Sure, converting is a lot of work, and
not many people are clued in about how to make it all
happen on cross-development systems and on Windows, but
if those of us who like autoconf pitch in, we can make
even the skeptics happy.

The main work involved in switching to autoconf is to move away from tests like
#ifdef BSD and move to #if HAVE_STRINGS_H,
i.e. use fine grained defines that can be automatically set
by an autoconf test at configure time.  Note that the
transition can be gradual, and doesn't even require autoconf;
you can switch to autoconf-style feature defines without
using autoconf.  And that won't hurt the strange-platform people one bit.

- Dan