[htdig-dev] Re: SSL troubles

[Lachlan A. <lh...@us...>]

Greetings Andy,

Thanks for the reminder.  I think this should be included in 3.2.0,=20
which we are preparing now.  Currently, we require an official bug=20
report before making any changes to CVS.  Could you please fill in a=20
bug report on sourceforge so we can track it properly?  Thanks.

Gabriele, this will require changing the configuration scripts again. =20
If I get stuck, could you please work your magic on them again?

My proposed solution is to remove the broken  CHECK_SSL  macro=20
entirely, and use  AC_CHECK_HEADERS(openssl/ssl.h)  which will define=20
the  HAVE_OPENSSL_SSL_H  macro, if it succeeds.  This means that=20
users include specific paths to  openssl/ssl.h  the same way as they=20
do to any other header we use.  It also means that the  ssl  code=20
can't be disabled, which is the current case, but appears not to be=20
the intention of the  CHECK_SSL  macro.

A trial patch is attached.  Running autoconf 2.57 produced a=20
=2E/configure  that worked on my system.  (Andy, I'll mail the new =20
configure  to you too, but don't want to post 800MB to this list!)  I=20
think it will fail if  openssl/ssl.h  is found but either  libssl  or =20
libcrypto  is missing.  Does anyone know how to make it skip the test=20
for  ssl.h  if one of the others fails?  Otherwise, we should add=20
HAVE_LIBCRYPTO  and  HAVE_LIBSSL  to our #ifdefs...

Cheers,
Lachlan

On Thu, 16 Oct 2003 01:09, And...@wi... wrote:
> Not to harp, but I tried building the  CVS a little bit ago and it
> still didn't set the HAVE_SSL_H stuff right, in that I had to go
> into the .h file and change the #define by hand.  Sorry, I don't
> have the details, I can't get to that machine right now, but I just
> thought I'd mention it.

--=20
lh...@us...
ht://Dig developer DownUnder  (http://www.htdig.org)