[fetchmail]fetchmail 6.2.9-rc3 available

[Matthias Andree]

On Mon, 19 Sep 2005, Jim Correia wrote:

> On Sep 19, 2005, at 8:45 PM, Matthias Andree wrote:
> 
> >>mxget.c: In function ‘getmxrecords’:
> >>mxget.c:64: error: ‘HEADER’ undeclared (first use in this function)
> >>
> >>On Mac OS X 10.4.2 it appears the relevant structure is defined in
> >><arpa/nameser_compat.h>.
> >
> >Same here on Linux - and it's included from arpa/nameser.h.
> 
> FWIW, I read the headers again. On Mac OS X 10.4.2 it is defined in   
> <arpa/nameser_compat.h> or <arpa_nameser8_compat.h>.
> 
> The latter appears to only be included by <apra/nameser.h> if  
> BIND_8_COMPAT us defined.
> 
> Adding -DBIND_8_COMPAT to the CFLAGS also seems to solve the problem.

Ah well. I wonder if fetchmail is using some "down deep" BIND interface
it should not be meddling with.

> >3. what config.log contains about res_search
> 
> Is this the output you are looking for?
> 
> configure:8288: checking for res_search
> configure:8345: gcc -o conftest -g -O2   conftest.c  >&5
> configure:8351: $? = 0
> configure:8355: test -z
>              || test ! -s conftest.err
> configure:8358: $? = 0
> configure:8361: test -s conftest
> configure:8364: $? = 0
> configure:8376: result: yes
> 
> configure:8391: checking for res_search
> configure:8479: result: yes
> configure:8482: result: using libc's resolver functions
> configure:8563: checking for crypt
> configure:8620: gcc -o conftest -g -O2   conftest.c  >&5
> configure:8626: $? = 0
> configure:8630: test -z
>              || test ! -s conftest.err
> configure:8633: $? = 0
> configure:8636: test -s conftest
> configure:8639: $? = 0
> configure:8651: result: yes
> 
> #define HAVE_RES_SEARCH 1

Yup. res_search is in -lc on Darwin, -lresolv on others. So apparently
some header and library information don't match. I find this MX lookup
gear ugly as the night and it's not as useful any more as it used to be.

I'll review all this later when I can peek around the corners in Darwin
on a kind person's system :)

I presume if I get this fixed, it breaks again on Solaris...

Kind regards,

-- 
Matthias Andree