Daniel Stenberg [Wed, 18 Feb 2004 12:22:56 +0000 (12:22 +0000)]
Made curl_accept() take a 'void *' instead of 'socklen_t *' in the 3rd
argument to also not force the casual includer to know about the socklen_t
type.
Daniel Stenberg [Wed, 18 Feb 2004 12:18:33 +0000 (12:18 +0000)]
Modified curl_accept() to take a 'void *' in the 2nd argument instead of
sockaddr *. This has the added benefit that source files that include
memdebug.h doesn't have to know about "sockaddr".
Daniel Stenberg [Wed, 18 Feb 2004 10:05:17 +0000 (10:05 +0000)]
No longer uses the 'ret' variable in the plain ipv4-version of
my_getaddrinfo() (caused a warning by the IRIX MIPSPro compiler). Also
clarified the situation for the 3-arg version of gethostbyname_r() with a huge
comment.
Daniel Stenberg [Wed, 18 Feb 2004 09:07:50 +0000 (09:07 +0000)]
The --enable-debug option really requires this to be built as part of curl.
When using it, we now set the include path to better find the devel curl
headers, and we check for the socklen_t type since the curl memdebug stuff
needs it.
Daniel Stenberg [Mon, 16 Feb 2004 13:14:55 +0000 (13:14 +0000)]
Fix verbosconnect() when ipv6-enabled to not assume that conn->serv_addr
is a valid pointer, but instead always depend on the passed-in dns pointer.
This happens to be NULL when the connection is re-used...
Daniel Stenberg [Mon, 16 Feb 2004 07:33:30 +0000 (07:33 +0000)]
Jeff Lawson pointed out that we need to check for a '5' in the version field
to properly work with SOCKS5 proxies. I also included some ascii art describing
the SOCKS5 response, as RFC1928 describes. Jeff provided details in bug
report #741841 and here: http://curl.haxx.se/mail/lib-2004-02/0181.html
Daniel Stenberg [Fri, 13 Feb 2004 09:50:23 +0000 (09:50 +0000)]
Ben Greear's SO_BINDTODEVICE patch that binds to a network interface "even
more" when the previous approach. Known to work on Linux, possibly on other
platforms as well.
Daniel Stenberg [Thu, 12 Feb 2004 16:02:55 +0000 (16:02 +0000)]
Make hostcache_fixoffset() take a long for offset, to fully work with 64bit
archs, also no longer typecast pointers to ints as that is a nono on 64bit
systems.
Daniel Stenberg [Thu, 12 Feb 2004 15:05:38 +0000 (15:05 +0000)]
If no nroff tool is found, or if no command line switch to nroff that converts
a man page to text is found, we disable the built-in manual stuff to still
be able to build.
Daniel Stenberg [Mon, 9 Feb 2004 07:12:33 +0000 (07:12 +0000)]
Ken Rastatter's fixes to improve portability of this example:
These minor changes remove portability issues with the this example and allow
it to run on Win32. Specifically:
* The use of pthread_create() has been replaced by g_thread_create(). This
removes the dependency on the pthreads library. Since this is an example using
GTK+, g_thread_create() is available as it is a part of glibc.
* The CURLOPT_FILE option is now referred to by its "newer name"
CURLOPT_WRITEDATA.
* The use of CURLOPT_WRITEFUNCTION has been added. As described in the docs,
this avoids the crashes when using a DLL under Win32.
* The output file has been renamed from "/tmp/test.curl" to "test.curl". It's
unlikely that there is a /tmp when in Win32 and other examples in libcurl
write their output files to the working directory.
Daniel Stenberg [Fri, 6 Feb 2004 12:13:20 +0000 (12:13 +0000)]
Rewrote the gethostbyname() check after Andrés García's provided patch
for finding it using mingw on windows.
I also made the script skip the search for gethostbyname_r and gethostbyaddr_r
when ipv6 is enabled.
Daniel Stenberg [Fri, 6 Feb 2004 07:28:49 +0000 (07:28 +0000)]
Make sure DynaGetFunction() returns a function pointer, not a data pointer.
The standards don't actually allow typecasts between data and functions so
some picky compilers warn about this.
Daniel Stenberg [Fri, 6 Feb 2004 07:15:27 +0000 (07:15 +0000)]
Remove the attempt to detect if we already tested the same source setup.
We really don't care, and so many other things could've changed to make the
new test interesting anyway.
Daniel Stenberg [Thu, 5 Feb 2004 21:40:05 +0000 (21:40 +0000)]
An attempt to only set both libz-related defines at the same time. We need
both the lib and the header present for both defines to be set. If only one
of the files is found, we issue a warning and set no define.