aboutsummaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2003-08-10don't claim the PASV connect is connected unless it *really* is!Daniel Stenberg
2003-08-08make sure the string is long enoughDaniel Stenberg
2003-08-08David Byron's fix for file:// URLs with drive letters included.Daniel Stenberg
2003-08-08chmod the cabundle file before we attempt to write to it, to makeDaniel Stenberg
make distcheck run fine
2003-08-08re-arranged the cwd/mkd stuff a bitDaniel Stenberg
2003-08-08Early Ehlinger's CURLOPT_FTP_CREATE_MISSING_DIRS patch was appliedDaniel Stenberg
2003-08-07use 644 for the chmodDaniel Stenberg
2003-08-06argh, it wasn't *that* easy to generate the ca-bundle header in the buildDaniel Stenberg
dir instead of the source dir, reverting that change
2003-08-06fix the treatment of the variable width specifier '*', which caused a bugDaniel Stenberg
in the urlglobbing just now, fixed in the debian bug tracker as Bug#203827
2003-08-06generate the ca-bundle.h in the build dir, and also make sure to chmodDaniel Stenberg
the file in the dist-hook to make distcheck run fine
2003-08-06include "share.h" for the cookie sharingDaniel Stenberg
2003-08-06make it build without ares supportDaniel Stenberg
make sure it set async false even when using ipv6 (made test case 20 fail before)
2003-08-06more fixDaniel Stenberg
2003-08-06updated the ares instructionDaniel Stenberg
2003-08-05LDFLAGS fix to make the GSSAPI build againDaniel Stenberg
2003-08-05added README.aresDaniel Stenberg
2003-08-05how to build with aresDaniel Stenberg
2003-08-05ares awareness/usage/support added. If configure --enable-ares is used, weDaniel Stenberg
build libcurl to use ares for asynch name resolves.
2003-08-05clean up the dir tree hierarchy in *_done() to make persistant connectionDaniel Stenberg
FTP use the correct directories! Reported in bug report #783116
2003-08-04 Jan Sundin reported a case where curl ignored a cookie that browsers don't,Daniel Stenberg
which turned up to be due to the number of dots in the 'domain'. I've now made curl follow the the original netscape cookie spec less strict on that part.
2003-08-04Dirk Manske's patch that introduces cookie support to the share interface.Daniel Stenberg
2003-08-03 Mark Fletcher provided an excellent bug report that identified a problemDaniel Stenberg
with FOLLOWLOCATION and chunked transfer-encoding, as libcurl would not properly ignore the body contents of 3XX response that included the Location: header.
2003-08-02When proxy authentication is used in a CONNECT request (as used for all SSLDaniel Stenberg
connects and otherwise enforced tunnel-thru-proxy requests), the same authentication header is also wrongly sent to the remote host. The name and password can then be captured by an evil host and possibly get used for malicious purposes.
2003-07-30Jörg Müller-Tolk updated this to build fine with 7.10.6Daniel Stenberg
2003-07-30Daniel Noguerol made the ftp code output "Accept-Ranges: bytes" in similarDaniel Stenberg
style like other faked HTTP headers when NOBODY and HEADER are used.
2003-07-30Reverted the 'filetime' struct field back to a 'long' as time_t is sometimesDaniel Stenberg
unsigned and we want this to be able to hold -1 for illegal/unset values.
2003-07-28clear http->send_buffer when we have freed the memory it pointed toDaniel Stenberg
2003-07-28François Pons brought a patch that once again made curl deal with ftp andDaniel Stenberg
"double slash" as indicating the root directory. In the RFC1738-fix of April 30, that ability was removed (since it is not the "right" way).
2003-07-25having it in CVS causes us problems *grrr*Daniel Stenberg
2003-07-25Andrés García updated with the added files etcDaniel Stenberg
2003-07-25With an unknown CA path, we undef the variable. To build properly withoutDaniel Stenberg
SSL/CA.
2003-07-25Removed #include <sys/resource.h>, as pointed out by Henry Bland we don'tDaniel Stenberg
need it.
2003-07-23When we re-use an existing connection we must make sure that we don'tDaniel Stenberg
accidentally re-use the connect_addr field, as that might no longer be around. Fix verified by Tracy Boehrer who basicly debugged and tracked down this problem.
2003-07-23minor code style fixDaniel Stenberg
2003-07-23Daniel Kouril's fix to make the GSS-Negotiate work fine.Daniel Stenberg
2003-07-23Juan F. Codagnone's fixes to build properly on Windows againDaniel Stenberg
2003-07-23Plain default version of this file, to allow users to build easier fromDaniel Stenberg
CVS. This will be updated by the configure script, and a default is placed here by the maketgz script.
2003-07-22reversed the check for GSSAPI when request that authDaniel Stenberg
2003-07-22CURLDEBUG, not MALLOCDEBUGDaniel Stenberg
2003-07-22More support for NTLM on proxies, now proxy state and nonce is stored inDaniel Stenberg
a separate struct properly.
2003-07-22The NTLM functions now take a 'proxy' argument as well.Daniel Stenberg
2003-07-22Added a separate struct for the proxyntlm data, as it will/can be differentDaniel Stenberg
than the remote server's. That is, both the server and the proxy can in fact require NTLM auth.
2003-07-21moved the proxyuser and proxypasswd fields from the sessionhandle to theDaniel Stenberg
connectdata to work as expected
2003-07-21adjusted to support NTLM for proxiesDaniel Stenberg
2003-07-21krb4-fixes for the moved user+password fields within the structsDaniel Stenberg
2003-07-20David Gardner pointed out in bug report 770755 that using the FTP command CWDDaniel Stenberg
with a blank argument is a bad idea. Now skip blanks.
2003-07-20If NTLM is requested, only re-use connections that have the exact sameDaniel Stenberg
credentials.
2003-07-19Access the user and passwd fields from the connectdata struct now insteadDaniel Stenberg
of the sessionhandle struct, as that was not good.
2003-07-19No longer stores user+password in the sessionhandle, now doing that in theDaniel Stenberg
connectdata struct instead. Each being an allocated pointer. The passwdgiven field was turned into a local variable in the only function it was being used.
2003-07-15Moved the NTLM credentials to the connectdata struct instead, as NTLMDaniel Stenberg
authenticates connections and not single requests. This should make it work better when we mix requests from multiple hosts. Problem pointed out by Cris Bailiff.