aboutsummaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2006-08-29Avoid Metaware's High-C warning "'=' encountered where '==' may have been ↵Gisle Vanem
intended."
2006-08-29Watcom lacks <sys/time.h>.Gisle Vanem
2006-08-29Added support for Watcom/DOS.Gisle Vanem
2006-08-29Updated dependency section.Gisle Vanem
2006-08-29Don't include zlib headers in dependency output.Gisle Vanem
2006-08-29Renamed config.dj -> config.dos.Gisle Vanem
2006-08-29Use config.dos instead. Updated generated dependencies.Gisle Vanem
2006-08-29Removed. New file is config.dos.Gisle Vanem
2006-08-29Renamed config.dj -> config.dos. Added #ifdef-section for djgpp.Gisle Vanem
2006-08-29BUFSIZE defined in Metaware's <stdio.h>. Undefine to avoid warning.Gisle Vanem
2006-08-29Support other MS-DOS compilers (MSDOS is a djgpp built-in define).Gisle Vanem
2006-08-29David McCreedy added CURLOPT_SOCKOPTFUNCTION and CURLOPT_SOCKOPTDATA toDaniel Stenberg
allow applications to set their own socket options.
2006-08-25Armel Asselin reported that the 'running_handles' counter wasn't updatedDaniel Stenberg
properly if you removed a "live" handle from a multi handle with curl_multi_remove_handle().
2006-08-22David McCreedy fixed a remaining mistake from the August 19 TYPE change.Daniel Stenberg
2006-08-22Peter Sylvester pointed out a flaw in the AllowServerConnect() in the FTPDaniel Stenberg
code when doing pure ipv6 EPRT connections.
2006-08-21Workaround for Cray UNICOS 9.0 to fix ftp.Dan Fandrich
2006-08-19Based on a patch by Armel Asselin, the FTP code no longer re-issues the TYPEDaniel Stenberg
command on subsequent requests on a re-used connection unless it has to.
2006-08-18Armel Asselin fixed a crash in the FTP code when using SINGLECWD mode andDaniel Stenberg
files in the root directory.
2006-08-18Andrew Biggs pointed out a "Expect: 100-continue" flaw where libcurl didn'tDaniel Stenberg
send the whole request at once, even though the Expect: header was disabled by the application. An effect of this change is also that small (< 1024 bytes) POSTs are now always sent without Expect: header since we deem it more costly to bother about that than the risk that we send the data in vain.
2006-08-16Minor portability fixes to get things running on UNICOS 9.0 on a Cray Y-MPDan Fandrich
2006-08-16Use gnutls_strerror() for clearer error message.Gisle Vanem
2006-08-15Use '_LIBICONV_VERSION' instead of variable '_libiconv_version'Gisle Vanem
to support older iconv versions.
2006-08-11Use __minix to detect Minix, which works on both ACK and GCC.Dan Fandrich
2006-08-09Only define the string prototypes in ANSI mode to reduce interference onDan Fandrich
systems that prototype them slightly differently.
2006-08-09added build info output.Gunter Knauf
2006-08-09fixed some web links.Gunter Knauf
2006-08-08Armel Asselin made the CURLOPT_PREQUOTE option work fine even whenDaniel Stenberg
CURLOPT_NOBODY is set true. PREQUOTE is then run roughly at the same place in the command sequence as it would have run if there would've been a transfer.
2006-08-08moved ugly NetWare hack to hostip.h so that hostip.c uses it too.Gunter Knauf
2006-08-08Fixed a flaw in the "Expect: 100-continue" treatment. If you did two POSTsDaniel Stenberg
on a persistent connection and allowed the first to use that header, you could not disable it for the second request.
2006-08-08Minix 3 doesn't have MSG_PEEKDan Fandrich
2006-08-07Allow again proper compilation outside of the source treeYang Tse
2006-08-04Initial stab at making libcurl compile under Minix 3.Dan Fandrich
2006-08-04Minor compatibility fixYang Tse
2006-08-04Added version info for iconv.Gisle Vanem
2006-08-04Added dependency for splay.obj.Gisle Vanem
2006-08-04oops, the previous commit was incomplete as we made an unconditional callDaniel Stenberg
to multi_runsingle() without it being really necessary or good
2006-08-04even when we get a single connection to deal with, we must still check forDaniel Stenberg
timeout'ed connections and possibly deal with them too
2006-08-04Fix compiler warningYang Tse
2006-08-04Avoid redundant check. configure script takes care of not definingYang Tse
HAVE_WINDOWS_H, HAVE_WINSOCK_H, HAVE_WINSOCK2_H, neither HAVE_WS2TCPIP_H when __CYGWIN__ is defined.
2006-08-03removed running_handles argument from multi_runsingle() since it wasn't reallyDaniel Stenberg
used anymore since multi->num_alive was introduced
2006-08-03Silence compiler warning 'unused parameter running_handles' in function ↵Yang Tse
multi_runsingle(). This is done here returning multi->num_alive in the running_handles parameter even when functions that call multi_runsingle() at this moment overwrite the returned value with the one that is valid when those functions curl_multi_perform() and multi_socket() have removed expired timers from the splay. Most probably, parameter 'running_handles' in function multi_runsingle() should be just removed.
2006-08-02keep count of the number of "alive" handles in a struct member, as otherwiseDaniel Stenberg
*multi_socket*() can't return the proper number
2006-08-02Mark Lentczner fixed how libcurl was not properly doing chunked encodingDaniel Stenberg
if the header "Transfer-Encoding: chunked" was set by the application. http://curl.haxx.se/bug/view.cgi?id=1531838
2006-08-01Maciej Karpiuk fixed a crash that would occur if we passed Curl_strerror()Daniel Stenberg
an unknown error number on glibc systems. http://curl.haxx.se/bug/view.cgi?id=1532289
2006-08-01spell-fixed a commentDaniel Stenberg
2006-07-31Silence warning: empty body in an if-statementYang Tse
2006-07-31Force compilation failure in case macros sread() or swrite() are not defined.Yang Tse
2006-07-31Provide definitions needed for macros sread() and swrite() in config file.Yang Tse
2006-07-30curl_multi_socket() and curl_multi_socket_all() got modified prototypes: theyDaniel Stenberg
both now provide the number of running handles back to the calling function.
2006-07-29Winsock and Cygwin need address family specification before bind(), this ↵Yang Tse
should be harmless for others.