5 \___|\___/|_| \_\_____|
10 - Changes done to buildconf script. Validate that autom4te and autoconf, as
11 well as aclocal and automake, versions match. Improve removal of previous
12 run generated files. Remove verbose debug logging of aclocal on Solaris.
14 Daniel Stenberg (5 Aug 2008)
15 - Yehoshua Hershberg found a problem that would make libcurl re-use a
16 connection with the multi interface even if a previous use of it caused a
17 CURLE_PEER_FAILED_VERIFICATION to get returned. I now make sure that failed
18 SSL connections properly close the connections.
20 Daniel Stenberg (4 Aug 2008)
21 - Test cases 1051, 1052 and 1055 were added by Daniel Fandrich on July 30 and
22 proved how PUT and POST with a redirect could lead to a "hang" due to the
23 data stream not being rewound properly when it had to in order to get sent
24 properly (again) to the subsequent URL. This is now fixed and these test
25 cases are no longer disabled.
28 - Autoconf 2.62 has changed the behaviour of the AC_AIX macro which we use.
29 Prior versions of autoconf defined _ALL_SOURCE if _AIX was defined. 2.62
30 version of AC_AIX defines _ALL_SOURCE and other four preprocessor symbols
31 no matter if the system is AIX or not. To keep the traditional behaviour,
32 and an uniform one across autoconf versions AC_AIX is replaced with our
33 own internal macro CURL_CHECK_AIX_ALL_SOURCE.
35 Daniel Stenberg (4 Aug 2008)
36 - Test case 1041 (added by Daniel Fandrich July 14th) proved a bug where PUT
37 with -C - sent garbage in the Content-Range: header. I fixed this problem by
38 making sure libcurl always sets the size of the _entire_ upload if an app
39 attemps to do resumed uploads since libcurl simply cannot know the size of
40 what is currently at the server end. Test 1041 is no longer disabled.
43 - No longer test availability of the gdi32 library, nor use it for linking, even
44 when we have been doing this since revision 1.47 of configure.ac 4 years and
45 5 months ago when cross-compiling a Windows target. We actually don't use any
46 function from the Windows GDI (Graphics Device Interface) related with drawing
47 or graphics-related operations.
49 Daniel Fandrich (1 Aug 2008)
50 - Added support for --append on SFTP uploads. Unfortunately, OpenSSH doesn't
51 support this so it goes untested.
54 - Configure process now checks if the preprocessor _REENTRANT symbol is already
55 defined. If it isn't currently defined a set of checks are performed to test
56 if its definition is required to make visible to the compiler a set of *_r
57 functions. Finally, if _REENTRANT is already defined or needed it takes care
58 of making adjustments necessary to ensure that it is defined equally for the
59 configure process tests and generated config file.
61 - Removed definition of CURL_CHECK_WORKING_RESOLVER from acinclude.m4 it has
62 not been in use since revision 1.81 of configure.in 6 years, 9 months ago.
64 Daniel Fandrich (31 Jul 2008)
65 - Fixed parsing of an IPv6 proxy address to support a scope identifier,
66 as well as IPv4 addresses in IPv6 format. Also, better handle the case
67 of a malformatted IPv6 address (avoid empty and NULL strings).
69 - Fixed a problem with any FTP URL or any URLs containing an IPv6 address
70 being mangled when passed to proxies when CURLOPT_PORT is also set
71 (reported by Pramod Sharma).
73 - User names embedded in proxy URLs without a password were parsed
74 incorrectly--the host name is treated as part of the user name and the
75 port number becomes the password. This can be observed in test 279
76 (was KNOWN_ISSUE #54).
78 Daniel Stenberg (30 Jul 2008)
79 - Phil Blundell added the CURLOPT_ADDRESS_SCOPE option, as well as adjusted
80 the URL parser to allow numerical IPv6-addresses to be specified with the
81 scope given, as per RFC4007 - with a percent letter that itself needs to be
82 URL escaped. For example, for an address of fe80::1234%1 the HTTP URL is:
83 "http://[fe80::1234%251]/"
85 - PHP's bug report #43158 (http://bugs.php.net/bug.php?id=43158) identifies a
86 true bug in libcurl built with OpenSSL. It made curl_easy_getinfo() more or
87 less always return 0 for CURLINFO_SSL_VERIFYRESULT because the function that
88 would set it to something non-zero would return before the assign in almost
89 all error cases. The internal variable is now set to non-zero from the start
90 of the function only to get cleared later on if things work out fine.
92 - Made the curl tool's -w option support the %{ssl_verify_result} variable
94 Daniel Fandrich (30 Jul 2008)
95 - Added test cases 1052 through 1055 to test uploading data from files
96 during redirects. Test cases 1052 and 1055 show problems (maybe the same
97 root cause as 1051) and are disabled.
99 - Fixed a couple of buffer overflows in the MS-DOS port of the curl tool.
101 Daniel Fandrich (29 Jul 2008)
102 - Fixed --use-ascii to properly convert text files on Symbian OS, MS-DOS
105 - Added test case 1051 to test Location: following with PUT, as reported
106 by Ben Sutcliffe. The test when run manually shows a problem in curl
109 Daniel Fandrich (28 Jul 2008)
110 - Fixed display of the interface bind address in the trace output when it's
113 - Added test cases 1045 through 1049 as simple tests of --interface using the
116 - Added test case 1050 to test --ftp-port with an IPv6 address
118 Daniel Stenberg (26 Jul 2008)
119 - David Bau filed bug report #2026240 "CURL_READFUNC_PAUSE leads to buffer
120 overrun" (http://curl.haxx.se/bug/view.cgi?id=2026240) identifying two
121 problems, and providing the fix for them:
123 - CURL_READFUNC_PAUSE did in fact not pause the _sending_ of data that it is
124 designed for but paused _receiving_ of data!
126 - libcurl didn't internally set the read counter to zero when this return
127 code was detected, which would potentially lead to junk getting sent to
130 Daniel Fandrich (26 Jul 2008)
131 - Added test 1044 to test large file support in ftp with -I.
133 - Eliminate a unnecessary socket creation in Curl_getaddrinfo for an IPv4
134 address in an IPv6 capable libcurl.
136 - Added feature in runtests.pl to select tests based on key word.
138 Daniel Fandrich (23 Jul 2008)
139 - Changed the long logfile elision code in runtests.pl to properly handle
142 - Changed references to TRUE and FALSE in the curl_easy_setopt man page to
143 1 and zero, respectively, since TRUE and FALSE aren't part of the
146 Daniel Stenberg (23 Jul 2008)
147 - I went over the curl_easy_setopt man page and replaced most references to
148 non-zero with the fixed value of 1. We should strive at making options
149 support '1' for enabling them mentioned explicitly, as that then will allow
150 us for to extend them in the future without breaking older programs.
152 Possibly we should even introduce a fancy define to use instead of '1' all
155 Yang Tse (21 Jul 2008)
156 - Use the sreadfrom() wrapper to replace recvfrom() in our code.
158 Yang Tse (20 Jul 2008)
159 - when recvfrom prototype uses a void pointer for arguments 2, 5 or 6 this will
160 now cause the definition, as appropriate, of RECVFROM_TYPE_ARG2_IS_VOID,
161 RECVFROM_TYPE_ARG5_IS_VOID or RECVFROM_TYPE_ARG6_IS_VOID.
163 Yang Tse (17 Jul 2008)
164 - RECVFROM_TYPE_ARG2, RECVFROM_TYPE_ARG5 and RECVFROM_TYPE_ARG6 are now defined
165 to the data type pointed by its respective argument and not the pointer type.
167 Yang Tse (16 Jul 2008)
168 - Configure process now checks availability of recvfrom() socket function and
169 finds out its return type and the types of its arguments. Added definitions
170 for non-configure systems config files, and introduced macro sreadfrom which
171 will be used on udp sockets as a recvfrom() wrapper.
173 Yang Tse (15 Jul 2008)
174 - Added description/comment to include paths used in several Makefile.am files.
175 Added automake option nostdinc to test servers makefile and modified libcurl
176 external headers include path for libtest programs.
178 Daniel Fandrich (14 Jul 2008)
179 - Added test1040 through test1043 to test -C - on HTTP. Test 1041 failed so
180 it's added to DISABLED.
182 Yang Tse (14 Jul 2008)
183 - HTTP_ONLY definition check in lib/setup.h is now done once that configuration
184 file has been included. In this way if symbol is defined in the config file
185 it will no longer be ignored. Removed inclusion of remaining system header
186 files from configuration files. Moved _REENTRANT definition up/earlier in
189 Yang Tse (11 Jul 2008)
190 - Added missing multiple header inclusion prevention definition for header
191 file content_encoding.h
193 Daniel Fandrich (11 Jul 2008)
194 - Fixed test 553 to pass the torture test.
196 Daniel Stenberg (11 Jul 2008)
197 - Daniel Fandrich found out we didn't pass on the user-agent properly when
198 doing "proxy-tunnels" with non-HTTP prototols and that was simply because
199 the code assumed the user-agent was only needed for HTTP.
201 Daniel Fandrich (10 Jul 2008)
202 - Changed slightly the SFTP quote commands chmod, chown and chgrp to only
203 set the attribute that has changed instead of all possible ones. Hopefully,
204 this will solve the "Permission denied" problem that Nagarajan Sreenivasan
205 reported when setting some modes, but regardless, it saves a protocol
206 round trip in the chmod case.
208 - Added test cases 1038 and 1039 to test Adrian Kreher's report that ftp
209 uploads with -C - didn't resume properly, but the tests pass.
211 Yang Tse (10 Jul 2008)
212 - Peter Lamberg filed bug report #2015126: "poll gives WSAEINVAL when POLLPRI
213 is set in fdset.events" (http://curl.haxx.se/bug/view.cgi?id=2015126) which
214 exactly pinpointed the problem only triggered on Windows Vista, provided
215 reference to docs and also a fix. There is much work behind Peter Lamberg's
216 excellent bug report. Thank You!
218 Daniel Fandrich (9 Jul 2008)
219 - Added tests 1036 and 1037 to verify resumed ftp downloads with -C -
221 Daniel Stenberg (9 Jul 2008)
222 - Andreas Schuldei improved Phil Blundell's patch for IPv6 using c-ares, and I
223 edited it slightly. Now you should be able to use IPv6 addresses fine even
224 with libcurl built to use c-ares.
226 Daniel Fandrich (9 Jul 2008)
227 - Fixed an OOM handling problem that cause test 11 to fail the torture test.
229 Daniel Fandrich (8 Jul 2008)
230 - Fixed test 554 to pass the torture test.
232 Daniel Fandrich (7 Jul 2008)
233 - Added test cases 1034 & 1035 to test IDN name conversion failures.
235 Daniel Stenberg (7 Jul 2008)
236 - Scott Barrett provided a test case for a segfault in the FTP code and the
237 fix for it. It occured when you did a FTP transfer using
238 CURLFTPMETHOD_SINGLECWD and then did another one on the same easy handle but
239 switched to CURLFTPMETHOD_NOCWD. Due to the "dir depth" variable not being
240 cleared properly. Scott's test case is now known as test 539 and it
243 Daniel Stenberg (3 Jul 2008)
244 - Phil Blundell provided a fix for libcurl's treatment of unexpected 1xx
245 response codes. Previously libcurl would hang on such occurances. I added
246 test case 1033 to verify.
248 - Introcuding a new timestamp for curl_easy_getinfo():
249 CURLINFO_APPCONNECT_TIME. This is set with the "application layer"
250 handshake/connection is completed. Which typically is SSL, TLS or SSH and by
251 using this you can figure out the application layer's own connect time. You
252 can extract the time stamp using curl's -w option and the new variable named
253 'time_appconnect'. This feature was sponsored by Lenny Rachitsky at NeuStar.
255 Daniel Fandrich (2 Jul 2008)
256 - Support Open Watcom C on Linux (as well as Windows).
258 Yang Tse (2 Jul 2008)
259 - The previously committed fix for bug report #1999181 prevented using the
260 monotonic clock on any system without an always supported POSIX compliant
261 implementation. Now the POSIX compliant configuration check is removed and
262 will fallback to gettimeofday when the monotonic clock is unavailable at
265 - The configure process will now halt when sed, grep, egrep or ar programs
266 can not be found among the directories in PATH variable.
268 Daniel Stenberg (1 Jul 2008)
269 - Rolland Dudemaine provided fixes to get libcurl to build for the INTEGRITY
272 Daniel Stenberg (30 Jun 2008)
273 - Made the internal printf() support %llu properly to print unsigned long longs.
275 - Stephen Collyer and Tor Arntsen helped identify a flaw in the range code
276 which output the range using a signed variable where it should rather use
279 Yang Tse (29 Jun 2008)
280 - John Lightsey filed bug report #1999181: "CLOCK_MONOTONIC always fails on
281 some systems" (http://curl.haxx.se/bug/view.cgi?id=1999181). The problem was
282 that the configure script did not use the _POSIX_MONOTONIC_CLOCK feature test
283 macro when checking monotonic clock availability. This is now fixed and the
284 monotonic clock will not be used unless the feature test macro is defined
285 with a value greater than zero indicating always supported.
287 Daniel Fandrich (25 Jun 2008)
288 - Honour --stderr with the -v option.
290 - Fixed a file handle leak in the command line client if more than one
291 --stderr option was given.
293 Daniel Stenberg (22 Jun 2008)
294 - Eduard Bloch filed the debian bug report #487567
295 (http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=487567) pointing out that
296 libcurl used Content-Range: instead of Range when doing a range request with
297 --head (CURLOPT_NOBODY). This is now fixed and test case 1032 was added to
300 Daniel Fandrich (21 Jun 2008)
301 - Stopped using ranges in scanf character sequences (e.g. %[a-z]) since that
302 is not ANSI C, just a common extension. This caused problems on
303 at least Open Watcom C.
305 Yang Tse (20 Jun 2008)
306 - Modified configuration script to actually verify if the compiler is good
307 enough at detecting compilation errors or at least it has been properly
308 configured to do so. Configuration heavily depends on this capability, so
309 if this compiler sanity check fails the configuration process will now fail.
311 Daniel Stenberg (20 Jun 2008)
312 - Phil Pellouchoud found a case where libcurl built with NSS failed to
313 handshake with a SSLv2 server, and it turned out to be because it didn't
314 recognize the cipher named "rc4-md5". In our list that cipher was named
315 plainly "rc4". I've now added rc4-md5 to work as an alias as Phil reported
316 that it made things work for him again.
318 - Hans-Jurgen May pointed out that trying SCP or SFTP over a SOCKS proxy
319 crashed libcurl. This is now addressed by making sure we use "plain send"
320 internally when doing the socks handshake instead of the Curl_write()
321 function which is designed to use the "target" protocol. That's then SCP or
322 SFTP in this case. I also took the opportunity and cleaned up some ssh-
323 related #ifdefs in the code for readability.
325 Daniel Stenberg (19 Jun 2008)
326 - Christopher Palow fixed a curl_multi_socket() issue which previously caused
327 libcurl to not tell the app properly when a socket was closed (when the name
328 resolve done by c-ares is completed) and then immediately re-created and put
329 to use again (for the actual connection). Since the closure will make the
330 "watch status" get lost in several event-based systems libcurl will need to
331 tell the app about this close/re-create case.
333 - Dengminwen found a bug in the connection re-use function when using the
334 multi interface with pipelining enabled as it would wrongly check for,
335 detect and close "dead connections" even though that connection was already
338 Daniel Fandrich (18 Jun 2008)
339 - Added SSH failure test cases 628-632
341 - Fixed a memory leak in the command-line tool that caused a valgrind error.
343 Daniel Stenberg (18 Jun 2008)
344 - Rob Crittenden brought a fix for the NSS layer that makes libcurl no longer
345 always fire up a new connection rather than using the existing one when the
346 multi interface is used. Original bug report:
347 https://bugzilla.redhat.com/show_bug.cgi?id=450140
349 Yang Tse (18 Jun 2008)
350 - Internal configure script improvement. No longer break out of shell "for"
351 statements from inside AC_FOO_IFELSE macros, otherwise temporary macro files
352 are not properly removed.
354 Daniel Fandrich (12 Jun 2008)
355 - Fixed curl-config --ca which wasn't being exported by configure.
357 Daniel Stenberg (11 Jun 2008)
358 - I did a cleanup of the internal generic SSL layer and how the various SSL
359 libraries are supported. Starting now, each underlying SSL library support
360 code does a set of defines for the 16 functions the generic layer (sslgen.c)
361 uses (all these new function defines use the prefix "curlssl_"). This
362 greatly simplified the generic layer in readability by involving much less
363 #ifdefs and other preprocessor stuff and should make it easier for people to
364 make libcurl work with new SSL libraries.
366 Hopefully I can later on document these 16 functions somewhat as well.
368 I also made most of the internal SSL-dependent functions (using Curl_ssl_
369 prefix) #defined to nothing when no SSL support is requested - previously
370 they would unnecessarily call mostly empty functions.
372 I've built libcurl with OpenSSL and GnuTLS and without SSL to test this and
373 I've also tried building with NSS but the NSS support is a mystery to me and
374 I failed to build libcurl with the NSS libraries I have installed. We really
375 should A) improve our configure script to detect unsuitable NSS versions
376 already at configure time and B) document our requirements better for the
379 Daniel Stenberg (10 Jun 2008)
380 - I made the OpenSSL code build again with OpenSSL 0.9.6. The CRLFILE
381 functionality killed it due to its unconditional use of
382 X509_STORE_set_flags...
384 Daniel Stenberg (8 Jun 2008)
385 - Due to the three new libcurl changes and the massive command line option
386 change I decided we'll mark it by bumping the next release number to 7.19.0!
388 - curl the tool now deals with its command line options somewhat differently!
389 All boolean options (such as -O, -I, -v etc), both short and long versions,
390 now always switch on/enable the option named. Using the same option multiple
391 times thus make no difference. To switch off one of those options, you need
392 to use the long version of the option and type --no-OPTION. Like to disable
393 verbose mode you use --no-verbose!
395 - Added --remote-name-all to curl, which if used changes the default for all
396 given URLs to be dealt with as if -O is used. So if you want to disable that
397 for a specific URL after --remote-name-all has been used, you muse use -o -
400 Daniel Stenberg (6 Jun 2008)
401 - Axel Tillequin and Arnaud Ebalard added support for CURLOPT_ISSUERCERT, for
402 OpenSSL, NSS and GnuTLS-built libcurls.
404 - Axel Tillequin and Arnaud Ebalard added support for CURLOPT_CRLFILE, for
405 OpenSSL, NSS and GnuTLS-built libcurls.
407 - Added CURLINFO_PRIMARY_IP as a new information retrievable with
408 curl_easy_getinfo. It returns a pointer to a string with the most recently
409 used IP address. Modified test case 500 to also verify this feature. The
410 implementing of this feature was sponsored by Lenny Rachitsky at NeuStar.
412 Version 7.18.2 (4 June 2008)
414 Daniel Fandrich (3 Jun 2008)
415 - Fixed a problem where telnet data would be lost if an EWOULDBLOCK
416 condition were encountered.
418 Marty Kuhrt (1 Jun 2008)
419 - Updated main.c to return CURLE_OK if PARAM_HELP_REQUESTED was returned
420 from getparameter instead of CURLE_FAILED_INIT. No point in returning
421 an error if --help or --version were requested.
423 Daniel Stenberg (28 May 2008)
424 - Emil Romanus found a problem and helped me repeat it. It occured when using
425 the curl_multi_socket() API with HTTP pipelining enabled and could lead to
426 the pipeline basically stalling for a very long period of time until it took
429 - Jeff Weber reported memory leaks with aborted SCP and SFTP transfers and
430 provided excellent repeat recipes. I fixed the cases I managed to reproduce
431 but Jeff still got some (SCP) problems even after these fixes:
432 http://curl.haxx.se/mail/lib-2008-05/0342.html
434 Daniel Stenberg (26 May 2008)
435 - Bug report #1973352 (http://curl.haxx.se/bug/view.cgi?id=1973352) identified
436 how the HTTP redirect following code didn't properly follow to a new URL if
437 the new url was but a query string such as "Location: ?moo=foo". Test case
438 1031 was added to verify this fix.
440 - Andreas Faerber and Scott McCreary made (lib)curl build for the Haiku OS.
442 Yang Tse (26 May 2008)
443 - David Rosenstrauch reported that header files spnegohelp.h and
444 openssl/objects.h were needed to compile SPNEGO support.
446 Daniel Fandrich (22 May 2008)
447 - Made sure to pass longs in to curl_easy_setopt where necessary in the
448 example programs and libtest code.
450 Daniel Stenberg (19 May 2008)
451 - When trying to repeat a multi interface problem I fell over a few multi
454 o with pipelining disabled, the state should never be set to WAITDO but
455 rather go straight to DO
457 o we had multiple states for which the internal function returned no socket
458 at all to wait for, with the effect that libcurl calls the socket callback
459 (when curl_multi_socket() is used) with REMOVE prematurely (as it would be
460 added again within very shortly)
462 o when in DO and DOING states, the HTTP and HTTPS protocol handler functions
463 didn't return that the socket should be waited for writing, but instead it
464 was treated as if no socket was needing monitoring so again REMOVE was
467 Daniel Stenberg (13 May 2008)
468 - Added test case 556 that uses curl_easy_send() and curl_easy_recv()
470 Daniel Stenberg (9 May 2008)
471 - Introducing curl_easy_send() and curl_easy_recv(). They can be used to send
472 and receive data over a connection previously setup with curl_easy_perform()
473 and its CURLOPT_CONNECT_ONLY option. The sendrecv.c example was added to
474 show how they can be used.
476 Yang Tse (9 May 2008)
477 - Internal time differences now use monotonic time source if available.
478 This also implies the removal of the winmm.lib dependency for WIN32.
480 Daniel Stenberg (9 May 2008)
481 - Stefan Krause reported a busy-looping case when using the multi interface
482 and doing CONNECT to a proxy. The app would then busy-loop until the proxy
483 completed its response.
485 Michal Marek (9 May 2008)
486 - Make Curl_write and it's callees accept a const pointer, in preparation
487 of tetetest's patch for curl_easy_send()
489 Daniel Stenberg (7 May 2008)
490 - Liam Healy filed the debian bug report #480044
491 (http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=480044) identifying a
492 segfault when using krb5 ftp, but the krb4 code had the same problem.
494 Yang Tse (7 May 2008)
495 - Christopher Palow provided the patch (edited by me) that introduces the
496 use of microsecond resolution keys for internal splay trees.
498 Daniel Stenberg (4 May 2008)
499 - Yuriy Sosov pointed out a configure fix for detecting c-ares when that is
502 Daniel Stenberg (3 May 2008)
503 - Ben Van Hof filed bug report #1945240: "libcurl sometimes sends body twice
504 when using CURL_AUTH_ANY" (http://curl.haxx.se/bug/view.cgi?id=1945240).
505 The problem was that when libcurl rewound a stream meant for upload when it
506 would prepare for a second request, it could accidentally continue the
507 sending of the rewound data on the first request instead of on the second.
508 Ben also provided test case 1030 that verifies this fix.
510 Daniel Stenberg (3 May 2008)
511 - Jean-Francois Bertrand reported a libcurl crash with CURLOPT_TCP_NODELAY
512 since libcurl used getprotobyname() and that isn't thread-safe. We now
513 switched to use IPPROTO_TCP unconditionally, but perhaps the proper fix is
514 to detect the thread-safe version of the function and use that.
515 http://curl.haxx.se/mail/lib-2008-05/0011.html
517 Daniel Stenberg (1 May 2008)
518 - Bart Whiteley provided a patch that made libcurl work properly when an app
519 uses the CURLOPT_OPENSOCKETFUNCTION callback to create a unix domain socket
522 Daniel Stenberg (29 Apr 2008)
523 - To make it easier for applications that want lots of magic stuff done on
524 redirections and thus cannot use CURLOPT_FOLLOWLOCATION easily, we now
525 introduce the new CURLINFO_REDIRECT_URL option that lets applications
526 extract the URL libcurl would've redirected to if it had been told to. This
527 then enables the application to continue to that URL as it thinks is
528 suitable, without having to re-implement the magic of creating the new URL
529 from the Location: header etc. Test 1029 verifies it.
531 Yang Tse (29 Apr 2008)
532 - Improved easy interface resolving timeout handling in c-ares enabled builds
534 Daniel Fandrich (28 Apr 2008)
535 - Added test 1028 to test an HTTP redirect to a FTP URL.
537 Daniel Stenberg (28 Apr 2008)
538 - Norbert Frese filed bug report #1951588: "Problem with curlftpfs and
539 libcurl" (http://curl.haxx.se/bug/view.cgi?id=1951588) which seems to be an
540 identical report to what Denis Golovan reported in
541 http://curl.haxx.se/mail/lib-2008-02/0108.html The FTP code didn't reset the
542 user/password pointers properly even though there might've been a new
543 struct/cconnection getting used.
545 Daniel Stenberg (26 Apr 2008)
546 - Reverted back to use automake 1.9.6 in the next release (from automake
547 1.10.1) since it *still* suffers from Solaris-related bugs. Our previous
548 automake 1.10 problem was reported in bug #1701360
549 (http://curl.haxx.se/bug/view.cgi?id=1701360) and this recent problem was
550 bug #1944825 (http://curl.haxx.se/bug/view.cgi?id=1944825). I have not
551 personally approached the automake team about either one of these but I
552 figure we need a Solaris 10 guy to do it!
554 Yang Tse (25 Apr 2008)
555 - Added 'timeout' and 'delay' attributes support for the test harness
556 <command> subsection.
558 Daniel Fandrich (24 Apr 2008)
559 - Made --stderr able to redirect all stderr messages.
561 Yang Tse (23 Apr 2008)
562 - Improve synchronization between test harness runtests.pl script
563 and test harness servers to minimize risk of false test failures.
565 Daniel Fandrich (22 Apr 2008)
566 - Added support for running on Symbian OS.
568 Daniel Fandrich (18 Apr 2008)
569 - Added test cases 1026 and 1027 to do some rudimentary tests on the --manual
572 Michal Marek (14 Apr 2008)
573 - allow disabling the typechecker by defining CURL_DISABLE_TYPECHECK, as
574 discussed in http://curl.haxx.se/mail/lib-2008-04/0291.html
576 Daniel Stenberg (14 Apr 2008)
577 - Stefan Krause reported a case where the OpenSSL handshake phase wasn't
578 properly acknowledging the timeout values, like if you pulled the network
579 plug in the midst of it.
581 - Andre Guibert de Bruet fixed a second case of not checking the malloc()
582 return code in the Negotiate code.
584 - Sandor Feldi reported bug #1942022
585 (http://curl.haxx.se/bug/view.cgi?id=1942022) pointing out a mistake in the
586 lib/Makefile.vc[68] makefiles' release-ssl-dll target.
588 - Brock Noland reported that curl behaved differently depending on which order
591 Daniel Stenberg (12 Apr 2008)
592 - Andre Guibert de Bruet found and fixed a case where malloc() was called but
593 was not checked for a NULL return, in the Negotiate code.
595 Daniel Fandrich (9 Apr 2008)
596 - Added test cases 1024 & 1025 to test a scenario similar to the one reported
597 by Ben Combee where libcurl would send the wrong cookie to a redirected
598 server. libcurl was doing the right thing in these test cases.
600 Michal Marek (7 Apr 2008)
601 - Fix the MIT / Heimdal check for good:
602 Define HAVE_GSSMIT if <gssapi/{gssapi.h,gssapi_generic.h,gssapi_krb5.h}> are
603 available, otherwise define HAVE_GSSHEIMDAL if <gssapi.h> is available.
605 Only define GSS_C_NT_HOSTBASED_SERVICE to gss_nt_service_name if
606 GSS_C_NT_HOSTBASED_SERVICE isn't declared by the gssapi headers. This should
607 avoid breakage in case we wrongly recognize Heimdal as MIT again.
609 Daniel Stenberg (5 Apr 2008)
610 - Alexey Simak fixed curl_easy_reset() to reset the max redirect limit properly
612 - Based on the Debian bug report #474224 that complained about the FTP error
613 message when libcurl doesn't get a 220 back immediately on connect, I now
614 changed it to be more specific on what the problem is. Also worth noticing:
615 while the bug report contains an example where the response is:
617 421 There are too many connected users, please try again later
619 we cannot assume that the error message will always be this readable nor
620 that it fits within a particular boundary etc.
622 Daniel Fandrich (3 Apr 2008)
623 - Added test627 to test SFTP with CURLOPT_NOBODY
625 Daniel Stenberg (3 Apr 2008)
626 - Setting CURLOPT_NOBODY to FALSE will now switch the HTTP request method to
627 GET simply because previously when you set CURLOPT_NOBODY to TRUE first and
628 then FALSE you'd end up in a broken state where a HTTP request would do a
629 HEAD by still act a lot like for a GET and hang waiting for the content etc.
631 - Scott Barrett added support for CURLOPT_NOBODY over SFTP
633 Daniel Fandrich (3 Apr 2008)
634 - Made sure that curl_global_init is called in all the multithreaded
637 Michal Marek (31 Mar 2008)
638 - Removed the generated ca-bundle.h file. The verbatim value of $ca and
639 $capath is known to configure, so it can be defined in config.h instead.
641 Daniel Stenberg (31 Mar 2008)
642 - Added CURLFORM_STREAM as a supported option to curl_formadd() to allow an
643 application to provide data for a multipart with the read callback. Note
644 that the size needs to be provided with CURLFORM_CONTENTSLENGTH when the
645 stream option is used. This feature is verified by the new test case
646 554. This feature was sponsored by Xponaut.
648 Daniel Fandrich (30 Mar 2008)
649 - Changed the makefile so the doc/examples/ programs are never built in a
650 normal build/install (only with the 'make check' target), so that a
651 build failure in the examples isn't fatal.
653 Version 7.18.1 (30 March 2008)
655 Daniel Stenberg (28 Mar 2008)
656 - Stephen Collyer pointed out that configure --with-libssh2 without a given
657 path didn't work properly.
659 Daniel Stenberg (27 Mar 2008)
660 - As found out and reported by Dan Petitt, libcurl didn't show progress/call
661 the progress callback for the first (potentially huge) piece of body data
662 sent together with the POST request headers in the initial send().
664 Daniel Stenberg (25 Mar 2008)
665 - Made setting the CURLOPT_SSL_CTX_FUNCTION option return a failure in case
666 libcurl wasn't built to use OpenSSL as that is a prerequisite for this
669 Daniel Stenberg (22 Mar 2008)
670 - Fixed the problem with doing a zero byte SCP transfer, verified with test
671 case 617 (which was added by Daniel Fandrich 5 Mar 2008).
673 Daniel Fandrich (20 Mar 2008)
674 - Fixed a problem where curl-config --protocols could erroneously show LDAPS
675 support when curl didn't even have regular LDAP support. It looks like
676 this could happen when the --enable-ldaps configure switch is given but
677 configure couldn't find the LDAP headers or libraries.
679 Michal Marek (20 Mar 2008)
680 - Added --with-ca-path=DIRECTORY configure option to use an openSSL CApath by
681 default instead of a ca bundle. The configure script will also look for a
682 ca path if no ca bundle is found and no option given.
684 - Fixed detection of previously installed curl-ca-bundle.crt
686 Daniel Fandrich (18 Mar 2008)
687 - Added test 626 to reproduce an infinite loop when given an invalid
688 SFTP quote command reported by Vincent Le Normand, and fixed it.
690 Michal Marek (18 Mar 2008)
691 - Added curl_easy_getinfo typechecker.
693 - Added macros for curl_share_setopt and curl_multi_setopt to check at least
694 the correct number of arguments.
696 Daniel Fandrich (13 Mar 2008)
697 - Added tests 622-625 to test SFTP/SCP uploads. Test 625 was an attempt to
698 reproduce the --ftp-create-dirs problem reported by Brian Ulm, but that
699 seems to need a call curl_easy_reset() which this test case doesn't do.
701 Daniel Stenberg (13 Mar 2008)
702 - Brian Ulm figured out that if you did an SFTP upload with
703 CURLOPT_FTP_CREATE_MISSING_DIRS to create a directory, and then re-used the
704 handle and uploaded another file to another directory that needed to be
705 created, the second upload would fail. Another case of a state variable that
706 wasn't properly reset between requests.
708 - I rewrote the 100-continue code to use a single state variable instead of
709 the previous two ones. I think it made the logic somewhat clearer.
711 Daniel Stenberg (11 Mar 2008)
712 - Dmitry Popov filed bug report #1911069
713 (http://curl.haxx.se/bug/view.cgi?id=1911069) that identified a race
714 condition in the name resolver code when the DNS cache is shared between
715 multiple easy handles, each running in simultaneous threads that could cause
718 - Added a macro for curl_easy_setopt() that accepts three arguments and simply
719 does nothing with them, just to make sure libcurl users always use three
720 arguments to this function. Due to its use of ... for the third argument, it
721 is otherwise hard to detect abuse.
723 Michal Marek (11 Mar 2008)
724 - Added a type checking macro for curl_easy_setopt(), needs gcc-4.3 and only
725 works in C mode atm (http://curl.haxx.se/mail/lib-2008-02/0267.html ,
726 http://curl.haxx.se/mail/lib-2008-02/0292.html )
728 Daniel Fandrich (10 Mar 2008)
729 - Added tests 618-621 to test SFTP/SCP transfers of more than one file
730 (test 620 tests the just-fixed problem reported by Brian Ulm).
732 Daniel Stenberg (9 Mar 2008)
733 - Brian Ulm reported a crash when doing a second SFTP transfer on a re-used
734 easy handle if curl_easy_reset() was used between them. I fixed it and Brian
735 verified that it cured his problem.
737 - Brian Ulm reported that if you first tried to download a non-existing SFTP
738 file and then fetched an existing one and re-used the handle, libcurl would
739 still report the second one as non-existing as well! I fixed it and Brian
740 verified that it cured his problem.
742 Michal Marek (6 Mar 2008)
743 - Fix the gssapi configure check to detect newer MIT Kerberos (patch by
746 Yang Tse (6 Mar 2008)
747 - Fix regression on Curl_socket_ready() and Curl_poll() so that these will
748 again fail on select/poll errors different than EINTR.
750 Daniel Fandrich (5 Mar 2008)
751 - Fixed the test harness so it will write out zero-length data files.
753 - Added tests 616 and 617 to see how SFTP and SCP cope with zero-length
754 files, as questioned by Mike Protts. SFTP does for me but SCP doesn't
755 so test 617 is disabled for now.
757 Daniel S (4 Mar 2008)
758 - Mike Protts brought a patch that makes resumed transfers work with SFTP.
760 Daniel S (1 Mar 2008)
761 - Anatoli Tubman found and fixed a crash with Negotiate authentication used on
762 a re-used connection where both requests used Negotiate.
764 Guenter Knauf (26 Feb 2008)
765 - Kaspar Brand provided a patch to support server name indication (RFC 4366).
767 Daniel S (25 Feb 2008)
768 - Kaspar Brand made GnuTLS-built libcurl properly acknowledge the option that
769 forces it to prefer SSLv3.
771 Daniel S (23 Feb 2008)
772 - Sam Listopad provided a patch in feature-request #1900014
773 http://curl.haxx.se/bug/feature.cgi?id=1900014 that makes libcurl (built to
774 use OpenSSL) support a full chain of certificates in a given PKCS12
777 Daniel S (22 Feb 2008)
778 - Georg Lippitsch made the src/Makefile.vc6 makefile use the same memory model
779 options as the lib/Makefile.vc6 already did.
781 Daniel S (21 Feb 2008)
782 - Zmey Petroff found a crash when libcurl accessed a NULL pointer, which
783 happened if you set the connection cache size to 1 and for example failed to
784 login to an FTP site. Bug report #1896698
785 (http://curl.haxx.se/bug/view.cgi?id=1896698)
787 Daniel S (20 Feb 2008)
788 - Fixed test case 405 to not fail when libcurl is built with GnuTLS
790 - Based on initial work done by Gautam Kachroo to address a bug, we now keep
791 better control at the exact state of the connection's SSL status so that we
792 know exactly when it has completed the SSL negotiation or not so that there
793 won't be accidental re-uses of connections that are wrongly believed to be
794 in SSL-completed-negotiate state.
796 - We no longer support setting the CURLOPT_URL option from inside a callback
797 such as the CURLOPT_SSL_CTX_FUNCTION one treat that as if it was a Location:
798 following. The patch that introduced this feature was done for 7.11.0, but
799 this code and functionality has been broken since about 7.15.4 (March 2006)
800 with the introduction of non-blocking OpenSSL "connects".
802 It was a hack to begin with and since it doesn't work and hasn't worked
803 correctly for a long time and nobody has even noticed, I consider it a very
804 suitable subject for plain removal. And so it was done.
806 Guenter Knauf (19 Feb 2008)
807 - We do no longer support SSLv2 by default since it has known flaws.
808 Kaspar Brand provided a patch for all supported SSL toolkits.
810 Daniel Fandrich (19 Feb 2008)
811 - Added test309 to test HTTP redirect to HTTPS URL
813 Daniel S (18 Feb 2008)
814 - We're no longer providing a very old ca-bundle in the curl tarball. You can
815 get a fresh one downloaded and created with 'make ca-bundle' or you can get
816 one from here => http://curl.haxx.se/docs/caextract.html if you want a fresh
817 new one extracted from Mozilla's recent list of ca certs.
819 The configure option --with-ca-bundle now lets you specify what file to use
820 as default ca bundle for your build. If not specified, the configure script
821 will check a few known standard places for a global ca cert to use.
823 Daniel S (17 Feb 2008)
824 - Jerome Muffat-Meridol helped me fix Curl_done() to close the current
825 connection by force when it was called before the entire request is
826 completed, simply because we can't know if the connection really can be
827 re-used safely at that point.
829 - Based on the same debugging logic, I've also made Curl_http_done() not
830 return CURLE_GOT_NOTHING if called "prematurely". This should have no real
831 effect to anything but the code makes more sense like this.
833 Daniel S (15 Feb 2008)
834 - Made the gnutls code path not even try to get the server cert if no peer
835 verification is requested. Previously it would even return failure if gnutls
836 failed to get the server cert even though no verification was asked for.
837 Public server showing the problem: https://www.net222.caisse-epargne.fr
839 - Fix my Curl_timeleft() leftover mistake in the gnutls code
841 - Pooyan McSporran found and fixed a flaw where you first would do a normal
842 http request and then you'd reuse the handle and replace the Accept: header,
843 as then libcurl would send two Accept: headers!
845 Daniel S (11 Feb 2008)
846 - Yang Tse pointed out a few remaining quirks from my timeout refactoring from
847 Feb 7 that didn't abort properly on timeouts. These are actually old
848 problems but now they should be fixed.
850 Yang Tse (10 Feb 2008)
851 - Bug report #1888932 (http://curl.haxx.se/bug/view.cgi?id=1888932) points out
852 and provides test program that demonstrates that libcurl might not set error
853 description message for error CURLE_COULDNT_RESOLVE_HOST for Windows threaded
854 name resolver builds. Fixed now.
856 Daniel Fandrich (8 Feb 2008)
857 - Added key words to all SSL-using tests so they can be skipped if necessary.
858 Removed a few unnecessary requires SSL statements.
860 Daniel S (8 Feb 2008)
861 - Mike Hommey filed and fixed bug report #1889856
862 (http://curl.haxx.se/bug/view.cgi?id=1889856): When using the gnutls ssl
863 layer, cleaning-up and reinitializing curl ends up with https requests
864 failing with "ASN1 parser: Element was not found" errors. Obviously a
865 regression added in 7.16.3.
867 Yang Tse (8 Feb 2008)
868 - Improved test harness SCP/SFTP start up server verification, doing a real
869 connection to the sftp server, authenticating and running a simple sftp
870 pwd command using the test harness generated configuration and key files.
872 Daniel S (8 Feb 2008)
873 - Günter Knauf added lib/mk-ca-bundle.pl which gets the Firefox ca bundle and
874 creates a suitable ca-bundle.crt file in PEM format for use with curl. The
875 recommended way to run it is to use 'make ca-bundle' in the build tree root.
877 Daniel Fandrich (7 Feb 2008)
878 - Added tests 1022 and 1023 to validate output of curl-config --version and
881 Daniel S (7 Feb 2008)
882 - Refactored a lot of timeout code into a few functions in an attempt to make
883 them all use the same (hopefully correct) logic to make it less error-prone
884 and easier to introduce library-wide where it should be used.
886 Yang Tse (6 Feb 2008)
887 - Fix an issue in strdup replacement function when dealing with absolutely
888 huge strings. Only systems without a standard strdup would be affected.
890 Daniel S (3 Feb 2008)
891 - Dmitry Kurochkin cleaned up the pipelining code and removed the need for and
892 use of the "is_in_pipeline" struct field.
894 - I wrote up and added the threaded-ssl.c example source code that shows how
895 to do multi-threaded downloads of HTTPS files with a libcurl that is built
896 with OpenSSL. It uses pthreads for the threading.
898 Daniel S (31 Jan 2008)
899 - Niklas Angebrand made the cookie support in libcurl properly deal with the
900 "HttpOnly" feature introduced by Microsoft and apparently also supported by
901 Firefox: http://msdn2.microsoft.com/en-us/library/ms533046.aspx . HttpOnly
902 is now supported when received from servers in HTTP headers, when written to
903 cookie jars and when read from existing cookie jars.
905 I modified test case 31 and 46 to also do some basic HttpOnly testing.
907 - Dmitry Kurochkin moved several struct fields from the connectdata struct to
908 the SingleRequest one to make pipelining better. It is a bit tricky to keep
909 them in the right place, to keep things related to the actual request or to
910 the actual connection in the right place.
912 Daniel S (29 Jan 2008)
913 - Dmitry Kurochkin fixed Curl_done() for pipelining, as it could previously
916 - Michal Marek fixed minor mistake in test case 553 that prevented it from
917 working on other IP-addresses or port numbers.
919 Version 7.18.0 (28 January 2008)
921 Daniel S (27 Jan 2008)
922 - Dmitry Kurochkin: In "real world" testing I found more bugs in
923 pipelining. Broken connection is not restored and we get into infinite
924 loop. It happens because of wrong is_in_pipeline values.
926 Daniel S (26 Jan 2008)
927 - Kevin Reed filed bug report #1879375
928 (http://curl.haxx.se/bug/view.cgi?id=1879375) which describes how libcurl
929 got lost in this scenario: proxy tunnel (or HTTPS over proxy), ask to do any
930 proxy authentication and the proxy replies with an auth (like NTLM) and then
931 closes the connection after that initial informational response.
933 libcurl would not properly re-initialize the connection to the proxy and
934 continue the auth negotiation like supposed. It does now however, as it will
935 now detect if one or more authentication methods were available and asked
936 for, and will thus retry the connection and continue from there.
938 - I made the progress callback get called properly during proxy CONNECT.
940 Daniel S (23 Jan 2008)
941 - Igor Franchuk pointed out that CURLOPT_COOKIELIST set to "ALL" leaked
942 memory, and so did "SESS". Fixed now.
944 Yang Tse (22 Jan 2008)
945 - Check poll.h at configuration time, and use it when sys/poll.h unavailable
947 Daniel S (22 Jan 2008)
948 - Dmitry Kurochkin removed the cancelled state for pipelining, as we agreed
949 that it is bad anyway. Starting now, removing a handle that is in used in a
950 pipeline will break the pipeline - it'll be set back up again but still...
952 Yang Tse (21 Jan 2008)
953 - Disable ldap support for cygwin builds, since it breaks whole build process.
954 Fixing it will affect other platforms, so it is postponed for another release.
956 Daniel S (18 Jan 2008)
957 - Lau Hang Kin found and fixed a problem with the multi interface when doing
958 CONNECT over a proxy. curl_multi_fdset() didn't report back the socket
959 properly during that state, due to a missing case in the switch in the
960 multi_getsock() function.
962 Yang Tse (17 Jan 2008)
963 - Don't abort tests 518 and 537 when unable to raise the open-file soft limit.
965 Daniel S (16 Jan 2008)
966 - Nathan Coulter's patch that makes runtests.pl respect the PATH when figuring
967 out what valgrind to run.
969 Yang Tse (16 Jan 2008)
970 - Improved handling of out of memory in the command line tool that afected
971 data url encoded HTTP POSTs when reading it from a file.
973 Daniel S (16 Jan 2008)
974 - Dmitry Kurochkin worked a lot on improving the HTTP Pipelining support that
975 previously had a number of flaws, perhaps most notably when an application
976 fired up N transfers at once as then they wouldn't pipeline at all that
977 nicely as anyone would think... Test case 530 was also updated to take the
978 improved functionality into account.
980 - Calls to Curl_failf() are not supposed to provide a trailing newline as the
981 function itself adds that. Fixed on 50 or something strings!
983 Daniel S (15 Jan 2008)
984 - I made the torture test on test 530 go through. This was actually due to
985 silly code left from when we switched to let the multi handle "hold" the dns
986 cache when using the multi interface... Of course this only triggered when a
987 certain function call returned error at the correct moment.
989 Daniel S (14 Jan 2008)
990 - Joe Malicki filed bug report #1871269
991 (http://curl.haxx.se/bug/view.cgi?id=1871269) and we could fix his hang-
992 problem that occurred when doing a large HTTP POST request with the
993 response-body read from a callback.
995 Daniel S (12 Jan 2008)
996 - I re-arranged the curl --help output. All the options are now sorted on
997 their long option names and all descriptions are one-liners.
999 - Eric Landes provided the patch (edited by me) that introduces the
1000 --keepalive-time to curl to set the keepalive probe interval. I also took
1001 the opportunity to rename the recently added no-keep-alive option to
1002 no-keepalive to keep a consistent naming and to avoid getting two dashes in
1003 these option names. Eric also provided an update to the man page for the new
1006 Daniel S (11 Jan 2008)
1007 - Daniel Egger made CURLOPT_RANGE work on file:// URLs the very same way it
1008 already worked for FTP:// URLs.
1010 - I made the curl tool switch from using CURLOPT_IOCTLFUNCTION to now use the
1011 spanking new CURLOPT_SEEKFUNCTION simply to take advantage of the improved
1012 performance for the upload resume cases where you want to upload the last
1013 few bytes of a very large file. To implement this decently, I had to switch
1014 the client code for uploading from fopen()/fread() to plain open()/read() so
1015 that we can use lseek() to do >32bit seeks (as fseek() doesn't allow that)
1016 on systems that offer support for that.
1018 Daniel S (10 Jan 2008)
1019 - Michal Marek made curl-config --libs not include /usr/lib64 in the output
1020 (it already before skipped /usr/lib). /usr/lib64 is the default library
1021 directory on many 64bit systems and it's unlikely that anyone would use the
1022 path privately on systems where it's not.
1024 - Georg Lippitsch brought CURLOPT_SEEKFUNCTION and CURLOPT_SEEKDATA to allow
1025 libcurl to seek in a given input stream. This is particularly important when
1026 doing upload resumes when there's already a huge part of the file present
1027 remotely. Before, and still if this callback isn't used, libcurl will read
1028 and through away the entire file up to the point to where the resuming
1029 begins (which of course can be a slow opereration depending on file size,
1030 I/O bandwidth and more). This new function will also be preferred to get
1031 used instead of the CURLOPT_IOCTLFUNCTION for seeking back in a stream when
1032 doing multi-stage HTTP auth with POST/PUT.
1034 - Nikitinskit Dmitriy filed bug report #1868255
1035 (http://curl.haxx.se/bug/view.cgi?id=1868255) with a patch. It identifies
1036 and fixes a problem with parsing WWW-Authenticate: headers with additional
1037 spaces in the line that the parser wasn't written to deal with.
1039 Daniel S (8 Jan 2008)
1040 - Introducing curl_easy_pause() and new magic return codes for both the read
1041 and the write callbacks that now can make a connection's reading and/or
1044 Daniel S (6 Jan 2008)
1045 - Jeff Johnson filed bug report #1863171
1046 (http://curl.haxx.se/bug/view.cgi?id=1863171) where he pointed out that
1047 libcurl's date parser didn't accept a +1300 time zone which actually is used
1048 fairly often (like New Zealand's Dailight Savings Time), so I modified the
1049 parser to now accept up to and including -1400 to +1400.
1051 Daniel S (5 Jan 2008)
1052 - Based on further discussion on curl-library, I reverted yesterday's SOCKS5
1053 code to instead introduce support for a new proxy type called
1054 CURLPROXY_SOCKS5_HOSTNAME that is used to send the host name to the proxy
1055 instead of IP address and there's thus no longer any need for a new
1056 curl_easy_setopt() option.
1058 The default SOCKS5 proxy is again back to sending the IP address to the
1059 proxy. The new curl command line option for enabling sending host name to a
1060 SOCKS5 proxy is now --socks5-hostname.
1062 Daniel S (4 Jan 2008)
1063 - Based on Maxim Perenesenko's patch, we now do SOCKS5 operations and let the
1064 proxy do the host name resolving and only if --socks5ip (or
1065 CURLOPT_SOCKS5_RESOLVE_LOCAL) is used we resolve the host name locally and
1066 pass on the IP address only to the proxy.
1068 Yang Tse (3 Jan 2008)
1069 - Modified test harness to allow SCP, SFTP and SOCKS4 tests to run with
1070 OpenSSH 2.9.9, SunSSH 1.0 or later versions. SOCKS5 tests need OpenSSH
1071 3.7, SunSSH 1.0 or later.
1073 Daniel S (2 Jan 2008)
1074 - I fixed two cases of missing return code checks when handling chunked
1075 decoding where a write error (or abort return from a callback) didn't stop
1076 libcurl's processing.
1078 - I removed the socklen_t use from the public curl/curl.h header and instead
1079 made it an unsigned int. The type was only used in the curl_sockaddr struct
1080 definition (only used by the curl_opensocket_callback). On all platforms I
1081 could find information about, socklen_t is 32 unsigned bits large so I don't
1082 think this will break the API or ABI. The main reason for this change is of
1083 course for all the platforms that don't have a socklen_t definition in their
1084 headers to build fine again. Providing our own configure magic and custom
1085 definition of socklen_t on those systems proved to work but was a lot of
1086 cruft, code and extra magic needed - when this very small change of type
1087 seems harmless and still solves the missing socklen_t problem.
1089 - Richard Atterer brought a patch that added support for SOCKS4a proxies,
1090 which is an inofficial PROXY4 variant that sends the hostname to the proxy
1091 instead of the resolved address (which is already supported by SOCKS5).
1092 --socks4a is the curl command line option for it and CURLOPT_PROXYTYPE can
1093 now be set to CURLPROXY_SOCKS4A as well.
1095 Daniel S (1 Jan 2008)
1096 - Mohun Biswas pointed out that --libcurl generated a source code with an int
1097 function but without a return statement. While fixing that, I also took care
1098 about adding some better comments for the generated code.