| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2003-08-20 | made curl_multi_info_read() set 'msgs_in_queue' to 0 even when it returns | Daniel Stenberg | |
| NULL! | |||
| 2003-08-19 | recent action | Daniel Stenberg | |
| 2003-08-19 | Now offering support for multiple -T on the same command line, just make | Daniel Stenberg | |
| sure you have one URL for each -T. A -T file name can also be "globbed" like -T "{file1,file2}". Test case 149 verifies this functionality. | |||
| 2003-08-19 | test case 149 is here | Daniel Stenberg | |
| 2003-08-19 | switch off globbing to enable [] in file names for -T | Daniel Stenberg | |
| 2003-08-19 | check the upload | Daniel Stenberg | |
| 2003-08-19 | teste multiple uploads in one command line | Daniel Stenberg | |
| 2003-08-19 | make sure the 'done' variable is always set to something in the | Daniel Stenberg | |
| Curl_is_resolved() function | |||
| 2003-08-19 | when --enable-debug is used to set debug options with gcc, use -Wno-long-long | Daniel Stenberg | |
| to inhibit long long warnings (ISO C90 does not support `long long') | |||
| 2003-08-19 | Respect HAVE_LONGLONG to support 'long long' | Daniel Stenberg | |
| 2003-08-19 | check for CRYPTO_cleanup_all_ex_data as well | Daniel Stenberg | |
| 2003-08-19 | CRYPTO_cleanup_all_ex_data() is not present in all OpenSSL versions so | Daniel Stenberg | |
| we need to check for its presence in the configure script | |||
| 2003-08-19 | better ignore pattern | Daniel Stenberg | |
| 2003-08-19 | test507 for multi with bad host name | Daniel Stenberg | |
| 2003-08-19 | test507 added | Daniel Stenberg | |
| 2003-08-19 | don't set done==TRUE if the host name doesn't resolve | Daniel Stenberg | |
| 2003-08-19 | Loren Kirkby pointed out that we need to call CRYPTO_cleanup_all_ex_data() | Daniel Stenberg | |
| when we cleanup the SSL stuff to not leak any memory. I wish this was documented anywhere. | |||
| 2003-08-18 | minor edit | Daniel Stenberg | |
| 2003-08-18 | minor edits | Daniel Stenberg | |
| 2003-08-18 | set the large-file support defines for the client too | Daniel Stenberg | |
| 2003-08-17 | setting WRITEFUNCTION or READFUNCTION to NULL will now reset the callback | Daniel Stenberg | |
| pointers to the internal default functions | |||
| 2003-08-15 | 7.10.7 | Daniel Stenberg | |
| 2003-08-15 | removed lots of "added in [version]" where [version] is resonably old | Daniel Stenberg | |
| 2003-08-14 | check for long long | Daniel Stenberg | |
| changed the use of AC_CHECK_TYPE as the previous approach is deprecated require 2.57 properly | |||
| 2003-08-14 | Possible code for large file support, added within #if 0 so far. | Daniel Stenberg | |
| 2003-08-14 | new Russian mirror both web and download | Daniel Stenberg | |
| 2003-08-14 | supprt for the new memlimit stuff | Daniel Stenberg | |
| 2003-08-14 | added one "added in blabla" and removed a few | Daniel Stenberg | |
| 2003-08-14 | Curl_SSL_InitSessions can return error, so check the return code and bail | Daniel Stenberg | |
| out if necessary | |||
| 2003-08-14 | Curl_llist_destroy() checks the input for non-NULL | Daniel Stenberg | |
| 2003-08-14 | new proto for Curl_hash_init | Daniel Stenberg | |
| 2003-08-14 | 1. check allocs | Daniel Stenberg | |
| 2. don't leave allocated memory behind when returning error | |||
| 2003-08-14 | return failure when the host cache creation fails | Daniel Stenberg | |
| 2003-08-14 | activate the new memory limit tests if requested | Daniel Stenberg | |
| only set cookiejar if selected | |||
| 2003-08-14 | return failure when an alloc function fails | Daniel Stenberg | |
| 2003-08-14 | prevent memory leak when going out of memory | Daniel Stenberg | |
| 2003-08-14 | allow out-of-memory testing by setting a limit. That number of memory | Daniel Stenberg | |
| allocation calls will succeed, the following will return NULL! | |||
| 2003-08-14 | better freeing when bailing out due to bad output glob | Daniel Stenberg | |
| 2003-08-14 | free data on failure | Daniel Stenberg | |
| 2003-08-14 | test87 verifies the new and better check for bad -o #[num] stuff | Daniel Stenberg | |
| 2003-08-14 | ignore the .pid files | Daniel Stenberg | |
| 2003-08-14 | ignore lib506 too | Daniel Stenberg | |
| 2003-08-14 | better report on why tests are skipped, and also show a count of the amount | Daniel Stenberg | |
| of test cases that were "considered". | |||
| 2003-08-14 | In case the output urlglob file name returned is NULL, then there was | Daniel Stenberg | |
| badness in the string and we help our user by returning an error. | |||
| 2003-08-14 | modified the #[num] code to be more robust, to return NULL on errors and | Daniel Stenberg | |
| to support numbers larger than 9 | |||
| 2003-08-14 | corrected this test case | Daniel Stenberg | |
| 2003-08-14 | test urlglobbing range requests | Daniel Stenberg | |
| 2003-08-14 | test86 added | Daniel Stenberg | |
| 2003-08-12 | David Byron's fix that makes this script use 'cygpath' instead of 'pwd' if | Daniel Stenberg | |
| this runs on windows, to find out the current working directory. | |||
| 2003-08-12 | 7.10.7-pre4 commit | Daniel Stenberg | |
 |
index : curl | |
| cURL mirror with patches applied | Ben |
| aboutsummaryrefslogtreecommitdiff |