platform/upstream/curl.git
21 years agofixing details for NTLM
Daniel Stenberg [Wed, 11 Jun 2003 16:14:45 +0000 (16:14 +0000)]
fixing details for NTLM

21 years agomore how I envision it _should_ work, but it still doesn't...
Daniel Stenberg [Wed, 11 Jun 2003 15:33:09 +0000 (15:33 +0000)]
more how I envision it _should_ work, but it still doesn't...

21 years agoto support "redirects" after the full body is transfered
Daniel Stenberg [Wed, 11 Jun 2003 15:31:40 +0000 (15:31 +0000)]
to support "redirects" after the full body is transfered

21 years agomade a nicer output for the decode test, as it served as a nice tool for me ;-)
Daniel Stenberg [Wed, 11 Jun 2003 15:31:06 +0000 (15:31 +0000)]
made a nicer output for the decode test, as it served as a nice tool for me ;-)

21 years agowhen we get the auth headers, we still need to read out the full body response
Daniel Stenberg [Wed, 11 Jun 2003 15:30:30 +0000 (15:30 +0000)]
when we get the auth headers, we still need to read out the full body response
as otherwise we can re-send requests on the same connection nicely

21 years agocorrect mistakes
Daniel Stenberg [Wed, 11 Jun 2003 14:05:13 +0000 (14:05 +0000)]
correct mistakes

21 years agodescribe the NTLM mechanism too
Daniel Stenberg [Wed, 11 Jun 2003 13:44:58 +0000 (13:44 +0000)]
describe the NTLM mechanism too

21 years agobasic NTLM support
Daniel Stenberg [Wed, 11 Jun 2003 13:44:31 +0000 (13:44 +0000)]
basic NTLM support

21 years agontlm added
Daniel Stenberg [Wed, 11 Jun 2003 13:42:53 +0000 (13:42 +0000)]
ntlm added

21 years agoInitial take at NTLM authentication. It doesn't really work at this point
Daniel Stenberg [Wed, 11 Jun 2003 13:38:55 +0000 (13:38 +0000)]
Initial take at NTLM authentication. It doesn't really work at this point
but the infrastructure is there.

21 years agoclarify the CUSTOMREQUEST and HTTPHEADER options slightly
Daniel Stenberg [Tue, 10 Jun 2003 13:06:38 +0000 (13:06 +0000)]
clarify the CUSTOMREQUEST and HTTPHEADER options slightly

21 years agoCURLOPT_HTTPAUTH docu
Daniel Stenberg [Tue, 10 Jun 2003 12:58:40 +0000 (12:58 +0000)]
CURLOPT_HTTPAUTH docu

21 years agoSet auth type differently, we use one CURLOPT_HTTPAUTH instead as we plan
Daniel Stenberg [Tue, 10 Jun 2003 12:49:16 +0000 (12:49 +0000)]
Set auth type differently, we use one CURLOPT_HTTPAUTH instead as we plan
to add more method in the future.

21 years agoDaniel Kouril for the HTTP negotiate stuff
Daniel Stenberg [Tue, 10 Jun 2003 12:25:01 +0000 (12:25 +0000)]
Daniel Kouril for the HTTP negotiate stuff

21 years agoDaniel Kouril's patch that adds HTTP negotiation support to libcurl was
Daniel Stenberg [Tue, 10 Jun 2003 12:22:19 +0000 (12:22 +0000)]
Daniel Kouril's patch that adds HTTP negotiation support to libcurl was
added.

21 years agowe fix more
Daniel Stenberg [Tue, 10 Jun 2003 12:07:10 +0000 (12:07 +0000)]
we fix more

21 years agomore generic
Daniel Stenberg [Tue, 10 Jun 2003 12:05:12 +0000 (12:05 +0000)]
more generic

21 years agoPass the error stream pointer to the URL globber, so that it can report
Daniel Stenberg [Tue, 10 Jun 2003 09:42:22 +0000 (09:42 +0000)]
Pass the error stream pointer to the URL globber, so that it can report
errors correctly to the user, if need be.

Also fixed so that a missing ] in the globbing process no longer leads
to core dump.

21 years agoWhen doing very big GET requests over HTTPS, we need to add some extra
Daniel Stenberg [Fri, 6 Jun 2003 14:58:26 +0000 (14:58 +0000)]
When doing very big GET requests over HTTPS, we need to add some extra
funky logic in order to make re-tries work fine with OpenSSL. This corrects
the problem David Orrell noticed.

21 years agoJust moved around some logic in Curl_write() to make it easier to debug.
Daniel Stenberg [Fri, 6 Jun 2003 14:56:50 +0000 (14:56 +0000)]
Just moved around some logic in Curl_write() to make it easier to debug.

21 years agospellfix
Daniel Stenberg [Fri, 6 Jun 2003 06:44:05 +0000 (06:44 +0000)]
spellfix

21 years agoReversed the logic to only include the <sys/select.h> header on systems
Daniel Stenberg [Thu, 5 Jun 2003 14:04:44 +0000 (14:04 +0000)]
Reversed the logic to only include the <sys/select.h> header on systems
known to really NEED it as another system that doesn't have it came up:
very old Linux libc5-based systems (as addition to all HPUX versions).

The only known system at this point is AIX.

21 years agoLDAP problem added as mention in bug report #735752
Daniel Stenberg [Tue, 3 Jun 2003 08:10:53 +0000 (08:10 +0000)]
LDAP problem added as mention in bug report #735752

21 years agoinclude the time headers just like we used to do in the curl/curl.h header
Daniel Stenberg [Tue, 3 Jun 2003 08:07:06 +0000 (08:07 +0000)]
include the time headers just like we used to do in the curl/curl.h header
once upon the time

21 years agowe want the time defines too
Daniel Stenberg [Tue, 3 Jun 2003 08:06:23 +0000 (08:06 +0000)]
we want the time defines too

21 years agoContent-Length: now overrides other means of knowing when the stream has
Daniel Stenberg [Tue, 3 Jun 2003 07:53:18 +0000 (07:53 +0000)]
Content-Length: now overrides other means of knowing when the stream has
ended.

21 years agoMake the Content-Length info override the Connection: close header, so that
Daniel Stenberg [Mon, 2 Jun 2003 14:57:08 +0000 (14:57 +0000)]
Make the Content-Length info override the Connection: close header, so that
libcurl will stop reading when the number of bytes have arrived and not wait
for a closed socket.

21 years agothe 500-599 test case range
Daniel Stenberg [Mon, 2 Jun 2003 14:48:27 +0000 (14:48 +0000)]
the 500-599 test case range

21 years agoignore more
Daniel Stenberg [Mon, 2 Jun 2003 13:55:40 +0000 (13:55 +0000)]
ignore more

21 years agoignore
Daniel Stenberg [Mon, 2 Jun 2003 13:53:13 +0000 (13:53 +0000)]
ignore

21 years agoDavid Byron's fix to get the progress-bar use the local size too when
Daniel Stenberg [Mon, 2 Jun 2003 13:42:42 +0000 (13:42 +0000)]
David Byron's fix to get the progress-bar use the local size too when
doing a resumed download.

21 years agomakefile fiddle
Daniel Stenberg [Mon, 2 Jun 2003 13:31:25 +0000 (13:31 +0000)]
makefile fiddle
changed how http requests are sent - now in one chunk more often
HPUX include fix in the external headers
better SSL work-arounds for bad SSL servers
modified error message when CURLE_HTTP_RETURNED_ERROR is returned

21 years agowork-around SSL implementation flaws better, pointed out in bug report
Daniel Stenberg [Mon, 2 Jun 2003 13:27:03 +0000 (13:27 +0000)]
work-around SSL implementation flaws better, pointed out in bug report
#745122.

21 years agomake a more descriptive error message when CURLE_HTTP_RETURNED_ERROR is
Daniel Stenberg [Mon, 2 Jun 2003 13:14:57 +0000 (13:14 +0000)]
make a more descriptive error message when CURLE_HTTP_RETURNED_ERROR is
returned

21 years agohaven't updates this in a loooong time
Daniel Stenberg [Wed, 28 May 2003 10:24:20 +0000 (10:24 +0000)]
haven't updates this in a loooong time

21 years agoPosting static data using POST and chunked encoded now also appends the
Daniel Stenberg [Wed, 28 May 2003 07:54:33 +0000 (07:54 +0000)]
Posting static data using POST and chunked encoded now also appends the
data to the initial request buffer, if the total post data is less than
100K.

21 years agoinclude sys/time.h, we didn't have a time() proto anymore. Did one of the
Daniel Stenberg [Tue, 27 May 2003 22:56:01 +0000 (22:56 +0000)]
include sys/time.h, we didn't have a time() proto anymore. Did one of the
changes in curl/curl.h make this occur?

21 years agoDocumented which rules the public headers must follow when we write
Daniel Stenberg [Tue, 27 May 2003 12:51:46 +0000 (12:51 +0000)]
Documented which rules the public headers must follow when we write
preprocessor checks for condititions.

21 years agooops, removed a # too many
Daniel Stenberg [Tue, 27 May 2003 12:51:15 +0000 (12:51 +0000)]
oops, removed a # too many

21 years agoremove usage of HAVE_* defines, we cannot and shall not depend on any such
Daniel Stenberg [Tue, 27 May 2003 12:45:51 +0000 (12:45 +0000)]
remove usage of HAVE_* defines, we cannot and shall not depend on any such
defines in the public external header files

21 years agosys/select.h is not present on HPUX, avoid including it
Daniel Stenberg [Tue, 27 May 2003 12:34:48 +0000 (12:34 +0000)]
sys/select.h is not present on HPUX, avoid including it

21 years agomade it work ;-)
Daniel Stenberg [Tue, 27 May 2003 12:18:00 +0000 (12:18 +0000)]
made it work ;-)

21 years agoif cvs update fails, attempt again after 5 seconds and retry 50 times
Daniel Stenberg [Tue, 27 May 2003 12:03:24 +0000 (12:03 +0000)]
if cvs update fails, attempt again after 5 seconds and retry 50 times
before giving up

21 years agoOnly build in lib and src by default, make the others dist-subdirs.
Daniel Stenberg [Tue, 27 May 2003 08:51:09 +0000 (08:51 +0000)]
Only build in lib and src by default, make the others dist-subdirs.
Make the test stuff get built when we run 'make test' instead.

21 years agoRudy Koento experienced problems with curl's recent habit of POSTing data in
Daniel Stenberg [Tue, 27 May 2003 08:33:08 +0000 (08:33 +0000)]
Rudy Koento experienced problems with curl's recent habit of POSTing data in
two separate send() calls, first the headers and then the data. I've now made
a fix that for static and known content that isn't to be chunked-encoded,
everything is now sent in one single system call again. This is also better
for network performance reasons.

21 years agoruns on DOS now
Daniel Stenberg [Tue, 27 May 2003 07:37:34 +0000 (07:37 +0000)]
runs on DOS now

21 years agoinclude file flaw and yet another socks5-fix
Daniel Stenberg [Tue, 27 May 2003 06:41:06 +0000 (06:41 +0000)]
include file flaw and yet another socks5-fix

21 years agoAnother socks5-fix. Make sure that when we use a socks-proxy, it is not the
Daniel Stenberg [Tue, 27 May 2003 06:28:25 +0000 (06:28 +0000)]
Another socks5-fix. Make sure that when we use a socks-proxy, it is not the
same as using a httpproxy so we must make sure to better check for http
proxies before we do HTTP proxy stuff. This included authorization and
URI usage in the request etc.

21 years agoCURLOPT_HTTPDIGEST is added
Daniel Stenberg [Tue, 27 May 2003 06:25:56 +0000 (06:25 +0000)]
CURLOPT_HTTPDIGEST is added

21 years agolanguage
Daniel Stenberg [Mon, 26 May 2003 12:32:22 +0000 (12:32 +0000)]
language

21 years agoftp ASCII transfers in general need fixing
Daniel Stenberg [Mon, 26 May 2003 08:19:06 +0000 (08:19 +0000)]
ftp ASCII transfers in general need fixing

21 years agoChris Lewis mentioned that he doesn't get WIN32 defined, only _WIN32 so we
Daniel Stenberg [Mon, 26 May 2003 07:57:53 +0000 (07:57 +0000)]
Chris Lewis mentioned that he doesn't get WIN32 defined, only _WIN32 so we
make an adjustment to catch this.

21 years agoeven more
Daniel Stenberg [Fri, 23 May 2003 11:24:39 +0000 (11:24 +0000)]
even more

21 years agoRicardo Cadime found a socket leak when listing directories without
Daniel Stenberg [Fri, 23 May 2003 11:14:09 +0000 (11:14 +0000)]
Ricardo Cadime found a socket leak when listing directories without
contents. Test cases 144 and 145 were added to verify the fix.

Now we deal with return code 450 properly and other codes also do proper
cleanup.

21 years agomore ftp testing using NLST and no contents and bad return code
Daniel Stenberg [Fri, 23 May 2003 11:10:35 +0000 (11:10 +0000)]
more ftp testing using NLST and no contents and bad return code

21 years agoRudy Koento's problem fixed, test case 66 verifies this.
Daniel Stenberg [Fri, 23 May 2003 09:47:57 +0000 (09:47 +0000)]
Rudy Koento's problem fixed, test case 66 verifies this.

21 years agotest 66 returns one line of data with no header (HTTP)
Daniel Stenberg [Fri, 23 May 2003 09:46:19 +0000 (09:46 +0000)]
test 66 returns one line of data with no header (HTTP)

21 years ago--digest added, --compressed rephrased
Daniel Stenberg [Fri, 23 May 2003 08:06:31 +0000 (08:06 +0000)]
--digest added, --compressed rephrased

21 years agoinclude digest.h for proto
Daniel Stenberg [Fri, 23 May 2003 06:44:24 +0000 (06:44 +0000)]
include digest.h for proto

21 years agoDJGPP fix by Gisle Vanem
Daniel Stenberg [Fri, 23 May 2003 06:43:14 +0000 (06:43 +0000)]
DJGPP fix by Gisle Vanem

21 years agomore more more
Daniel Stenberg [Thu, 22 May 2003 22:47:48 +0000 (22:47 +0000)]
more more more

21 years agowarning-free is better
Daniel Stenberg [Thu, 22 May 2003 22:45:38 +0000 (22:45 +0000)]
warning-free is better

21 years agoIntroducing --digest
Daniel Stenberg [Thu, 22 May 2003 22:40:01 +0000 (22:40 +0000)]
Introducing --digest

21 years agoBetter Digest stuff
Daniel Stenberg [Thu, 22 May 2003 22:39:38 +0000 (22:39 +0000)]
Better Digest stuff

21 years agoAdded CURLOPT_HTTPDIGEST support
Daniel Stenberg [Thu, 22 May 2003 22:38:46 +0000 (22:38 +0000)]
Added CURLOPT_HTTPDIGEST support
SOCKS5 fix as suggested by Jis in bugreport #741841.

21 years agoDocument the <dataNUM> thing we use,
Daniel Stenberg [Thu, 22 May 2003 22:37:00 +0000 (22:37 +0000)]
Document the <dataNUM> thing we use,

21 years agoDigest support added
Daniel Stenberg [Thu, 22 May 2003 22:36:39 +0000 (22:36 +0000)]
Digest support added

21 years agoDigest testing added
Daniel Stenberg [Thu, 22 May 2003 22:36:22 +0000 (22:36 +0000)]
Digest testing added

21 years agoproper header added
Daniel Stenberg [Thu, 22 May 2003 16:23:27 +0000 (16:23 +0000)]
proper header added

21 years agohush the compiler
Daniel Stenberg [Thu, 22 May 2003 16:12:30 +0000 (16:12 +0000)]
hush the compiler

21 years agoInitial Digest support. At least partly working.
Daniel Stenberg [Thu, 22 May 2003 16:09:54 +0000 (16:09 +0000)]
Initial Digest support. At least partly working.

21 years agoDavid Balazic pointed out the lack of checks for a valid %XX code when
Daniel Stenberg [Wed, 21 May 2003 15:53:59 +0000 (15:53 +0000)]
David Balazic pointed out the lack of checks for a valid %XX code when
we unescape a string. We now check and decode only valid %XX strings.

21 years agofix the makefile in packages/DOS too
Daniel Stenberg [Wed, 21 May 2003 08:12:52 +0000 (08:12 +0000)]
fix the makefile in packages/DOS too

21 years ago Gisle Vanem made curl build with djgpp on DOS.
Daniel Stenberg [Wed, 21 May 2003 08:08:48 +0000 (08:08 +0000)]
 Gisle Vanem made curl build with djgpp on DOS.

21 years agoGisle Vanem's fix to make the 'curl -M' output nicer
Daniel Stenberg [Wed, 21 May 2003 07:21:44 +0000 (07:21 +0000)]
Gisle Vanem's fix to make the 'curl -M' output nicer

21 years agomissing semicolon, by Gisle Vanem
Daniel Stenberg [Tue, 20 May 2003 12:44:55 +0000 (12:44 +0000)]
missing semicolon, by Gisle Vanem

21 years agoGisle Vanem's code for not trusting h_aliases to always be non-NULL
Daniel Stenberg [Tue, 20 May 2003 09:41:39 +0000 (09:41 +0000)]
Gisle Vanem's code for not trusting h_aliases to always be non-NULL

21 years agoRemind about the gpg command lines
Daniel Stenberg [Tue, 20 May 2003 06:33:13 +0000 (06:33 +0000)]
Remind about the gpg command lines

21 years agosupport user name and password in proxy environment variables
Daniel Stenberg [Mon, 19 May 2003 13:14:26 +0000 (13:14 +0000)]
support user name and password in proxy environment variables

21 years agothe proxy environment variables now may contain user name and password
Daniel Stenberg [Mon, 19 May 2003 13:09:41 +0000 (13:09 +0000)]
the proxy environment variables now may contain user name and password

21 years agoremove debug output
Daniel Stenberg [Mon, 19 May 2003 13:08:48 +0000 (13:08 +0000)]
remove debug output

21 years agosetenv support added to allow test cases to require a set of environment
Daniel Stenberg [Mon, 19 May 2003 13:06:10 +0000 (13:06 +0000)]
setenv support added to allow test cases to require a set of environment
variables

21 years ago7.10.5 commit
Daniel Stenberg [Mon, 19 May 2003 11:45:10 +0000 (11:45 +0000)]
7.10.5 commit

21 years agoknown AIX ipv6 problems
Daniel Stenberg [Fri, 16 May 2003 10:57:53 +0000 (10:57 +0000)]
known AIX ipv6 problems

21 years agoSkip any preceeding dots from the domain name of cookies when we keep them
Daniel Stenberg [Thu, 15 May 2003 22:28:19 +0000 (22:28 +0000)]
Skip any preceeding dots from the domain name of cookies when we keep them
in memory, only add it when we save the cookie. This makes all tailmatching
and domain string matching internally a lot easier.

This was also the reason for a remaining bug I introduced in my overhaul.

21 years agochange the order of the in_addr_t tests, so that 'unsigned long' is tested
Daniel Stenberg [Thu, 15 May 2003 21:13:36 +0000 (21:13 +0000)]
change the order of the in_addr_t tests, so that 'unsigned long' is tested
for first, as it seems to be what many systems use

21 years ago1. George Comninos' progress meter fix
Daniel Stenberg [Thu, 15 May 2003 08:13:19 +0000 (08:13 +0000)]
1. George Comninos' progress meter fix
2. I also added the pre-releases and dates to the log

21 years agodocumented CURLOPT_FTP_USE_EPRT
Daniel Stenberg [Wed, 14 May 2003 09:03:51 +0000 (09:03 +0000)]
documented CURLOPT_FTP_USE_EPRT

21 years agoGeorge Comninos provided a fix that calls the progress meter when waiting
Daniel Stenberg [Wed, 14 May 2003 06:31:00 +0000 (06:31 +0000)]
George Comninos provided a fix that calls the progress meter when waiting
for FTP command responses take >1 second.

21 years agoSetup and use CURL_INADDR_NONE all over instead of INADDR_NONE. We setup
Daniel Stenberg [Tue, 13 May 2003 12:12:17 +0000 (12:12 +0000)]
Setup and use CURL_INADDR_NONE all over instead of INADDR_NONE. We setup
the define accordingly in the hostip.h header to work nicely all over.

21 years agobefore using if2ip(), check if the address is an ip address and skip it if
Daniel Stenberg [Tue, 13 May 2003 12:11:31 +0000 (12:11 +0000)]
before using if2ip(), check if the address is an ip address and skip it if
it is.

21 years agolibtool 1.4.2 is enough
Daniel Stenberg [Tue, 13 May 2003 09:38:09 +0000 (09:38 +0000)]
libtool 1.4.2 is enough

21 years agofix comment
Daniel Stenberg [Tue, 13 May 2003 09:37:45 +0000 (09:37 +0000)]
fix comment

21 years agobefore checking for network interfaces using if2ip(), check that the given
Daniel Stenberg [Mon, 12 May 2003 13:06:48 +0000 (13:06 +0000)]
before checking for network interfaces using if2ip(), check that the given
name isn't an ip address

21 years agono more complaining when I have 1.5 and it tests for 1.4.2
Daniel Stenberg [Mon, 12 May 2003 13:05:11 +0000 (13:05 +0000)]
no more complaining when I have 1.5 and it tests for 1.4.2

21 years agofixes from the last week+
Daniel Stenberg [Mon, 12 May 2003 12:49:22 +0000 (12:49 +0000)]
fixes from the last week+

21 years agoDan F clarified the CURLOPT_ENCODING description after his changes to
Daniel Stenberg [Mon, 12 May 2003 12:47:35 +0000 (12:47 +0000)]
Dan F clarified the CURLOPT_ENCODING description after his changes to
allow "" to enable all support formats.

21 years agoDan Fandrich added --compressed docu
Daniel Stenberg [Mon, 12 May 2003 12:46:45 +0000 (12:46 +0000)]
Dan Fandrich added --compressed docu

21 years agosetting ENCODING to "" means enable-all-you-support
Daniel Stenberg [Mon, 12 May 2003 12:45:57 +0000 (12:45 +0000)]
setting ENCODING to "" means enable-all-you-support

21 years ago Dan Fandrich changed CURLOPT_ENCODING to select all supported encodings if
Daniel Stenberg [Mon, 12 May 2003 12:45:14 +0000 (12:45 +0000)]
 Dan Fandrich changed CURLOPT_ENCODING to select all supported encodings if
 set to "".  This frees the application from having to know which encodings
 the library supports.