platform/upstream/curl.git
20 years agoRalph Mitchell's update that allows this script to take an existing
Daniel Stenberg [Fri, 7 Nov 2003 08:06:03 +0000 (08:06 +0000)]
Ralph Mitchell's update that allows this script to take an existing
directory name and build/run/test curl in there instead of trying to update
from CVS. Using this approach, this script can be used to test daily tarballs
etc.

20 years agothe resource file itself
Daniel Stenberg [Fri, 7 Nov 2003 07:20:15 +0000 (07:20 +0000)]
the resource file itself

20 years agoremoved the resource lines I shouldn't have added in the firt place
Daniel Stenberg [Fri, 7 Nov 2003 07:13:56 +0000 (07:13 +0000)]
removed the resource lines I shouldn't have added in the firt place

20 years agodefine TRUE and FALSE here as some platforms don't have them otherwise
Daniel Stenberg [Fri, 7 Nov 2003 07:02:35 +0000 (07:02 +0000)]
define TRUE and FALSE here as some platforms don't have them otherwise

20 years agoCURLDEBUG, no underscore
Daniel Stenberg [Thu, 6 Nov 2003 16:04:19 +0000 (16:04 +0000)]
CURLDEBUG, no underscore

20 years agoGisle Vanem's libcurl.dll version info patch applied
Daniel Stenberg [Thu, 6 Nov 2003 15:09:30 +0000 (15:09 +0000)]
Gisle Vanem's libcurl.dll version info patch applied

20 years agodon't allow options that we KNOW can't be used
Daniel Stenberg [Thu, 6 Nov 2003 14:16:16 +0000 (14:16 +0000)]
don't allow options that we KNOW can't be used

20 years agomark all mentioned options better
Daniel Stenberg [Thu, 6 Nov 2003 13:51:29 +0000 (13:51 +0000)]
mark all mentioned options better

20 years agoconverted most .I lines to \fI ones
Daniel Stenberg [Thu, 6 Nov 2003 13:34:28 +0000 (13:34 +0000)]
converted most .I lines to \fI ones

20 years agoegd, not edg
Daniel Stenberg [Thu, 6 Nov 2003 13:31:42 +0000 (13:31 +0000)]
egd, not edg

20 years agoDefine TRUE and FALSE unless already present. I've moved away those
Daniel Stenberg [Thu, 6 Nov 2003 13:07:54 +0000 (13:07 +0000)]
Define TRUE and FALSE unless already present. I've moved away those
definitions from the global curl header and thus this needs to do it
themselves.

20 years agoremoved the _last since that is by definition not an error code
Daniel Stenberg [Thu, 6 Nov 2003 12:52:48 +0000 (12:52 +0000)]
removed the _last since that is by definition not an error code

20 years agonew defines for version info
Daniel Stenberg [Thu, 6 Nov 2003 11:37:54 +0000 (11:37 +0000)]
new defines for version info

20 years agoInsert the major, minor and patch numbers into the curl/curl.h header as
Daniel Stenberg [Thu, 6 Nov 2003 11:34:19 +0000 (11:34 +0000)]
Insert the major, minor and patch numbers into the curl/curl.h header as
well. Removed the depedency on perl.

20 years agoWe offer the version number "in parts" as well by introducing three new
Daniel Stenberg [Thu, 6 Nov 2003 11:33:08 +0000 (11:33 +0000)]
We offer the version number "in parts" as well by introducing three new
defines.

20 years agothings
Daniel Stenberg [Thu, 6 Nov 2003 08:35:01 +0000 (08:35 +0000)]
things

20 years agonew mirror added
Daniel Stenberg [Thu, 6 Nov 2003 08:32:43 +0000 (08:32 +0000)]
new mirror added

20 years agoupdated somewhat
Daniel Stenberg [Thu, 6 Nov 2003 08:15:04 +0000 (08:15 +0000)]
updated somewhat

20 years agounitialized variable fix, reported by both Marty Kuhrt and benjamin gerard
Daniel Stenberg [Thu, 6 Nov 2003 07:55:45 +0000 (07:55 +0000)]
unitialized variable fix, reported by both Marty Kuhrt and benjamin gerard

20 years agoMac OS X framework build files, contributed by Matt Veenstra
Daniel Stenberg [Thu, 6 Nov 2003 07:42:02 +0000 (07:42 +0000)]
Mac OS X framework build files, contributed by Matt Veenstra

20 years agoMoved the TRUE and FALSE #defines to lib/setup.h instead, as they don't
Daniel Stenberg [Thu, 6 Nov 2003 07:21:32 +0000 (07:21 +0000)]
Moved the TRUE and FALSE #defines to lib/setup.h instead, as they don't
belong in the public header file. They are not in our name space so we
should not set them globally.

20 years agominor format edits
Daniel Stenberg [Wed, 5 Nov 2003 15:52:00 +0000 (15:52 +0000)]
minor format edits

20 years agouse roffit to make HTML from man pages
Daniel Stenberg [Wed, 5 Nov 2003 15:51:25 +0000 (15:51 +0000)]
use roffit to make HTML from man pages
http://daniel.haxx.se/projects/roffit/

20 years agouse .BR to refer to other man pages to make it easier for parsers to know
Daniel Stenberg [Tue, 4 Nov 2003 13:30:27 +0000 (13:30 +0000)]
use .BR to refer to other man pages to make it easier for parsers to know

20 years agoconsistant with other man pages
Daniel Stenberg [Tue, 4 Nov 2003 13:27:28 +0000 (13:27 +0000)]
consistant with other man pages

20 years ago.TP and .B replaced with .IP
Daniel Stenberg [Tue, 4 Nov 2003 13:03:45 +0000 (13:03 +0000)]
.TP and .B replaced with .IP

20 years ago.IP instead of .TP and .B
Daniel Stenberg [Tue, 4 Nov 2003 12:59:10 +0000 (12:59 +0000)]
.IP instead of .TP and .B

20 years agohighlight more mentioned options
Daniel Stenberg [Tue, 4 Nov 2003 12:58:03 +0000 (12:58 +0000)]
highlight more mentioned options

20 years agoUse .IP instead of .TP and .B for each item.
Daniel Stenberg [Tue, 4 Nov 2003 12:55:53 +0000 (12:55 +0000)]
Use .IP instead of .TP and .B for each item.

20 years agospell!
Daniel Stenberg [Tue, 4 Nov 2003 12:07:19 +0000 (12:07 +0000)]
spell!

20 years agothree changes went in
Daniel Stenberg [Tue, 4 Nov 2003 07:24:56 +0000 (07:24 +0000)]
three changes went in

20 years agoTroels Walsted Hansen fixed the MSVC makefiles to build fine on Windows.
Daniel Stenberg [Tue, 4 Nov 2003 06:59:25 +0000 (06:59 +0000)]
Troels Walsted Hansen fixed the MSVC makefiles to build fine on Windows.

20 years agospellfix in the generated comment by Kevin Roth
Daniel Stenberg [Tue, 4 Nov 2003 06:52:26 +0000 (06:52 +0000)]
spellfix in the generated comment by Kevin Roth

20 years agoKevin Roth's fix to make correct cygwin packages
Daniel Stenberg [Tue, 4 Nov 2003 06:50:50 +0000 (06:50 +0000)]
Kevin Roth's fix to make correct cygwin packages

20 years agoblanked, starting all over again
Daniel Stenberg [Sat, 1 Nov 2003 11:40:07 +0000 (11:40 +0000)]
blanked, starting all over again

20 years agostarting on 7.10.9 in CVS
Daniel Stenberg [Sat, 1 Nov 2003 11:33:51 +0000 (11:33 +0000)]
starting on 7.10.9 in CVS

20 years ago7.10.8!
Daniel Stenberg [Sat, 1 Nov 2003 11:22:48 +0000 (11:22 +0000)]
7.10.8!

20 years agoftp goes UTC
Daniel Stenberg [Fri, 31 Oct 2003 22:17:13 +0000 (22:17 +0000)]
ftp goes UTC

20 years agogmtime() really can't even return NULL, can it?
Daniel Stenberg [Fri, 31 Oct 2003 21:43:22 +0000 (21:43 +0000)]
gmtime() really can't even return NULL, can it?

20 years agoback to GMT again
Daniel Stenberg [Fri, 31 Oct 2003 21:37:03 +0000 (21:37 +0000)]
back to GMT again

20 years agoassume that MDTM returns the timestamp as UTC
Daniel Stenberg [Fri, 31 Oct 2003 21:36:43 +0000 (21:36 +0000)]
assume that MDTM returns the timestamp as UTC

20 years agomy own system had a bad valgrind, so check that it works at least somewhat
Daniel Stenberg [Fri, 31 Oct 2003 21:34:39 +0000 (21:34 +0000)]
my own system had a bad valgrind, so check that it works at least somewhat
before we use it

20 years agoDebian bug report #218046, a minor typo.
Daniel Stenberg [Fri, 31 Oct 2003 18:43:12 +0000 (18:43 +0000)]
Debian bug report #218046, a minor typo.

20 years agoremoved compiler warnings from my latest fix
Daniel Stenberg [Fri, 31 Oct 2003 09:49:10 +0000 (09:49 +0000)]
removed compiler warnings from my latest fix

20 years agonumerical options check
Daniel Stenberg [Fri, 31 Oct 2003 09:27:15 +0000 (09:27 +0000)]
numerical options check

20 years agoAdded a new parse-numeric-parameters function so that options that take
Daniel Stenberg [Fri, 31 Oct 2003 09:22:25 +0000 (09:22 +0000)]
Added a new parse-numeric-parameters function so that options that take
a numeric argument can better bail out if given a weird parameter. This
catches situations such as "-y -Y 2000" or "-O -C [URL]" etc.

20 years agonewer, nicer
Daniel Stenberg [Fri, 31 Oct 2003 08:08:21 +0000 (08:08 +0000)]
newer, nicer

20 years agoSet version string to -CVS. This string is generated by maketgz at release-
Daniel Stenberg [Thu, 30 Oct 2003 13:30:21 +0000 (13:30 +0000)]
Set version string to -CVS. This string is generated by maketgz at release-
time.

20 years agoMake the CVS version clone the libcurl version string, this is generated
Daniel Stenberg [Thu, 30 Oct 2003 13:28:58 +0000 (13:28 +0000)]
Make the CVS version clone the libcurl version string, this is generated
by the maketgz script so in releases this is replaced by an actual string.

20 years agoignore .dist files here
Daniel Stenberg [Thu, 30 Oct 2003 13:07:36 +0000 (13:07 +0000)]
ignore .dist files here

20 years agoMade the dist-hook more generic for adding more files named .dist into the
Daniel Stenberg [Thu, 30 Oct 2003 12:48:19 +0000 (12:48 +0000)]
Made the dist-hook more generic for adding more files named .dist into the
dist archive. Preparing for a new maketgz.

20 years ago--head on file://, some cleanups
Daniel Stenberg [Thu, 30 Oct 2003 10:30:41 +0000 (10:30 +0000)]
--head on file://, some cleanups

20 years agomore help text cleanups
Daniel Stenberg [Thu, 30 Oct 2003 09:15:47 +0000 (09:15 +0000)]
more help text cleanups

20 years agosilly me, I was meaning to do this change already as discussed on the libcurl
Daniel Stenberg [Thu, 30 Oct 2003 09:13:04 +0000 (09:13 +0000)]
silly me, I was meaning to do this change already as discussed on the libcurl
list, we get the time in GMT and not localtime

20 years agocurl --head now reports info "headers" on file:// URLs as well
Daniel Stenberg [Thu, 30 Oct 2003 09:08:16 +0000 (09:08 +0000)]
curl --head now reports info "headers" on file:// URLs as well

20 years agoupdated for the cut off GMT string
Daniel Stenberg [Thu, 30 Oct 2003 09:06:52 +0000 (09:06 +0000)]
updated for the cut off GMT string

20 years agowhen getting headers only, stop pretending the reported time is GMT, as it
Daniel Stenberg [Thu, 30 Oct 2003 09:06:24 +0000 (09:06 +0000)]
when getting headers only, stop pretending the reported time is GMT, as it
isn't. In fact, it might be, but we have no idea.

20 years agoonly use siginterrupt() if it really exists on the platform
Daniel Stenberg [Thu, 30 Oct 2003 07:32:04 +0000 (07:32 +0000)]
only use siginterrupt() if it really exists on the platform

20 years agoCheck for siginterrupt, used by the test suite web server sws.
Daniel Stenberg [Thu, 30 Oct 2003 07:31:02 +0000 (07:31 +0000)]
Check for siginterrupt, used by the test suite web server sws.

20 years agoOnly compile the ipv6-section for ipv6-enabled libcurls. Should save
Daniel Stenberg [Thu, 30 Oct 2003 07:13:13 +0000 (07:13 +0000)]
Only compile the ipv6-section for ipv6-enabled libcurls. Should save
us some trouble.

20 years agoAdded proper #ifdef's for the #include files when checking for in_addr_t,
Daniel Stenberg [Thu, 30 Oct 2003 07:08:45 +0000 (07:08 +0000)]
Added proper #ifdef's for the #include files when checking for in_addr_t,
which made the test start working nicely on BeOS. Shard verified it.

20 years agoStop using stunnel.pm, we pass in the path from the main script instead.
Daniel Stenberg [Wed, 29 Oct 2003 16:27:43 +0000 (16:27 +0000)]
Stop using stunnel.pm, we pass in the path from the main script instead.
Also made it easier to stop the test suite with control-c.

20 years agocleaned up the --help output. Tried to unify the language. Shortened a bunch
Daniel Stenberg [Wed, 29 Oct 2003 16:11:36 +0000 (16:11 +0000)]
cleaned up the --help output. Tried to unify the language. Shortened a bunch
of explanations.

20 years agoDavid Hull made the file: URL parser also accept the somewhat sloppy file
Daniel Stenberg [Wed, 29 Oct 2003 09:53:21 +0000 (09:53 +0000)]
David Hull made the file: URL parser also accept the somewhat sloppy file
syntax: file:/path. I added test case 203 to verify this.

20 years agostunnel.pm is no more
Daniel Stenberg [Wed, 29 Oct 2003 09:13:53 +0000 (09:13 +0000)]
stunnel.pm is no more

20 years agostunnel.pm is removed
Daniel Stenberg [Wed, 29 Oct 2003 09:13:36 +0000 (09:13 +0000)]
stunnel.pm is removed

20 years agonow check for and use valgrind, inserted and modified the check for stunnel
Daniel Stenberg [Wed, 29 Oct 2003 09:12:47 +0000 (09:12 +0000)]
now check for and use valgrind, inserted and modified the check for stunnel
and we no longer use the stunnel.pm

20 years agoremoved, functionality added to runtests.pl
Daniel Stenberg [Wed, 29 Oct 2003 09:12:09 +0000 (09:12 +0000)]
removed, functionality added to runtests.pl

20 years agoAnother glibc resolve name fix
Daniel Stenberg [Tue, 28 Oct 2003 13:06:15 +0000 (13:06 +0000)]
Another glibc resolve name fix

20 years agopossibly uninitialized variable
Daniel Stenberg [Tue, 28 Oct 2003 09:28:11 +0000 (09:28 +0000)]
possibly uninitialized variable

20 years agooverlapping memory chunks with strcpy(), detected by the friendly valgrind
Daniel Stenberg [Tue, 28 Oct 2003 09:17:15 +0000 (09:17 +0000)]
overlapping memory chunks with strcpy(), detected by the friendly valgrind

20 years agoslightly quieter torture test by default
Daniel Stenberg [Mon, 27 Oct 2003 06:40:48 +0000 (06:40 +0000)]
slightly quieter torture test by default

20 years agoJames Bursa's fix to prevent failf() to write outside its buffer boundary
Daniel Stenberg [Mon, 27 Oct 2003 06:35:47 +0000 (06:35 +0000)]
James Bursa's fix to prevent failf() to write outside its buffer boundary

20 years agoJames Bursa found an ERRORBUFFFER overflow
Daniel Stenberg [Sun, 26 Oct 2003 15:42:21 +0000 (15:42 +0000)]
James Bursa found an ERRORBUFFFER overflow

20 years agosnprintf() made a single-byte buffer overflow, as it could write a zero
Daniel Stenberg [Sun, 26 Oct 2003 15:37:45 +0000 (15:37 +0000)]
snprintf() made a single-byte buffer overflow, as it could write a zero
outside its given buffer. Discovered and reported by James Bursa.

20 years agoavoid dependence on the order of the fields in the timeval struct, pointed out
Daniel Stenberg [Sat, 25 Oct 2003 06:03:50 +0000 (06:03 +0000)]
avoid dependence on the order of the fields in the timeval struct, pointed out
by Gisle Vanem

20 years agosupport building ares-enabled too
Daniel Stenberg [Fri, 24 Oct 2003 22:30:40 +0000 (22:30 +0000)]
support building ares-enabled too

20 years agotests the new -C - fix
Daniel Stenberg [Fri, 24 Oct 2003 21:58:43 +0000 (21:58 +0000)]
tests the new -C - fix

20 years agoHTTP resume fix
Daniel Stenberg [Fri, 24 Oct 2003 21:58:11 +0000 (21:58 +0000)]
HTTP resume fix

20 years agoResuming a download of an already downloaded document, that is trying to get
Daniel Stenberg [Fri, 24 Oct 2003 21:54:34 +0000 (21:54 +0000)]
Resuming a download of an already downloaded document, that is trying to get
a range of a document beyond its size, caused libcurl to "hang" until the
server closed the connection and then it returned error 18.

This is bad. This way, we don't return any error at all, which isn't nice
either, as we need to alert the app somehow that the request range was out
of size.

20 years agosolaris build improvement
Daniel Stenberg [Fri, 24 Oct 2003 21:27:22 +0000 (21:27 +0000)]
solaris build improvement

20 years agoBased on David Hull's fix in bug report 804599, we now check for solaris
Daniel Stenberg [Fri, 24 Oct 2003 21:26:04 +0000 (21:26 +0000)]
Based on David Hull's fix in bug report 804599, we now check for solaris
and gcc, to set the -mimpure-text link flag for linking the lib better.

20 years agodon't select() forever, set a timeout so at least the test fails nice
Daniel Stenberg [Fri, 24 Oct 2003 20:58:34 +0000 (20:58 +0000)]
don't select() forever, set a timeout so at least the test fails nice

20 years agobetter name, it isn't https
Daniel Stenberg [Fri, 24 Oct 2003 20:39:39 +0000 (20:39 +0000)]
better name, it isn't https

20 years agoc-ares is the new name and ares_version() was added
Daniel Stenberg [Fri, 24 Oct 2003 20:30:14 +0000 (20:30 +0000)]
c-ares is the new name and ares_version() was added

20 years agoIntroducing ares_version(), so that we can have apps get version info about
Daniel Stenberg [Fri, 24 Oct 2003 20:28:04 +0000 (20:28 +0000)]
Introducing ares_version(), so that we can have apps get version info about
what particular ares version that is being used.

20 years agomake it work with Negotiate support enabled too
Daniel Stenberg [Fri, 24 Oct 2003 20:21:01 +0000 (20:21 +0000)]
make it work with Negotiate support enabled too

20 years agoruntests -t introduced
Daniel Stenberg [Fri, 24 Oct 2003 13:12:58 +0000 (13:12 +0000)]
runtests -t introduced

20 years agoimproved the torture testing
Daniel Stenberg [Fri, 24 Oct 2003 13:00:19 +0000 (13:00 +0000)]
improved the torture testing

20 years agosupports -l now to display the actual line that was prevented by memlimit
Daniel Stenberg [Fri, 24 Oct 2003 12:59:58 +0000 (12:59 +0000)]
supports -l now to display the actual line that was prevented by memlimit

20 years agoif malloc fails, we must bail out nicely
Daniel Stenberg [Fri, 24 Oct 2003 12:58:59 +0000 (12:58 +0000)]
if malloc fails, we must bail out nicely

20 years agoImproved checks for bad memory situations and proper behaviour when we
Daniel Stenberg [Fri, 24 Oct 2003 12:58:29 +0000 (12:58 +0000)]
Improved checks for bad memory situations and proper behaviour when we
run out of memory.

20 years agobetter bailing-out cleanup if a malloc fails in the DNS cache
Daniel Stenberg [Fri, 24 Oct 2003 12:57:23 +0000 (12:57 +0000)]
better bailing-out cleanup if a malloc fails in the DNS cache

20 years agofree all memory on failure before bailing out, not really necessary but
Daniel Stenberg [Fri, 24 Oct 2003 12:56:27 +0000 (12:56 +0000)]
free all memory on failure before bailing out, not really necessary but
my upcoming automated test gets crazy if not

20 years agothe actual release-notes is new too
Daniel Stenberg [Fri, 24 Oct 2003 08:56:08 +0000 (08:56 +0000)]
the actual release-notes is new too

20 years agoIntroducing -t to "torture" the memory allocations/failing/bail-outing in
Daniel Stenberg [Fri, 24 Oct 2003 08:53:59 +0000 (08:53 +0000)]
Introducing -t to "torture" the memory allocations/failing/bail-outing in
curl and libcurl. -t is not used anywhere automated yet, and it does already
identify memory leaks on failed allocations. Work to do.

20 years agoproduce a single summary of the amount of allocations on -v
Daniel Stenberg [Fri, 24 Oct 2003 08:09:33 +0000 (08:09 +0000)]
produce a single summary of the amount of allocations on -v

20 years agodisable-eprt, fixed CA cert verfication
Daniel Stenberg [Fri, 24 Oct 2003 06:12:39 +0000 (06:12 +0000)]
disable-eprt, fixed CA cert verfication

20 years agoGeorg Horn's patch. -k is no longer mutually exclusive with --cacert and
Daniel Stenberg [Thu, 23 Oct 2003 07:46:22 +0000 (07:46 +0000)]
Georg Horn's patch. -k is no longer mutually exclusive with --cacert and
--capath. Using -k together with one of those just means that the result
of the CA cert check is ignored (but displayed if -v is used).

20 years agoGeorg Horn's fixes to do different CA cert verifications. They can now be
Daniel Stenberg [Thu, 23 Oct 2003 07:44:55 +0000 (07:44 +0000)]
Georg Horn's fixes to do different CA cert verifications. They can now be
done even if the result is ignored, as some sites seem to require that.