Age | Commit message (Collapse) | Author |
|
|
|
and we can't depend on configure-defines in it. This logic is borrowed from
the public curl headers.
|
|
|
|
|
|
|
|
|
|
|
|
release
|
|
|
|
|
|
|
|
|
|
that this file is out of date!
|
|
|
|
|
|
operation!
|
|
|
|
|
|
cleared up some ssize_t/size_t mixups
|
|
|
|
|
|
previously.
|
|
|
|
|
|
to warn
|
|
they may not be of the size size
|
|
pointer arithmetic (ptr1 - ptr2) and to do that properly on 64bit we need
long
|
|
|
|
|
|
revision number
|
|
refuse to remove the dir otherwise!
|
|
on Windows.
|
|
|
|
|
|
newer versions
|
|
|
|
limit-rate. It does work on the expense of the rate limiter.
|
|
int to various types
|
|
compilers
|
|
This works by magicly starting up a new thread that can be killed when the
timeout is reached. testtesttest!
|
|
|
|
one to prevent that.
|
|
|
|
just to avoid making a totally empty file. Just to avoid compiler warnings.
|
|
and I doubt anyone else uses it.
Better preprocessor magic for the O flag (for curl_off_t printing) to prevent
compiler warnings.
|
|
|
|
It is working with earlier versions, but when I use it with 3.0.3, I can't
get it to ignore errors in "system headers" with -isystem so we get
excessive amounts of warnings on SSL headers which is very annoying.
|
|
without proto
|
|
it collides with another MAX define on some platforms (like netbsd 1.6.1)
|
|
|