On Wed, Jul 23, 2008 at 11:01:21PM +0200, Daniel Stenberg wrote:
> On Mon, 14 Jul 2008, Dan Fandrich wrote:
>> I've added test cases 1040 and test1041 which test -C - on HTTP, and
>> test1041 looks like it shows up a problem which might be related to
>> yours. It generates the header:
>>
>> Content-Range: bytes 18446744073709551615-/100
>>
>> where that first integer just happens to be the signed 64-bit value -1
>> but displayed as unsigned. Something is definitely fishy somewhere.
>
> Isn't this pretty exactly what KNOWN_BUGS #8 is about?

KNOWN_BUG #8 explains why the results aren't optimal (i.e. the already-
uploaded portion of the file can't be skipped), but I would expect curl
to do something sane in this case, namely send a header of:

  Content-Range: bytes 0-99/100

and unconditionally upload the entire contents of the file.

>>> Dan

-- 
http://www.MoveAnnouncer.com              The web change of address service
          Let webmasters know that your web site has moved
-------------------------------------------------------------------
List admin: http://cool.haxx.se/cgi-bin/mailman/listinfo/curl-users
FAQ:        http://curl.haxx.se/docs/faq.html
Etiquette:  http://curl.haxx.se/mail/etiquette.html