On Thu, 13 Dec 2007, Georg Lippitsch wrote:

> One more question: Do we need a seperate "CURLOPT_SEEKDATA"? Because the
> data passed to a seekfunction is usually to same that is passed to the read
> function (a stream in most cases). So, CURLOPT_READDATA would be passed to
> both, read function and seek function. IMO, there is no need for seperate
> seekdata.

I think we should have a separate CURLOPT_READDATA only for the reason that
all existing callbacks have such a one (all *FUNCTION options comes with a
*DATA companion) and it makes sense to stay consistent!

-- 
  Commercial curl and libcurl Technical Support: http://haxx.se/curl.html