On Fri, 20 Jan 2006, Jason Sachs wrote:

> A stumper... there's something weird with one of the USDA webpages, Firefox
> sees it fine but curl ends prematurely somewhere in the middle.

Have you checked and seen that Firefox indeed get the exact same data? If so,
do you have any theory how it can decide that the what-seems-to-be-the-final
chunk isn't actually the final one?

A little work-around to get the data without chunks would be to make the
request using HTTP 1.0, with the -0 / --http1.0 option.

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