On 4/22/05, Daniel Stenberg <daniel-curl_at_haxx.se> wrote:
> On Fri, 22 Apr 2005, RSN RSN wrote:
>
> > yes it does but --trace[-ascii] has a lot of suplufous data and it is in a
> > seperate file.
>
> Yes it has "suplufous data", but that's in addition to the data you want. And
> the extra fluff is fairly easy to filter out I would say.
>
> > The point is that it is that all headers are required to be in the same
> > file. Why split the headers into seperate files? This currently requires to
> > have both files open at the same time to read and cross-refernce everything.
> >
> > Why not have requests and reply and the data in the same file?
>
> --trace makes one single file. It contains the lot. The only difference your
> suggested -i or -D changes would make, is that it wouldn't also contain the
> "info" texts like the --trace output does.
>
> --
> Commercial curl and libcurl Technical Support: http://haxx.se/curl.html
>

It states that the -i and -D provide the headers yet they only provide
the replys. This is only half the entire process. Why not complete the
function or rename the functions to say that they only display HTTP
replys. The question of why the headers have been split has not been
answered. The people at Ethereal found it neccessary to include this
function.
Received on 2005-04-22