On Sat, 25 May 2013, Venkatesh Prabu Narayanan wrote:

Please stop trying to act stupid and just continue to top-post while I
complain on everyone else on this list who does it. Okay?

> Further to my previous post, I have tried returning the option
> 'CURL_READFUNC_ABORT' and now the operation aborted as expected. Now I have
> two questions,
>
> 1. Is it safe to set large content length as in my case ?. Does it affect
> any other behavior ?

It is safe from our (libcurl's client-side) aspect. If it is safe in your use
of it with your server and environment? It's not possible for us to know for
sure.

> 2. After sending last block of data (less than 16K), server sends a
> handshake protocol (only 7 bytes in size) to the client. This protocol
> wasn't received in the Write Call back function (you can see this in my
> first post in this thread).

But you returned CURL_READFUNC_ABORT, right? That means you ask the transfer
to abort. Or when was that 7 bytes supposed to arrive? I think I'm not fully
understanding what you're doing.

> Also is there any minimum size need to be filled in the socket so as to call
> the write call back function ?

Nope.

-- 
  / daniel.haxx.se
-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-library
Etiquette:  http://curl.haxx.se/mail/etiquette.html