Random_the_seed() takes about 25-30 seconds to complete and perhaps this
contributes to the SSL handshake failing. Should it take this long?

The reason I suspect this might be part of the problem is because only when
the call to random_the_seed() is made does the SSL transport fail. All
subsequent attempts pass and random_the_seed() isn't called.

With respect to the details - when random_the_seed() is called, the error
that is received in 10053 - "Software caused connection abort".

Just wondering if others have experienced this and if there is a way to
speed up random_the_seed().

Thanks,

-----Original Message-----
From: Daniel Stenberg [mailto:daniel_at_haxx.se]
Sent: Friday, October 03, 2003 8:53 AM
To: Curl Mailinglist
Subject: Re: SSL transport failure during handshake

On Wed, 1 Oct 2003, Wiebe, Kevin wrote:

> The first time I open my application and attempt to transport over SSL,
the
> transport fails. Each subsequent attempt succeeds. I noticed the first
> attempt calls random_the_seed(),but each subsequent one doesn't.

Right, we consider seeding the random seed once to be enough.

> Perhaps this is part of the problem.

What makes you suspect this?

> Has anyone else experienced a similar issue, and if so, any solutions?

You're not giving us much details, so it is next to impossible to even guess
what you're experiencing.

-- 
 Daniel Stenberg -- curl: been grokking URLs since 1998
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf