On Wed, 2008-01-30 at 12:37 -0800, David Byrne wrote:
> Julius,
> 
> Thanks for replying. I saw AllowAllHostnameVerifier, but decided to write my 
> own so I could report on the errors. However, this only solves part of my 
> problem. Even if the verifier passes the cert along, the TrustManager used by 
> SSLSocket will still throw an exception. I'm writing a null logic 
> TrustManager for that.
> 

The host verification process can take place only after an SSL session
has been successfully established. So, the trust manager needs to verify
the certificate chain first.

Oleg


> Note that I might be way off on this. I'm still trying to wrap myself around 
> Java's SSL implementation.
> 
> Thanks,
> David
> 
> 
> ----- Original Message ----
> From: Julius Davies <[EMAIL PROTECTED]>
> To: HttpClient User Discussion <[email protected]>
> Sent: Wednesday, January 30, 2008 1:16:26 PM
> Subject: Re: AbstractVerifier in 4.x (was Invalid SSL Certs)
> 
> 
> Hi, 
> David,
> 
> Hmmm... 
> and 
> there's 
> also 
> an 
> AllowAllHostnameVerifier 
> already 
> in 
> there.
>  
> But 
> it's 
> also 
> lacking 
> a 
> public 
> constructor, 
> so 
> it's 
> no 
> use 
> to 
> you 
> at
> the 
> moment.
> 
> People 
> are 
> encouraged 
> to 
> create 
> new 
> JIRA 
> tickets 
> with 
> patches 
> attached!  
> ;-)
> 
> 
> yours,
> 
> Julius
> 
> 
> On 
> Jan 
> 30, 
> 2008 
> 9:59 
> AM, 
> David 
> Byrne 
> <[EMAIL PROTECTED]> 
> wrote:
> > 
> A 
> quick 
> follow-up 
> on 
> AbstractVerifier. 
> Is 
> there 
> a 
> reason 
> that 
> the 
> constructor 
> isn't 
> set 
> to 
> public? 
> It 
> makes 
> it 
> difficult 
> to 
> extend 
> outside 
> of 
> the 
> package.
> >
> > 
> Thanks,
> > 
> David 
> Byrne
> >
> >
> 


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to