that seems fine to me, but it should still report an warning to the
log. so that it's easier to see the auth URL was null.

peter

On 5/27/05, KiYun Roe <[EMAIL PROTECTED]> wrote:
> Hi,
> 
> Let me put it this way: the corresponding code in HTTPSampler checks for a
> null return from the auth manager and sets authorization only if the return
> is not null. In other words, authorization is optional for any given URL.
> I'd like to suggest that the web service sampler should follow the same
> pattern.
> 
> -- KiYun
> 
> -----Original Message-----
> From: Peter Lin [mailto:[EMAIL PROTECTED]
> Sent: Friday, May 27, 2005 12:26 PM
> To: JMeter Developers List
> Subject: Re: cvs commit:
> jakarta-jmeter/src/protocol/http/org/apache/jmeter/protocol/http/sampler
> WebServiceSampler.java
> 
> hmm, good point. I hadn't thought about that. Now I'm chaining my
> mind. Seems like defaulting to "/" might be confusing. it's probably
> safer to just log a warning.
> 
> peter
> 
> On 5/27/05, KiYun Roe <[EMAIL PROTECTED]> wrote:
> > Hi,
> >
> > I think the situation might arise when the auth manager is set at a higher
> > level and shared by several samplers, including one or more that don't
> > require security.
> >
> > -- KiYun
> >
> > -----Original Message-----
> > From: Peter Lin [mailto:[EMAIL PROTECTED]
> > Sent: Friday, May 27, 2005 12:10 PM
> > To: JMeter Developers List
> > Subject: Re: cvs commit:
> > jakarta-jmeter/src/protocol/http/org/apache/jmeter/protocol/http/sampler
> > WebServiceSampler.java
> >
> > that's true, but it doesn't make much sense to me to add an Auth
> > manager and then not provide an URL. I suppose I could check and
> > default to "/" in the event it's null, but there's no gaurantee the
> > login will work :)
> >
> > in either case, I should probably report the error. Webservices is
> > already dog slow, so it's not like a minor check will make much
> > difference for "slow as molasses services".
> >
> > peter
> >
> > On 5/27/05, KiYun Roe <[EMAIL PROTECTED]> wrote:
> > > Hi,
> > >
> > > It seems to me that AuthManager#getAuthForURL() can return null, so that
> > > this version of the patch is susceptible to null pointer exceptions if
> > there
> > > is no authorization for the current URL. Is that not the case?
> > >
> > > -- KiYun
> > >
> 
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 
>

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

Reply via email to