Hi Thijs,

On 03/07/2013 04:14 PM, Thijs Kinkhorst wrote:
> Op donderdag 7 maart 2013 15:40:31 schreef Dario Minnucci:
>>> Fixing this would therefore require a bit of reworking of how
>>> simpleSAMLphp  tracks IdP's internally.
>>
>> I didn't look at the code enough to provide or propose a definitive and
>> elegant solution but I guess adding the reason why the IdP was not
>> considered to the backtrace presented on the browser could help.
> 
> So my point is that this is not known at that codepoint so the reason cannot 
> be added. SSP works like this: it reads at one point the set of IdP's 
> configured in saml20-idp-remote.php. At that time already, it will skip any 
> IdP that has expired (and logs a warning). After that, the set of IdP's 
> that's 
> used around the code doesn't include that IdP at all.
> 
> At the point where this error happens, the entityId is held against SSP's 
> internal list of IdP's, which doesn't contain that IdP because it was already 
> dropped way earlier. So we cannot really change that error because we do not 
> know at that point anything about the IdP.
> 
> Changing this would be good, but it would require to load all IdP's, keep 
> track of the expired status in the internal data structure, and check at each 
> point where it's used whether it's expired, if so, error out with specific 
> message, if not, continue. Possible, but not a matter of augmenting the error 
> message.
>  

ACK, thanks for the deep explanation.

Regards,

-- 
 Dario Minnucci <mid...@debian.org>
 Phone: +34 902884117 | Fax: +34 902024417 | Support: +34 807450000
 Key fingerprint = BAA1 7AAF B21D 6567 D457  D67D A82F BB83 F3D5 7033


Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to