[undertow-dev] When DIGEST support?

Anil Saldhana Anil.Saldhana at redhat.com
Tue Jun 11 15:48:45 EDT 2013


Jason, https://issues.jboss.org/browse/PLINK-190

On 06/11/2013 02:33 PM, Jason Greene wrote:
> On Jun 11, 2013, at 2:28 PM, Jason Greene <jason.greene at redhat.com> wrote:
>
>> On Jun 11, 2013, at 2:20 PM, Anil Saldhana <Anil.Saldhana at redhat.com> wrote:
>>
>>> On 06/11/2013 02:17 PM, Bill Burke wrote:
>>>> On 6/11/2013 3:07 PM, Jason Greene wrote:
>>>>>> On Jun 11, 2013, at 12:26 PM, Anil Saldhana<Anil.Saldhana at redhat.com>  wrote:
>>>>>>
>>>>>>>> On 06/11/2013 07:58 AM, Anil Saldhana wrote:
>>>>>>>>>> On Jun 11, 2013, at 7:23 AM, Jason Greene<jgreene at redhat.com>  wrote:
>>>>>>>>>>
>>>>>>>>>>>> Hmm I guess that means we need a picketlink IDM subsystem?
>>>>>>>>>> We have a subsystem ready.
>>>>>>>> https://docs.jboss.org/author/display/PLINK/PicketLink+Subsystem
>>>>>>>> Feedback welcome.
>>>>>> I like the typed syntax, but I do not like the fact this is using JPA over JDBC. It's going to create a dependency mess (now all wildfly distributions that want to use database auth will require Java EE), and it will slow server initialization for very little gain.
>>>>>>
>>>> They have a file store.
>>> The Identity Store implementations can be:
>>> a) database via jpa
>>> b) file system  (created exclusively for Wildfly)
>>> c) ldap
>>> d) mix and match
>>>
>>> Regarding Jason's concern about database authentication requiring EE api
>>> (due to JPA) slowing server startup, that is going to be tough as the
>>> alternative would be straight JDBC.
>>
>> Why is that a problem though? JDBC isn't that hard, all vendors these days support portable ANSI 92 sql, and the complexity of queries that an IDM needs to execute is small.
>>
> I guess I should be having this discussion on the security dev list :)
>
> -Jason


More information about the undertow-dev mailing list