Sounds good!

On Mon, Feb 8, 2010 at 12:59 PM, Kalle Korhonen
<[email protected]> wrote:
> Yeah we should, that's what I implied in the PS. I can take a look -
> I'll open another jira for it.
>
> Kalle
>
>
> On Mon, Feb 8, 2010 at 9:46 AM, Les Hazlewood <[email protected]> wrote:
>> I'm wondering - shouldn't we do this for all of our support modules then?
>>
>> On Mon, Feb 8, 2010 at 12:32 PM, Tauren Mills <[email protected]> wrote:
>>> Kalle,
>>> I see you already committed this change!  Thanks for the prompt action.  I
>>> agree that very few situations would not specify Spring dependency somewhere
>>> else.
>>> Tauren
>>>
>>> On Mon, Feb 8, 2010 at 8:41 AM, Les Hazlewood <[email protected]> wrote:
>>>>
>>>> Hi Kalle,
>>>>
>>>> +1
>>>>
>>>> I think this is a good idea - I must be "having a case of the Mondays"
>>>> ;).  Yes, almost everyone who uses shiro-spring integration will most
>>>> certainly specify their Spring dependency elsewhere.
>>>>
>>>> Cheers,
>>>>
>>>> Les
>>>>
>>>> On Mon, Feb 8, 2010 at 11:30 AM, Kalle Korhonen
>>>> <[email protected]> wrote:
>>>> > Yea, the problem originates from us depending on spring-all but many
>>>> > users likely depending on only those specific Spring libs they need,
>>>> > in which case artifact ids don't match and you end with duplicate libs
>>>> > in your classpath. We should mark Spring as scope provided, presumably
>>>> > people don't use shiro-spring as the end dependency to get Spring. If
>>>> > I don't hear otherwise, I'll create a JIRA and fix it.
>>>> >
>>>> > Kalle
>>>> >
>>>> > PS. Les, for commons-logging the situation is much the same. The
>>>> > proper way is to lobby the projects to mark common-logging as
>>>> > provided, but it's a more difficult problem to solve since there may
>>>> > not be any other end dependency for commons logging unless you
>>>> > provided one (either slf or commons-logging). We should evaluate all
>>>> > of our third-party dependencies and mark them as provided if at all
>>>> > possible - it's more difficult to go from compiled to provided than
>>>> > the other way around.
>>>> >
>>>> >
>>>> > On Mon, Feb 8, 2010 at 8:03 AM, Les Hazlewood <[email protected]>
>>>> > wrote:
>>>> >> Hi Tauren,
>>>> >>
>>>> >> Yep, this is pretty customary in Maven poms as I understand it, but
>>>> >> I'd love to hear if there is a better way.  I have to do this
>>>> >> regularly for any dependency that pulls in commons-logging - I have to
>>>> >> manually exclude it since I use SLF4J's version instead.
>>>> >>
>>>> >> If there's a better way, I'm all ears!
>>>> >>
>>>> >> Cheers,
>>>> >>
>>>> >> Les
>>>> >>
>>>> >> On Mon, Feb 8, 2010 at 7:01 AM, Tauren Mills <[email protected]>
>>>> >> wrote:
>>>> >>> Can something be done to make Shiro support either Spring 2.5.6 or
>>>> >>> Spring
>>>> >>> 3.0 more seamlessly?  I recently upgraded from 2.5.6 to 3.0 and found
>>>> >>> that
>>>> >>> Shiro was including spring 2.5.6 as a dependency.  I had to manually
>>>> >>> exclude
>>>> >>> it in my pom.
>>>> >>> <dependency>
>>>> >>>                 <groupId>org.apache.shiro</groupId>
>>>> >>>                 <artifactId>shiro-spring</artifactId>
>>>> >>>                 <version>${shiro.version}</version>
>>>> >>> <exclusions>
>>>> >>> <exclusion>
>>>> >>> <groupId>org.springframework</groupId>
>>>> >>> <artifactId>spring</artifactId>
>>>> >>> </exclusion>
>>>> >>> </exclusions>
>>>> >>> </dependency>
>>>> >>> The good news is that this doesn't seem to have caused any problems
>>>> >>> for my
>>>> >>> application.  I saw another project that manages to support both
>>>> >>> versions in
>>>> >>> their pom, but can't remember which project it was or how they did it.
>>>> >>>  I'll
>>>> >>> try to find it if it would help.
>>>> >>> Of course I'm open to suggestions if there is a better way to deal
>>>> >>> with
>>>> >>> this.
>>>> >>> Thanks,
>>>> >>> Tauren
>>>> >>>
>>>> >>> On Fri, Jan 22, 2010 at 5:59 PM, Jason <[email protected]> wrote:
>>>> >>>>
>>>> >>>> ah, the question wasnt how to check, but how to configure.
>>>> >>>> I mean how do u abandon the shiro.ini roles & permissions mechanism
>>>> >>>> for
>>>> >>>> TextConfigurationRealm etc and model it in spring instead.
>>>> >>>> I thought in a previous post you mentioned this should be done.
>>>> >>>>
>>>> >>>> re the useage though I'm really keen for an aop solution, and an xml
>>>> >>>> element something like this:
>>>> >>>> <shiro:requires permissions="user:edit"/>
>>>> >>>> that I could just drop into any spring bean would be great. far
>>>> >>>> better for
>>>> >>>> me than annotations.
>>>> >>>>
>>>> >>>> Cheers
>>>> >>>> Jason.
>>>> >>>>
>>>> >>>>
>>>> >>>> Les Hazlewood wrote:
>>>> >>>>>
>>>> >>>>> Hi Jason,
>>>> >>>>>
>>>> >>>>> What do you mean by 'configure permissions' exactly?  Typically
>>>> >>>>> permission checks in standalone applications are done by explicitly
>>>> >>>>> checking (subject.isPermitted(blah)) or using Shiro's
>>>> >>>>> @RequiresPermissions annotation.
>>>> >>>>>
>>>> >>>>> Regards,
>>>> >>>>>
>>>> >>>>> Les
>>>> >>>>>
>>>> >>>>> On Thu, Jan 21, 2010 at 8:13 PM, Jason Eacott
>>>> >>>>> <[email protected]>
>>>> >>>>> wrote:
>>>> >>>>>>
>>>> >>>>>> thanks!
>>>> >>>>>> now how do I configure permissions etc in spring for a standalone
>>>> >>>>>> app?
>>>> >>>>>>
>>>> >>>>>>
>>>> >>>>>> Les Hazlewood wrote:
>>>> >>>>>>>
>>>> >>>>>>> The upcoming Shiro 1.0 release will have improved Spring
>>>> >>>>>>> application
>>>> >>>>>>> support, especially for Spring web applications.
>>>> >>>>>>>
>>>> >>>>>>> In Shiro-enabled Spring web apps today, there was often a hybrid
>>>> >>>>>>> configuration - you would usually define an INI-based Shiro Filter
>>>> >>>>>>> in
>>>> >>>>>>> web.xml and configure it via INI mechanisms.  But often you would
>>>> >>>>>>> configure the SecurityManager and its dependencies (Realms, etc)
>>>> >>>>>>> in
>>>> >>>>>>> applicationContext.xml.  In Shiro 1.0, you will be able to
>>>> >>>>>>> configure
>>>> >>>>>>> all of Shiro in your Spring files and only touch web.xml only when
>>>> >>>>>>> setting up Shiro for the first time.
>>>> >>>>>>>
>>>> >>>>>>> There are many benefits for Spring users when configuring Shiro
>>>> >>>>>>> entirely in Spring instead of in web.xml:
>>>> >>>>>>>
>>>> >>>>>>> 1) Shiro configuration can live along side where you configure the
>>>> >>>>>>> rest of your application - no need to flip back between web.xml
>>>> >>>>>>> and
>>>> >>>>>>> spring files when making configuration changes.
>>>> >>>>>>> 2) Shiro configuration can leverage Spring-specific configuration
>>>> >>>>>>> benefits, such as PropertyPlaceholderConfigurer for properties
>>>> >>>>>>> based
>>>> >>>>>>> configuration at startup, spring-managed lifecycles (init-method,
>>>> >>>>>>> destroy-method), circular dependency checks, and more.
>>>> >>>>>>> 3) Custom javax.servlet.Filters that you could use in Shiro's
>>>> >>>>>>> powerful
>>>> >>>>>>> url-pattern-based filter chain definitions can also be defined in
>>>> >>>>>>> Spring and acquired automatically at startup.
>>>> >>>>>>>
>>>> >>>>>>> The current documentation for all of this is located here:
>>>> >>>>>>>
>>>> >>>>>>> http://cwiki.apache.org/confluence/display/SHIRO/Spring
>>>> >>>>>>>
>>>> >>>>>>> Please feel free to review and offer suggestions/improvements.
>>>> >>>>>>>  The
>>>> >>>>>>> mechanisms documented (using Spring's DelegatingFilterProxy and
>>>> >>>>>>> the
>>>> >>>>>>> new ShiroFilterFactoryBean) have been tested and the two spring
>>>> >>>>>>> web
>>>> >>>>>>> sample applications have been updated to use this approach.
>>>> >>>>>>>
>>>> >>>>>>> Early adopters are encouraged to use this newer support before 1.0
>>>> >>>>>>> is
>>>> >>>>>>> released as there probably won't be any significant changes to
>>>> >>>>>>> this
>>>> >>>>>>> mechanism before then.  (SecurityManager configuration might be
>>>> >>>>>>> simplified via a Spring FactoryBean as well, but that won't affect
>>>> >>>>>>> web
>>>> >>>>>>> configuration).
>>>> >>>>>>>
>>>> >>>>>>> Please give it a try and let us know what you think!
>>>> >>>>>>>
>>>> >>>>>>> Best,
>>>> >>>>>>>
>>>> >>>>>>> Les
>>>> >>>>>>>
>>>> >>>>>
>>>> >>>
>>>> >>>
>>>> >>
>>>> >
>>>
>>>
>>
>

Reply via email to