[Freeipa-devel] [Patch] Re-base memberOf plug-in off of current FDS memberOf plug-in

[Rob Crittenden]

Nathan Kinder wrote:
> Rob Crittenden wrote:
>> Nathan Kinder wrote:
>>> Rob Crittenden wrote:
>>>> Nathan Kinder wrote:
>>>>> This fixes many, many issues with the memberOf plug-in.
>>>>>
>>>>> -NGK
>>>>>
>>>>>
>>>>
>>>> Wow, a lot to go through. I think I grok most of this.
>>>>
>>>> A couple of comments/questions.
>>>>
>>>> Most of the functions are renamed in a more generic way, I'm 
>>>> presuming to make it easier to keep in sync with the FDS version of 
>>>> the plugin, right?
>>> Yes.  I want to make it as easy as possible to diff between the IPA 
>>> and FDS memberOf plug-in code.  This should help in any future fix 
>>> porting work.
>>>>
>>>> But some functions weren't renamed, ipamo_postop_init() for one. Do 
>>>> we want to go ahead and do that?
>>> This one function is the entry function for the plug-in.  The 
>>> configuration entry in dse.ldif has an attribute with this function 
>>> name that is used to load the plug-in.  I didn't want to do anything 
>>> that required us to change the configuration entry since we'd need to 
>>> handle making those changes during an upgrade.
>>>>
>>>> Some comment blocks look really strange in the patch too:
>>>>
>>>> - * Online tasks interface (to support import, export, etc)
>>>> - * After some cleanup, we could consider making these public.
>>>> - */
>>>> + *  * Online tasks interface (to support import, export, etc)
>>>> + *   * After some cleanup, we could consider making these public.
>>>> + *    */
>>> That's caused by vim trying to be smart.  We can clean those up.
> I've attached a patch that cleans up these mangled comments.
> 
> -NGK
>>>>
>>>> What is the purpose of dont_allow_that()?
>>> The FDS memberof plug-in allows dynamic configuration changes.  In 
>>> IPA, we are hardcoding the attribute that memberOf uses, but I wanted 
>>> to leave the dynamic config code in place for the most part.  We 
>>> register callback functions to be invoked when one performs different 
>>> operations against the memberOf plug-in's config entry.  We don't 
>>> want to allow anyone to modify, rename, or delete our config entry 
>>> while the server is running, so we register dont_allow_that() as the 
>>> callback for these types of operations.  This will result in a 
>>> DSE_UNWILLING_TO_PERFORM being returned to the client attempting to 
>>> modify the config entry.
>>>
>>> -NGK
>>
>> Ok, then ack
>>
>> rob
> 

ack

Pushed both to master.

rob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3245 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://listman.redhat.com/archives/freeipa-devel/attachments/20080718/339abd6a/attachment.bin>