Stephen Hahn wrote:
> * Shawn Walker <[EMAIL PROTECTED]> [2008-09-09 00:11]:
>> Shawn Walker wrote:
>>> Danek Duvall wrote:
>>>> history.py:
>>>>
>>>>   - I'm curious why you went with module-level functions, rather than
>>>>     wrapping everything in a class.  That would help get rid of all the
>>>>     "global" declarations.
>>> Because this way, the object doesn't have to be passed around.  You can 
>>> call history from anywhere, any module, at any time, and be guaranteed 
>>> that anything you do is for the current operation.
>>>
>>> This particular data structure is a recommendation from one of the 
>>> authors of Python (Alex Martelli), that you can find in the Python 
>>> Cookbook, recipe 6.16, "In most cases, you don't need either of them 
>>> [singleton or borg]. Just write a Python module, with functions and 
>>> module-global variables, instead of defining a class, with methods and 
>>> per-instance attributes."
>> To be clear, the whole point of doing this was so that, later on, when 
>> we go to implement bugs such as 2022 (client should provide operational 
>> intent to server), I didn't want to have to pass around an object to 
>> every place where we might need to get intent information.
>>
>> By using a module, I was able to get all the benefits of a singleton 
>> (shared state, single instance) without complex mechanisms.
>>
>> This allows the client to get access to intent information at any place 
>> throughout the code just by using history.method() without having to 
>> rely on the caller to pass in a reference to it.
> 
>   I understand that.  How do you suggest I diff two image's histories?

I would perform a diff on the output of "pkg history" for each image.

-- 
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to