[EMAIL PROTECTED] wrote:
>>> It took me a minute to figure this out.  The way this is implemented is
>>> kindof sneaky.  I had to notice that every function starts with a call
>> OK maybe sneaky, but the intent wasn't to be.
> 
> Not accusing you of malicious intent. :) 
> 
>>> to get_env() and get_env() checks is_BEenv.  It might just be easier to
>>> check is_BEenv directly inside the methods.  There are a couple of other
>>> possible approaches too:
>> Yeah I wasn't sure what the OO convention was, wrap a variable in a
>> method or just access it directly. If it is most conventional to
>> access the member directly I'll do that.
> 
> My take is that if you're the class, it's okay to access the variable
> directly.  If you're outside the class, you may want to use the accessor
> function.  Python has some additional functionality that lets you do

Right. I initially thought some of the methods might be called from outside the 
class but since none are, I'll access the class variables directly.

> neat things here, but I don't know if any of that is applicable.
> 
>>> 1. Create an abstract BootEnv class, define BootEnvZFS and BootEnvNull
>>> classes.  When we can import libbe, use the BootEnvZFS class, otherwise
>>> use the BootEnvNull class.  Presumably, the null class has methods that
>>> don't actually do any work.
>>>
>>> You could also just make the parent class have null methods, and then
>>> have BootEnvZFS do work.  There are a few different ways you could go
>>> with this general idea.  Stephen/Danek may want to say more about this,
>>> too.
>> Each of these are great ideas but wouldn't the first one create
>> duplicate code?
> 
> Not exactly.  The idea would be that you have the same methods defined
> for each class, but only BootEnvZFS would contain code that does any
> work.
> 
> I like the progress that we're making and I'd like to get a couple of
> other reviewers to sign off on your wad too.
> 

Yes me too especially before I send out what is hopefully the last webrev. ;-)

Tim

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

Reply via email to