On Wed, Oct 28, 2009 at 4:14 PM, Luke Kanies <[email protected]> wrote:
>
> On Oct 28, 2009, at 11:45 AM, Nigel Kersten wrote:
>
>>
>> On Wed, Oct 28, 2009 at 11:38 AM, Luke Kanies <[email protected]>
>> wrote:
>>>
>>> On Oct 28, 2009, at 11:31 AM, Nigel Kersten wrote:
>>> [...]
>>>>> That I see.  If you're just downloading facts from the server to
>>>>> set
>>>>> the environment, though, how is that fundamentally different from
>>>>> the
>>>>> server owning what environment the client belongs to?
>>>>
>>>> The facts look at local configuration on the client such as the
>>>> debconf datastore or a property list.
>>>>
>>>>>  Are your facts
>>>>> just simple strings, or are they making complicated calculations
>>>>> based
>>>>> on something on the client?
>>>>
>>>> It's not particularly complicated.
>>>>
>>>> It works out whether it's a Mac, Linux or Solaris box.
>>>> It works out whether it's a desktop, laptop or server
>>>> It works out whether the owner of the machine has decided to
>>>> configure
>>>> this machine to use unstable or testing.
>>>
>>>
>>> So you've got different environments for different platforms, then?
>>> Did you do that for scaling reasons, or just because the code for
>>> them
>>> is that different, or what?
>>
>> Ah. So that each relevant team can control their own release process
>> and follow their own schedule.
>>
>> We all commit to a common development "environment" that is only used
>> for the most close to the metal development testing.
>>
>> Then each team chooses to integrate from the common development
>> environment to their own unstable environment on their own schedule.
>> They then control the integration from unstable->testing and
>> testing->stable.
>>
>> We only commit to the common dev environment.
>
> That all seems great.  I just don't get where it changes on a given
> host.

the last component of the environment (testing, unstable, stable) is
determined by local configuration files on the client. debconf for
linux, plists for Mac.

>
>>> There's definitely some critical concept missing here...
>>
>> was that it ? :) I'm a bit flat out today, but happy to jump on a
>> phone or do it in IRC if we need more of a back and forth conversation
>> about this.
>
> A call might be helpful.  Ping me on IRC?

will do.

>
> --
> Humphrey's Law of the Efficacy of Prayer:
>     In a dangerous world there will always be more people around whose
>     prayers for their own safety have been answered than those whose
>     prayers have not.
> ---------------------------------------------------------------------
> Luke Kanies | http://reductivelabs.com | http://madstop.com
>
>
> >
>



-- 
nigel

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Puppet Developers" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/puppet-dev?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to