On Fri, Dec 17, 2010 at 2:40 PM, Stefan Schulte
<[email protected]> wrote:
> What about replace => false. If the file is not there we dont need
> checksums because we know we have to copy the file. If the file is
> present we dont need to calculate checksums because we wont replace
> anyway. I think thats a valid usecase where you have a source property
> and no checksum parameter.

We will be adding more intelligent behavior around checksumming so we
simply ignore checksumming in situations like this, but not right now.

That would be a significant enough change it would be introduced in a
major version, not a minor.



>
>> One of the problems with this model, though, is that there's a semantic 
>> difference between writing the file out and not - if you've got a service 
>> that needs to get restarted if the file changes, but not if it doesn't, then 
>> you can't just write it out every time.
>>
>> On Dec 16, 2010, at 7:49 PM, Trevor Vaughan wrote:
>> > I disagree.
>> >
>> > Look at the concat module. Why waste resources checksumming when it
>> > would be more efficient to just write out the file?
>> >
>> > Read + Checksum + (potential) Write is more I/O and CPU intensive than
>> > Write.
>> >
>> > Trevor
>> >
>> > On 12/16/2010 07:25 PM, Nigel Kersten wrote:
>> >> On Thu, Dec 16, 2010 at 2:55 PM, Jesse A Wolfe <[email protected]> wrote:
>> >>
>> >> Note that we changed the desired behavior to fail with a syntax error 
>> >> instead.
>> >>
>> >> It doesn't make sense to not checksum and specify content/source.
>
> -Stefan
>



-- 
Nigel Kersten - Puppet Labs -  http://www.puppetlabs.com

-- 
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