On Tue, Mar 15, 2016 at 8:10 AM, Udondan <[email protected]> wrote:

> I also agree it would have made more sense if the dynamic include was
> optional and the default behavior was the version 1 way. But I also see how
> this is hard to justify now that Ansible 2 works as it does, it's hard to
> go back. If Ansible follows semver that would actually mean we have Ansible
> 3 knocking on our door if the behavior is switched back.
>
> Ideally, in my opinion, the new functionality would have been added with a
> different name, for example "import" instead of "include". This would have
> allowed a clean way to migrate without breaking things. It also makes it
> possible to distinguish between two complete different features by name.
>


Agreed, hind sight is 20/20. That is also something we considered, however
again at this point it would break playbooks as written for 2.0.



> We have adopted Ansible 2 but I would not mind if the behavior changed
> back to the old static include. In any case, I will adjust our roles. Most
> of the includes do not need to be dynamic and I would set them to static
> just to get rid of the hundreds blue include statements during playtime.
>
> Now, I'm wondering what happens if a static include includes a file which
> has a dynamic include. :)
>

Thanks for this, this is something I need to test with my feature branch. I
don't think I recursively parse tasks brought in from static includes...

-- 
You received this message because you are subscribed to the Google Groups 
"Ansible Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/ansible-project/CAMFyvFi-%2B1B9wNpazV%3DnoY%3DhY_%2Bpsdyh98c8XZ53zpRcd4KmBA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to