> Author: sauer
> Link to topic: https://cfengine.com/forum/read.php?3,24472,24485#msg-24485
> I use that value as a descriptive name, the same way other promises use it;
> usually I choose a more descriptive version of the bundle name, but it always
> describes the purpose in one or two words. :)
> methods:
> any::
> "securetty" usebundle => harden_securetty;
> use_alpha_config|!cfengine_3_0::
> "vars" usebundle => harden_vars;
That definitely seems more useful than "any".
> I'm having difficulty envisioning a scenario where using the name of a module
> promise provides any added value - either in functionality or just prettier
> syntax
It would be a prettier and more consistent syntax for bundles that fulfill a
promise for the name of something passed. For example, a files promise might
look like:
files:
"/my/file" perms => mog( "644", "root", "root" );
It's fulfilling some desired state for /my/file.
Now let's say I want a bundle to fulfill some desired state for /my/file.
Today it seems that we have to do this:
methods:
"not-used" usebundle => my_bundle("/my/file", "some-other-parameter");
But what I'd rather do is:
methods:
"/my/file" usebundle => my_bundle("some-other-parameter");
I realize this may not be useful for desirable for all kinds of bundle calls.
But I can see it being useful in many situations. That way the language could
be used consistently for methods: promises and other kinds of promises.
> While I can see the theoretical argument of having the name be accesible
> somehow, it'd have to be a new variable; the this.promiser variable anywhere
> in a called bundle already has a different meaning, and there's no attribute
> of the methods promises (other than possibly "comment") where
> $(this.promiser) would make any sense.
A new variable would be fine.
The reason I thought of $(this.promiser) is because it seems like
$(this.promiser) references that same left-hand-side value for files:
iterations and searches.
You're calling that left-hand-side value a comment in methods: calls, but it's
definitely not a comment value for other types of promises. So I'm just
suggesting that it be made consistent, and it seems like all we need is a way
for the called bundle to access that left-hand-side value.
> On an unrelated note, Tod, did you ever work in the Champaign, IL office?
> Your name looks familiar, but it's been ~6 years. :)
I occasionally work with people there, so we definitely may have crossed paths
in the past. Nice to hear from you again if so. :)
--
Tod Oace, Intel Corporation <t...@intel.com>
_______________________________________________
Help-cfengine mailing list
Help-cfengine@cfengine.org
https://cfengine.org/mailman/listinfo/help-cfengine
