On Fri, 16 Dec 2011 08:06:32 +0100 Mark Burgess <m...@cfengine.com> wrote:
MB> Please forget you ever saw this.
Saw what?!?!?
MB> This might look superficially attractive to a programmer but it
MB> suffers from many problems, not the least of which would be a major
MB> increase in the size and complexity of the parser. This approach
MB> does not satisfy the criteria for good knowledge management. It
MB> appeals only to programmers - and would allow programmers to gratify
MB> themselves at the expense of communicating policy to other
MB> stakeholders.
You'd rather have us create 10 bundles to express 10 variations in
system policy? That may make sense to the *parser* but not to the
human. Inheritance won't solve this problem and will very probably
introduce a new one ("what got defined up the inheritance chain?").
Please remember all the cfengine users are programmers by necessity.
MB> If there were ever a valid use case for sharing data as json, an
MB> intermediate function interface would be a defensible
MB> approach. Embedding a sub-language within cf3 would repeat the
MB> mistakes made in the cf2 which led to chaos and feature creep.
MB> The goal of cf3 core is to be small and stable. We are looking at
MB> other mechanisms to simplify body templating.
I look forward to a better approach :)
Thanks
Ted
_______________________________________________
Help-cfengine mailing list
Help-cfengine@cfengine.org
https://cfengine.org/mailman/listinfo/help-cfengine
