[OCF] ocf-shellfuncs standardization - Comments? Objections?

[Lon Hohberger]

On Wed, 2005-06-22 at 19:18 -0600, Alan Robertson wrote:

> > It makes a great deal of sense to enforce a standard convention on the
> > name/call semantics for a logging function.
> 
> OK.  I agree.  But, what exactly do you mean...
> 
> Do you mean you agree, but it should be more stringent, or that you 
> agree, and that's that, or that you disagree :-).

Having a set logging function name / calling conventions allows
implementation of different logging functions depending on the
requirements of the developer - without having to change how the RAs
call it.

So, what the function does should probably be left up to the
implementation details / developer requirements / whatever.  It allows
you to swap around how you log stuff without changing the RA (except
perhaps the shell-functions script I source).

I personally wouldn't require RA's to use a set logging function name /
call semantics, but I'd certainly "strongly recommend" that they do.

The call semantics you provided are fine -- I think they'd work well.

This will all get pushed back to the OCF RA API draft, right? ;)

-- Lon