Re: Bug 025: Instances of TestMode

On Wed, May 11, 2011 at 5:10 PM, Sean B. Palmer <sean@miscoranda.com> wrote:

> The latter class has a similar enumeration of subclasses to
> earl:OutcomeValue, and yet earl:OutcomeValue has (default?)
> instances for those subclasses and earl:TestMode does not.

Sorry, I got this the wrong way around. This bug report should read:

Given that there is an enumerated subclass for each instance of
earl:OutcomeValue, why is there no enumerated subclass for each
earl:TestMode instance?

The error in the original bug report shows my thinking. It would be
more sensible to have the classes, which allow some form of
extensibility by class, than the instances. So much so that I had
remembered earl:TestMode has having only the classes and no instances,
rather than only the instances and no classes.

This still, then, compares to Bug 018, arguing that the classes should
be around and the instances removed.

I discovered that this bug incorrectly referred to classes rather than
instances by reviewing my hand drawn EARL data model diagram. This is
a good demonstration of why the provisions described in Bug 001 would
be so useful.

-- 
Sean B. Palmer, http://inamidst.com/sbp/

Received on Wednesday, 11 May 2011 19:17:11 UTC