Another Test::Harness problem [was: FAIL Path-Class-0.16]

[Andy Armstrong]

On 1 Nov 2007, at 18:02, Ovid wrote:
> I thought we agreed that until the new TAP comes out, we will have  
> to support TODO in plan.  In fact, I *thought* somebody checked in a  
> change, but I can't recall.


I'm pretty sure we've never had that code. The decision that I recall  
was to do with relaxing the parsing of # skip to allow, e.g., #  
skipping.

The todo in plan code would be moderately involved - we'd have to  
decide which layer was responsible for associating the stored todo  
numbers with incoming tests.

That is: do we keep it all in the parser and fake the todo status of  
results or is the harness responsible for remembering which tests are  
flagged as todo.

-- 
Andy Armstrong, Hexten