[rt.cpan.org #30603] T:H 3.0 breaks POE

[Andy Armstrong via RT]

<URL: http://rt.cpan.org/Ticket/Display.html?id=30603 >

On Sun Nov 11 21:40:27 2007, andreas.koenig.7os6VVqR at franz.ak.mind.de wrote:
> Come on, Andy, in all fairness. Please show the world what exactly the
> bug in POE's testsuite is.

Ah, sorry Andreas. You're quite right. I wasn't being terse on purpose - just replied in a 
hurry.

The bug is that when POE wants to skip all its tests it emits something like

1..0 # not going to run these tests

The accepted TAP syntax for that is 

1..0 # SKIP not going to run these tests

In the case where we're parsing TAP version 12 or earlier (i.e. we see no version line) we 
currently relax the rules to allow things like

1..0 # skipping merrily away

but we still barf if there's nothing resembling 'skip' at all.

We certainly could allow 

1..0 # anything

for older TAP because we have the TAP version 13 line to trigger strict mode. I didn't make 
the change on POE's account though because, having spoken to Chris about it in Vienna I got 
the impression that they'd be fixing it in POE's test suite.

Given that Ovid's just taken me to task for breaking POE too (I guess it's like sticking needles 
in a little effigy given that they're namesakes :) I'm quite happy to release T::H 3.01 with a 
fix.