Stas Bekman wrote:
> Geoffrey Young wrote:
> 
>>
>> Stas Bekman wrote:
>>
>>> I'm working out the details of using the new CLONE_SKIP feature. I was
>>> thinking that it'd be useful to make a new branch for tha purpose. The
>>> main problem is that at the moment we can't run the new tests
>>> conditionally, since there is nothing to condition the support for
>>> CLONE_SKIP :(
>>
>>
>>
>> sure, go ahead and create a new branch.  I hope this one merges better,
>> though ;)
> 
> 
> Me too :) should we call it 'ithreads'?
> 
> svn.apache.org/viewcvs.cgi/perl/modperl/branches/ithreads?

sounds good.  maybe the unstable- prefix like apreq and the rename did is
also a good


> I think 5.005_0x is a history now. At least I've noticed that more and
> more modules on CPAN no longer bother to support it. I've stopped trying
> to support it in my modules, where it relies on other modules which
> already don't support it.

actually, we need T::B 0.49 or greater for things to work, so we probably
need to bundle it or require it as an external dependency.  IIRC no perl has
shipped with that version or higher yet.


>> personally I'm very much in favor of dumping Test.pm if we can make
>> Test::Builder do what we need.
> 
> 
> Well, we can't yank Test.pm, because we need special straps under
> mod_perl. such as 'plan $r, ...'. 

you can yank Test.pm and still have those straps.  at least I think you can.
 the current -withtestmore  allows plan() to work properly with $r, for example.

> Though we could make Test.pm a proper
> subclass of Test::More.

there's already Test::Legacy for that, which might be thing work
experimenting with and adding our foo to if need be.

--Geoff

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to