Another CPANPLUS failure
Andy Armstrong
n at rciss.us
Mon Oct 15 12:49:40 UTC 2007
On 15 Oct 2007, at 13:38, David Golden wrote:
> I believe one big difference between CPAN::Reporter and CPANPLUS is
> the way test processes are executed. CPAN::Reporter executes all
> commands in a command shell and tees the output to capture it for
> reporting. I think that CPANPLUS uses IPC::Cmd, which uses either
> IPC::Run or IPC::Open3 and reads the output stream directly.
> TAP::Harness::Iterator::Process also uses IPC::Open3, maybe something
> is getting gummed up in all the filehandle redirection?
Oh goody :)
I'm now attempting to reproduce our test failures with CPANPLUS.
Unfortunately I'm getting entirely *different* failures.
This is probably premature, ill-informed opinion - but based on this
and one previous dalliance with CPANPLUS I'm prepared to say I'm not
a fan :)
--
Andy Armstrong, Hexten
More information about the tapx-dev
mailing list