On Jun 23, 2005, at 3:42 PM, Randal L. Schwartz wrote:
"Rafael" == Rafael Garcia-Suarez <[EMAIL PROTECTED]>
writes:
Rafael> Change 24956 on 2005/06/23 by [EMAIL PROTECTED]
Rafael> Add a pseudo-test to clean up data files for
Archive::Tar tests
Does that mean you depend on all tests being run, and on the order of
runnings tests?
Is that wise?
Why not just clean up after each test? Surely, each test file knows
what temp files it is creating.
Sorry for being slow on the reply, but here it is :)
The tar files are ones created with gnutar and serve as a reference
point for tests.
We could have them generated from A::T as well, but any error in A::T's
creation code
may not be caught by it's untarring code.
That was the only rationale to add the files as binary files, rather
than generate them
(be that from a 00.t file, or from a BEGIN block that checks, etc)
I prefer the binary files, but if it's a problem for core, then we can
change it of
course.
--
Jos Boumans
"If superman is so smart, why does he wear underpants over his
trousers?"
CPANPLUS http://cpanplus.sf.net