Revision history for Perl extension PerlIO::subfile.

0.01  Sun Feb 18 17:16:25 2001
	- original version; created by h2xs 1.21 with options
		-A -n PerlIO::subfile

0.02  Sat Apr 21 17:39:11 BST 2001
      Change to use the modified layer API of 5.7.1
      As the pushed argument is now an SV, allow an SV that is an IV/UV
      (but not a PV) is treated as a relative end value. (which will make
      things easier for ex::lib::zip

0.03  Sun Aug 26 12:05:12 BST 2001
      By 5.7.2 it seems that the pushed argument is passed in as &sv_undef
      rather than NULL. I wasn't testing SvOK() and was getting use of
      uninitialized value errors.
      Changed tests to use Test::More, set $!=0 whenever we report it on !ok().
      Fixed test "Binmode should take us to the start of the subfile" to
      actually test tell(), rather than another needless seek.
      Fixed => to => in two of the assertions. Fixed a bug in another assertion.
      Looks like I'd not built against a -g perl for a while.

0.04  Sat Sep  1 12:25:23 BST 2001
      Uncommented :subfile-on-a-pipe tests. They now work. (perlio improved).
      Loop through the tests for 3 layers.
      Mon Oct 29 20:28:38 GMT 2001
      perlio.c layer API has changed again with the addition of a dup()
      function for cloning layers across. (both within and across ithreads)

0.05  Mon Mar 18 21:03:52 GMT 2002
      Catch up with the 5.7.3 perlio API changes.

0.06  Thu Jul 11 20:31:28 GMT 2002
      Catch up with the 5.8.0-RC1 perlio API changes.
      The perlio layer now reports a file position on unseekable streams,
      which messes things up when pushing a subfile onto (say) a pipe.
      Mark the test as a TODO, as it doesn't affect ex::lib::zip
      And I had a bug. Which valgrind found for me straight away.
      Use valgrind.
      It's far more effective than coffee.

0.07  Fri Jul  4 22:52:08 BST 2003
      Wow! It seems that I never tested this with a threaded perl build.
      How lax. Neither did CPAN testers. And strangely, PerlIO-gzip had been
      tested with a threaded build, because all the macros are correct there.

0.08  Mon Mar 13 23:06:51 GMT 2006
      It seems that 5.9.x doesn't like me unless I use () for XSLoader::load