-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Tue, 2 Apr 2002, Dave Rolsky wrote:
And all I'm suggesting is that because there is no other way to do this
stuff, that when you change the text in the generated Makefile, that
you at least document that it has changed, so people can test their
On Sun, 7 Apr 2002, Edward Avis wrote:
Of course I wouldn't be the one doing it anyway, so it's really none of
my business, but I just want to say that as a MakeMaker _user_ I
wouldn't expect such a service or have any right to expect it. I think
that people who do ugly hacks against the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Sun, 7 Apr 2002, Dave Rolsky wrote:
But please get off your high horse and recognize that for a long time
MM has been the only way to install things (that has now changed with
Module::Build) and MM has no decent interface for customization.
Oh,
On Monday, April 8, 2002, at 08:35 AM, Edward Avis wrote:
So please, use whatever means necessary to clean out the Augean stable
that is MakeMaker -
On Monday, April 8, 2002, at 08:44 AM, Dave Rolsky wrote:
But please get off your high horse ...
It's those damn high horses that created
On Monday, April 8, 2002, at 10:28 AM, Dave Rolsky wrote:
Seriously, I like the plan of trying to get Module::Build into shape for
5.8.1.
Likewise. I may propose a lightning talk at TPC for marketing.
This probably means updating CPAN to be Module::Build-aware. Does
Andreas
read this
On Mon, Apr 08, 2002 at 12:01:04PM +1000, Ken Williams wrote:
Andreas didn't respond. He may not have time (or interest) in it,
perhaps p5p is the best place to go.
Andreas was generally offline dealing with Real Life for the past many
months, but he seems to be back on. So I'd just try
On Tue, 2 Apr 2002, Michael G Schwern wrote:
And even though there are ways around it for tests, what about for the
install process? What if I need to run a few as part of the 'make
install' bit? How else can I make that happen besides fiddling with the
Makefile text itself?
There is
On Mon, 1 Apr 2002, Michael G Schwern wrote:
So... what is it you were doing with the test* methods and how? Let's
see what we can preserve.
I was doing this:
my $test = $self-SUPER::test(_);
# %MY::APACHE is set in makeconfig.pl.
# If we are not going to test with Apache there
On Mon, Apr 01, 2002 at 09:46:42AM -0600, Dave Rolsky wrote:
On Mon, 1 Apr 2002, Michael G Schwern wrote:
So... what is it you were doing with the test* methods and how? Let's
see what we can preserve.
I was doing this:
my $test = $self-SUPER::test(@_);
# %MY::APACHE is set
On Mon, 1 Apr 2002, Michael G Schwern wrote:
Hmmm. This can't be preserved. Wouldn't it be a lot less magical to
just put the %ENV logic into your tests, or a library in t/lib which
your tests use?
I'm not sure what you mean about putting it _in_ the tests. We have to
set the variables
At least some of the various targets in the Makefiles produced by the
newer EU::MM appear to have changed. For example, the test_dynamic target
has changed quite a bit.
But these bits of text were really part of the public API of the module,
in that you could always overload the test method in
11 matches
Mail list logo
