This is obviously an esoteric condition, and one that can be remedied on
my setup at least by using t/all.t, but I am noting this bug here for
future reference.
this is just weird :)
without probing further, I'd say that all.t isn't a bad idea. stas and
I had a series of arguments long ago on what it meant when mod_perl
wasn't available but folks were using Test*Perl features. I'm still of
the belief that tests should individually probe for what they need via
need(), either in each individual test or an all.t.
since Apache::Reload is being released solo, I think it's perfectly
valid to make sure mod_perl is available via an all.t test - there are
many, many (obscure or not) circumstances where mod_perl might not be
available at test time, none of which are a reflection on the code in
Apache::Reload. imho, if the test environment doesn't have everything
it needs it's not a test failure :)
so, I guess this doesn't help us figure out where Apache-Test is going
wrong. but it's all I have at the moment :)
--Geoff
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
