Andy Lester wrote:
I just committed a patch that autodetects the HASATTRIBUTE_xxx
attributes for GCC. It's kinda funky under Solaris, but still passes.
Please let me know what happens on other platforms. I'm off to bed.
I'm assuming you're speaking of revisions to config/init/attributes.pm.
It runs without incident on Darwin.
I didn't look at config/init/attributes.pm too closely when it was first
committed to trunk (other than to create an RT scheduling it for unit
tests). On closer examination, however, I believe it is misnamed. It
does not fit the description provided in docs/configuration.pod for an
*initialization* test:
"I<Initialization steps> are run before any other steps. They do tasks
such as preparing the configuration system's data structures and
checking the F<MANIFEST>."
Rather, it fits the description for what docs/configuration.pod calls a
*probe*:
"Probes are automated tests of some feature of the computer. These
should be used wherever a value will not often need to be modified by
the user. A step containing probes is an I<automatic step>. Automatic
steps should be in the F<config/auto> folder."
Note that init::attributes::runstep(), like the runstep() methods of all
other "auto" steps, has the C-compiler run tests over a code sample.
init::attributes should therefore be renamed auto::attributes. If
should probably also be moved later in the configuration process so that
it sits beside other "init" steps, but I don't have the time now to
investigate exactly where it should sit.
kid51
