Re: perl -wc

["Kyle R. Burton" <mortis@voicenet.com>]

> > are you testing all 1030 files each time you run the test, or only the
> > files that have changed since the last test?
> 
> I thought of that, too, but I'm not sure it's sufficient.  The tester
> would have to know about which modules depend on which other modules.
> You'd probably have to set up something with make, and then remember
> to keep it updated as new modules are added and dependencies change.


Thank you for the responses.  I'll look into using the SIG WARN idea,
it looks to be the most promising - the fork/exec itself is not what I
was necessarily worried about taking up the time, it was the
re-compilation of all the commonly used modules (if 800 of the modules
all use Foo::Bar, then Foo::Bar [and all it's dependencies] will be
'perl -wc'd 800 times, wasting a lot of time).

As far as the mtime stuff goes, I want to be able to run this against
not just my local working copy, including my changes, but those that
other developers commit, and it'd be nice to be able to run it on the
codebase in newly installed environments (along with the unit tests)
to show that the code is of high quality in that environment as well.

Make tends to be slow for similar reasons.  I've used Ant and Maven
and in situations like this I'm envious of the speed of those tools.
Especially for C/C++ - It'd be nice if gcc could be used in a similar
way.

Thanks again for the replies.

Kyle


-- 

------------------------------------------------------------------------------
Wisdom and Compassion are inseparable.
        -- Christmas Humphreys
mortis@voicenet.com                            http://www.voicenet.com/~mortis
------------------------------------------------------------------------------
-
**Majordomo list services provided by PANIX <URL:http://www.panix.com>**
**To Unsubscribe, send "unsubscribe phl" to majordomo@lists.pm.org**