Whoops.

http://groups.google.com/group/perl.cpan.testers/browse_thread/thread/72bcf48d152c17ad/d09e5c5f95dec79d

http://groups.google.com/group/perl.cpan.testers/browse_thread/thread/270d01eb25695b8a/044491c0d79e2cff

I am wondering, does the rapid use of multiple transport_args or other
commands, interrupt or confuse the connection to SMTP?



On Thu, Apr 16, 2009 at 9:51 AM, David Golden <[email protected]> wrote:

> On Thu, Apr 16, 2009 at 9:27 AM, M W487 <[email protected]> wrote:
> > Sorry, 1.53_02 did not work.
> >
> >
> http://mail.google.com/mail/?safe=on&safe=on&safe=on&shva=1#sent/120ae60088f6bc1a
>
> This is a link to a gmail email -- I can't see your emails since I'm
> not you.  :-)
>
> Add "email_to=y...@email" to .cpanreporter/config.ini and mails will
> go to you instead of to [email protected].  Then use "show
> original" in gmail and save that to a text file and send that as an
> attachment to the list.  (Don't forward -- we'll lose the header
> information.)
>
> Until I see some of the emails using 1.53_02, I don't want to make
> guesses as to what's happening.
>
> If you promise not to send me more than one or two sample reports, you
> can even set 'email_to' to my email address and let them go to me
> directly.
>
> > Why does Test::Reporter: transport_args occur three times?
> Test::Reporter:
> > report twice?
>
> Apparently, Test::Reporter send() calls report(), from() and subject()
> in null context to initialize them (they are cached).  I couldn't tell
> you why.  T::R is not the nicest codebase.  I do as little as I need
> to in maintaining it and don't spend cycles cleaning it up unless I'm
> fixing bugs.
>
> > Why do I have a huge PATH, that seems to grow without end if I only run
> the
> > smoker?  (Occasionally, I go into CPAN by hand, and install a module just
> to
> > clean out the build (?) directory and thus PATH).
>
> PATH or PERL5LIB?  CPAN.pm adds any successfully tested distribution
> to PERL5LIB.  This is essential to make sure that prerequisites that
> are tested but not installed are available for subsequent tests.  If a
> smoker is repeatedly calling CPAN's test(), then PERL5LIB will grow to
> be huge. (The same happens with distributions with massive dependency
> chains.)
>
> CPAN::Reporter::Smoker calls $CPAN::META->reset_tested() between
> distributions to make CPAN forget previously successful tests and
> reset PERL5LIB.  If minismokebox isn't doing that, then that's a bug
> in minismokebox.
>
> -- David
>

Reply via email to