On Tue, Feb 11, 2014 at 3:40 PM, Kevin A. McGrail <[email protected]> wrote:
> (2) You can use C<< eval { HTML::Parser->VERSION(3.46) >> for your
> comparison which won't get picked up by mistake.
>
> I can look at changing that for 3.4.1 but SA 3.4.0 is effectively a
> released artifact at this point so it sounds like we'll need a 3.4.1
> release to handle this. I've been handling the release all day and night
> now so I'm running on empty. With the eval above, how would you recommend
> implementing it because I don't know the C<< syntax you are showing. This
> is the current block. A little direction would help a lot!
>
>
Sorry. That was POD code quoting, which I use occasionally in email. I
mean this:
unless ( eval { HTML::Parser->VERSION(3.46) } ) {
...
}
The VERSION method (see "perldoc UNIVERSAL") with an argument is an
assertion so it needs to be wrapped with eval.
> Finally, could you tell me more for a second about the real world
> impact of this indexing issue? From a CPAN lay-perspective, I was
> able to run cpan and do install Mail-SpamAssassin and things appear
> to have worked.
> http://search.cpan.org/~kmcgrail/Mail-SpamAssassin-3.4.0/ also looks
> ok.
>
>
As long as "Mail::SpamAssassin" was indexed, then people can install it.
The things that failed to index did not get updated. So, for example,
compare these index entries:
Mail::SpamAssassin 3.004000
K/KM/KMCGRAIL/SpamAssassin/Mail-SpamAssassin-3.4.0.tar.gz
Mail::SpamAssassin::Conf bogus
K/KM/KMCGRAIL/Mail-SpamAssassin-3.3.2.tar.gz
Should some crazy person do "cpan Mail::SpamAssassin::Conf" they would get
the 3.3.2 tarball.
Realistically, you probably don't need to care.
David
--
*David Golden* <[email protected]>
*Take back your inbox!* → http://www.bunchmail.com/
Twitter/IRC: @xdg
