[CCing firstname.lastname@example.org]On Sun, Dec 11, 2016 at 11:11:56AM +0000, Christopher Hoskin wrote:
Amazon::S3 is a fork of Net::Amazon::S3.
“This need for this module arose from some work that needed to work
with S3 and would be distributed, installed and used on many various
environments where compiled dependencies may not be an option.
Net::Amazon::S3 used XML::LibXML tying it to that specific and often
difficult to install option. In order to remove this potential barrier
to entry, this module is forked and then modified to use XML::SAX via
Since Net::Amazon::S3 is already packaged for Debian as
libnet-amazon-s3-perl, the motivation for the fork does not apply to
us. Also, Net::Amazon::S3 seems more actively maintained upstream
(most recent release 2014 as opposed to 2009 for Amazon::S3).
I’m therefore wondering if patching RT to use Net::Amazon::S3 might be
a better option? (I don’t know how much work this would involve?)
I did think of this, but I assume that it wouldn’t be in RT upstream’s
interest to accept such a patch (for the reason stated in the above
justification for the fork), nor in Debian’s interest to permanently
deviate from upstream in this way.
As for the concern about the Amazon::S3 being unmaintained - I
haven’t done a detailed investigation but I would guess that the fact
that RT adopted it in their new release means that it is at least good
enough for them.
Upstream: can anyone comment on the decision to go with Amazon::S3
rather than Net::Amazon::S3?