[bop-devel] Hosted Payment

[Ivan Kohler]

On Sat, Jul 19, 2014 at 06:36:50AM +0200, Stefan Hornburg (Racke) wrote:
> I would be very interested to extend BOP in a manner to support these 
> callbacks, which are used in one or the other form by any gateway 
> today (at least if you want to evade PCI compliance).
> 
> The question is if we:
> 
> 1. extend BOP itself
> 2. write a subclass of BOP adding methods needed for redirect

#2 is the approach we took with Business::OnlineThirdPartyPayment 
#(working name).

I don't see any reason it couldn't be folded back into B:OP.  For now I 
don't see any immediate reason to do so -- we have a lot of production 
use of vanilla B:OP and there's value in being able to iterate the 
third-party/hosted/redirect codebase on its own right now.


> 3. write BOP 2nd generation

I've seen this attempted several times without traction.  I don't 
personally have the need or tuits to spearhead something like this, but 
I wouldn't stand in the way of someone else doing so.

I'd recommend backwards-compatibility for end users (i.e. application 
authors using Business::OnlinePayment) and gateway module authors.

Personally I plain to maintain stable B:OP, continue work on 
B:OnlineThirdPartyPayment (and hopefully rename it), do incremental 
refactoring, review/accept patches, write new gateways modules, etc.


> (Moo,

There's more "mini-Moose" candidates than Moo.  Has that proliferation 
stabilized?  What are the best ones now?

Are there tradeoffs vs. installability on old perl?  I know new versions 
are all the rage these days, but supporting back to say, 5.8 still seems 
useful for now for a simple module.


> get rid of HTTPS::Any, ...)

I uploaded Net::HTTPS::Any 0.11 which fixes the test failures.

Rather than "get rid of Net::HTTPS::Any", if you still perceive a 
problem here that needs solving, what is it?


-- 
Ivan Kohler
President and Head Geek, Freeside Internet Services, Inc.  http://freeside.biz/
Debian GNU/Linux developer  |  CPAN author  |  cat person  |  ski addict