Pre-3.6 code freeze

[Patrick R. Michaud]

On Wed, Jul 20, 2011 at 08:35:14AM -0400, Andrew Whitworth wrote:
> First question is,
> of course, what is our current list of target systems? That's a list
> that we should be more proactive in defining, and more conscientious
> about maintaining. It may very well be that we no longer consider
> Windows XP to be a "supported platform". Microsoft no longer supports
> it, and without a platform champion maybe we don't want to either.

It's defined already -- from docs/project/support_policy.pod:

    We support recent versions of the three major operating system 
    families: GNU/Linux, Mac OS X, and Microsoft Windows.  
    Any version less than two years old counts as "recent".

> Things like making hardlinks, or filenames with strange encoding
> shouldn't be in Parrot core if they aren't available on all our target
> platforms. 

I've always felt that part of what makes Parrot attractive to
HLL devs is that it can take care of many of these sorts of
compatibility issues for us.  Is fine if some of them get pushed
out to libraries... but if that just means the libraries are
even less supported than the core, it's not a win.

> Also, with the ability to get NCI access to the necessary
> underlying functions, maybe we don't need a built-in ability at all:

Given the history and current state of NCI support (especially
its inability to access arbitrary C functions), I'd be a bit
hesitant to accept this approach as a viable workaround.

Pm