[OpenIndiana-discuss] Second Look at OI "Stable" Proposal (was Re: Proposal: OpenIndiana Stable Branch)

[Ken Gunderson]

On Sun, 2011-02-13 at 19:58 +0000, Deano wrote:
> Hi Ken,
> 
> Most of us agree that the more packages and more of a cross section the
> stable build has, the better. 
> However the main issue is supporting it, it's not simply packaging, its
> maintaining it to a high enough standard to be considered "stable", if
> someone can come forward and not only offer a package but be willing to look
> after it, follow the security alerts, package and patch it, then great.
> For now we have some great people supporting many packages and
> consolidation, but getting more into the stable list requires more people
> interested in helping.

I understand that.  Be that as it may, however, consider the perception
of a potential end user who hears about OI's "stable" server release and
takes if for a test spin, only to learn that there's not even one
single, solitary supported IMAP package.  Last I looked, email was #1
use of the Internet, followed by WWW.  Granted it's been a couple years
since I last checked, but if email is not #1 I'll wager it's close
behind as #2.

Hence, in m view, that seems like a huge sector to be blowing off.
Especially in light of comparatively rich support on the web framework
front; apache, php, java, tomcat, python, ruby, & perl.  As well as
having other major Internet service bases covered between Bind &
Sendmail, and not one, but two RDBMS systems.  An IMAP option is, ohmo,
is an essential component server side and the lack thereof not only
leaves a huge gaping hole, but also potentially paints a target for
negative evaluations and reviews.

I  don't really speak for the OI dev community so did not feel like it
was my place to solicit help from e.g. Cyrus dev community.  But I
suspect that if someone of a bit higher profile approached them they'd
be receptive.  Problem solved, no? 

I say Cyrus becuase I've used pretty much all of these at one time or
another and in modern times the choice pretty much boils down to Cyrus
or Dovecot.  Dovecot is less feature complete and under more active
development.  Cyrus would therefore likely involve less effort on the
pkg maintenance side.

My $0.02. I'm just trying to help.  

-- 
Ken Gunderson <kgunders at teamcool.net>