On Fri, May 18, 2007 at 01:40:11PM -0500, Steve Stallion wrote:
> On 5/18/07, Brian Gupta <[EMAIL PROTECTED]> wrote:
> >Why don't we first focus on those applications that don't exist in
> >Solaris currently? Also, in parallel, we should work on an updated
> >packaging format.
> 
> Whoa... slow down Tiger.
> 
> SVR4 is by far a more complete and cohesive packaging format than is
> available on any other UNIX or Linux distribution. Essentially there
> is no technical merit in tossing SVR4 packaging - not by a long shot.

What? Can you please compare what you think makes it complete? I mean
explain what you mean by the format? This assertion is hard to swallow
without your explaining what your experience is and what leads you to
say this.

> That said, SVR4 could certainly use a little help with respect to
> usability (pkgadd -D anyone?).
> 
> This was posted earlier, but please take an indepth look into the
> packaging guide and see what is provided out of the box - I think you
> will agree this beats the pants off of RPM, .deb, and POTS (plain old
> tarballs).

I don't see any really strong recommendations for the format
there. 61-characters for the NAME, population of dependencies left up
to the developer and no structure given to express dependancies, 256
characters for DESC, VERSION is defined as X.Y[.Z] and REV= is
similarly limited. There's the 16-character CATEGORY, which isn't
enough to allow people to create readable categories.

There is a lot of cruft, like HOTLINE, ULIMIT, ORDER, etc. Why there
is even a MAXINST parameter is a puzzle to me. 

There aren't good installation and removal script hooks. With
equivelants of rpm's %pre and %post macros things like ISTATES and
RSTATES could be done away with easily (as well as MAXINST, come to
think of it).

Why are SUNW_OBSOLETES, SUNW_REQUIRES, and SUNW_INCOMPAT restricted to
patches?

After going over that document, which expands on some of the packaging
work on solaris I've done in the past, I've got to say that it's a
pretty pale shadow of rpm and deb, and a scratch above POTS.

The philosophy of using patches means that the provisions for patches
is a big advantage, but except for that I'd have to give it a D.

-Peter

-- 
The 5 year plan:
In five years we'll make up another plan.
Or just re-use this one.

_______________________________________________
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org

Reply via email to