Re: Roadmap: and now the positive side

[Chris B. Vetter]

On Thu, 23 Oct 2003 18:42:00 +0100 (BST)
Nicola Pero <nicola@brainstorm.co.uk> wrote:
> The main problem in GNUstep usually/often is an incarnation of the
> never-ending dycotomy -
>  * we're building a system better than the others
>  * we're building a portable, compatible system
> 'better than the others' usually imply 'different'.  'different'
> implies'incompatible'.  So the two objectives are somewhat
> contrasting.
> 'Difficult' doesn't mean impossible, also because there are so many
> nuances (how much/in which way better ?  how much/in which way
> compatible?).

Yes, but isn't that precisely the point? What you're saying is a bit
"wishy-washy" because it doesn't _clearly_ say what we really want.

If you take a look at questions users new to GNUstep pose, and look at
this particular thread, it should be obvious that we

  a) do not really know what GNUstep is supposed to be, because
  b) do not really know where it's headed (because of a)

Should we focus on "building something better" or more on "portability"?
As it looks, right now, we (that is GNUstep) try both, and apparently we
are getting nowhere. As I said, GNUstep is "drifting along" but not
"going forward".

Heck, we can't even agree, whether GNUstep is a bundle of libraries, an
API, an implementation of an API (OpenStep), a development or desktop
environment, or all of it, or something different entirely.

In my naivety, I think what is needed is a clearly outlined definition
what GNUstep is supposed to represent, eg. "A development environment
that implements the API as specified by OpenStep" and a roadmap ala
 - feature freeze
 - release version a
 - bug fixing
 - release version b
 - ensure a complete/correct implementation of A
 - release version c
 - add support for B
 - release version d

At least that would give people some clear direction and something to
look forward to use and work for.

As I said, that may be a bit too naive on my part...

-- 
Chris