next step forward (My understanding of the current situation, also an overview of previous discussion)

[Zhang Weiwu]

The discussion become difficult to track for people who just want to
have an overview, so I try to summarize what I understand now. I hope
people who follow my post be brief and ignore implementation details so
we can constantly focus on a framework.  I also hope my English is not a
problem and thanks for tolerating this.

GNUStep has been 15 years old, based on the design more than twenty
years ago and still constantly being improved. The project in general
suffer from lack of effective communication, management, coordination
and, perhaps caused by all former lackness, lack of development effort
including community and high quality code.

A product, be it opensource or commercial, can roughly have 3 aspects of
focus:

Product development
        Write source code, analysis user requirement to define and
        prioritize features, product design, co-ordinate development
        teams, finding/weighting/managing problems (bug tracking),
        communication and definition of a common view of the product,
        quality control, arrangement of development resources.

Marketing
        Find and defining target audiance, segment audiances and plan
        strategy accordingly, find the right way to reach and message
        the target audiance, co-ordinate public presentation of the
        teams and products, manage distribution and delivery (e.g.  let
        which product reach user, which reach developer), find partners
        and helpers.

Support
        Provide support faclity, organize support effort, define and
        manage communication strategy and policy.

The 3 aspects are closely connected, for example understanding of user's
requirement is foundation of all 3 aspect, support have direct
relationship with product development and is depended by marketing. So
it is unlikely a loose communicating team with ideas in separate parts
of separate aspects can solve the GNUStep problem.

All of these require innovation, execution power (a.k.a. decide
something and it really can be done), belief that things can be done,
and the willingness to work with other people, focusing so much on
improvements to be tolerating. In general, it is damn hard to make a
good product.

So far the discussion touched every aspect and every sub-item as I
listed in these aspects, although most discussion is on the first item
of the first aspect: writing code. There are a few things people can
agree on /now/, e.g.

* the code in multiple parts need to be polished and an agreed
        coding convention is needed. It is benificial if gnustep and
        desktop environment based on it can agree on the same coding
        convention.
* the website should be reworked. It should be more useful and
        give positive impression to visitors, at minimum that things are
        bing improved
* the effort in mulitple related sub-projects should be
  co-ordinated
* effective measurement must be taken to get developers attention
        on gnustep and gets them to use it

Although in certain bigger area we couldn't get agreed opinion yet.

* should gnustep target not only developers but also users, and
        what developers / users should we target to.
* competibility with other DE.
* user interface
* reference platform / compatibility with MacOS

I suggest

* people discuss with the frame work we said (3 aspects etc) in
  mind so we can work on the whole pictuer instead of pieces of
  jigsaw
* discussion on-topic, each topic should be discussed in their own
  threads
* work towards to have some kind of management commitee which
        deliver a plan for gnustep roadmap and discuss the roadmap. Note
        this is not a roadmap of product, but a roadmap in gnustep in
        general, including bug tracking, website management framework
        etc.
* set goals and organize people work on them. Keep managed,
        communicated etc. Some kind of management commitee is needed for
        this.

I think it is easy to a) have an opinion on any topic but difficult to
b) make a really good product. I am trying to push from the current
situation a), to get result b).

To go from a) to b) I have this in my head:
1. Throw all the topics for discussion, one per thread, so people have
time to
   comment, show opinion and reason it;
2. Then people compile roadmaps. I would if I can get some time doing it and
   the level of understanding that actually can make it;
3. Organize a commitee to run the roadmap;



_______________________________________________
Discuss-gnustep mailing list
Discuss-gnustep@...
http://lists.gnu.org/mailman/listinfo/discuss-gnustep