Tim Churches wrote:
> The experience of the GNUmed project with the open source
> wxWindows/wxPython as a cross-platform GUI environment has not been a
> happy one. 
These statements are correct but the causal link is not.
To a certain extent the choice of platform is almost
irrelevant, if the organisational apparatus is there
you will eventually succeed.
As Horst has pointed out, nowadays it requires deliberate effort
not to write cross-platform software,  .NET runs on Linux, Firefox is ahead
in the web-standards game, etc.

Gnumed's problem has been a lack of a clear solution for key technical
choices around internal architecture and how data moves around from
database to client and back. It's boring and obtuse, but crucial for
multiple
developers to be able to co-operate on the project. Again, what's
important is
not so much the actual technical decision, but the fact that it is made,
clearly documented,
and adhered to.

My 0.02c: getting external players on board such as divisions, unis and
Colleges is well and good,
but it should occur a posteriori, not a priori. What I mean is, a basic
working system exists,
documented inter-module APIs and the GUI-to-database pathway are clearly
documented, *then* others can
come on board, the point is development can continue while we are
waiting, and doesn't stop
if (when) they decide to pull out or don't play a productive role.

I started a few pages on the OzDocIT a few months ago to jump start
exactly this process. As
expected, I ended up with 7 or so proposals, with one vote each!

The simplest solution is a 'dictator' type, which most other open-source
projects use, who
can declare solutions by fiat when consensus doesn't emerge naturally.
Horst is the obvious
choice, but the limitation has been his time-availability, it's
difficult to play this role unless
the person is actively involved in coding.
The alternative is a technical panel, which is model followed by
FreeBSD, postgres and a few
other projects, who agree to be bound by a majority vote of the panel.
Is their support for this? Who should be on the panel? I'll kick things
off my nominating Horst, Syan Tan, Tim Churches and Richard Terry.

Ian H
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk

Reply via email to