[Joerg Mayer <jmayer@xxxxxxxxx>]

On Thu, 4 Mar 2004, Richard Sharpe wrote:
> I think we need some discussion around how we make such radical changes in
> future.

When I posted patches to implement shared libs I received some negative
feedback. I received no such feedback when I moved the dissection stuff
into libethereal. I also stated that I would wait until after the release
of 0.10.2 to put things into cvs. I did that. Because of the problem
feedback received on shared libs - I decided to start with the easy
part (famous last words).
If I had expected much trouble, I would have gone with the idea of a
branch, but it didn't look that problematic at the time I made that
decision and so I took the way that seemd the simpler one: make the
patch to head, fix some oversights and be done with it. No tracking
and porting of patches from head branch to shared branch.

Depending on how serious poeple feel the remaining problems are:
We can still create a branch and then back out the changes as they
are limited to files related to the build process. Please note
however, that this will not prevent people from having to update
their build tools - which seems to be at the core of your problems.
Also note, that I wonder whether more people would be willing to
test a developemnt branch that a patch sent to the list.

 Ciao
            Jï¿œrg

-- 
Joerg Mayer                                           <jmayer@xxxxxxxxx>
We are stuck with technology when what we really want is just stuff that
works. Some say that should read Microsoft instead of technology.