Re: Proposed: evolution

[Dan Winship <danw novell com>]

On Tue, 2004-07-20 at 15:39 -0400, Seth Nickell wrote:
> What concerns me is the precedent this sets for Evolution itself. Is
> this indicative of a general problem we are going to have where the
> whole platform gets a change made but evolution hangs back? Its not a
> particularly hard change to make technically, but Evolution appears to
> have a requirement the rest of the platform does not have ("runs on
> old GNOME releases").

The requirement is not
        
        (Released versions of) Evolution must run on old GNOME releases

it's

        *Beta* versions of Evolution must run on the most popular
        releases of the most popular distros, so that we will have
        enough people testing it to find lots of bugs

At the moment, that means "Evolution 1.5 must not require GNOME 2.6".

To some extent, this is because 2.0 was a fairly ambitious release, and
since future releases will have to be smaller in scope to keep up with
the GNOME release cycle, there won't be as many changes and they won't
need as much testing, so always requiring the latest-and-greatest won't
be as much of an issue.

But then again, maybe it will still be an issue. Evolution is huge, and
unlike other large things in the GNOME platform like gtk+ and nautilus,
it has competition. Everyone running a beta release of GNOME ends up
testing gtk+ and nautilus, but not everyone running beta GNOMEs is going
to end up testing evolution. So it may continue to be the case in the
future that the pool of people-who-run-GNOME-betas isn't large enough to
ensure that evolution gets enough testing to meet GNOME release goodness
criteria, in which case we'd want to make sure that people running non-
beta versions of GNOME could test it too. But this wouldn't be
evolution's fault. The same thing would end up happening for other very
large apps accepted into the platform too.

[As for why an #ifdef patch didn't get committed sooner, I dunno.]

-- Dan