cross-site bug tracking

[Luis Villa]

On 4/12/07, Greg Dekoenigsberg <gdk at redhat.com> wrote:
> > Well, Canonical has. I think their execution has been poor, but they
> > have had the right vision for several years now.
>
> Their vision is wrong because, no matter the architecture decisions
> they've made, they *think* they're smart enough to build it as a
> proprietary application -- and they're not.  This is *exactly* the class
> of application that *must* be open source if it's got any hope of success.

Method of production is orthogonal to goals of production. Don't
confuse the two, or make the mistake of writing off what they are
trying to do just because how they are doing it is dumb.

[Or to put it another way- RH has gone a long, long time with a
proprietary build system. There is very little reason to think that
Canonical can't similarly succeed with a proprietary development
infrastructure.]

> > (I think GNOME would be happy to switch away from bugzilla to
> > something distributed, but has ~ 0 manpower to write the tool in the
> > first place. You obviously know mozilla better than I do, but I'd be
> > shocked if they are ready to move away from bugzilla- too tied into
> > it.)
>
> And this is the problem in a nutshell.  Everyone agrees with the idea in
> theory, but no one has the manpower to make it happen in practice.

No one besides Canonical (and maybe rpath). They may not be doing it
well, but they are doing it, and they will get smarter with time. Red
Hat and Novell should see this as a serious threat to their business
models, and should be investing appropriately.

Luis