Dell's Linux Survey

[nodata]

Am Mittwoch, den 21.03.2007, 13:25 -0400 schrieb Matthew Flaschen:
> nodata wrote:
> > Am Dienstag, den 20.03.2007, 19:28 -0400 schrieb Matthew Flaschen:
> >> nodata wrote:
> >>> Am Dienstag, den 20.03.2007, 13:45 -0500 schrieb Dick Dowdell:
> >>>> FYI - Industry pundit on the topic:
> >>>> http://www.infoworld.com/article/07/03/19/12OPopenent_1.html?source=NLC-OPENT&cgd=2007-03-20
> >>> It's a shame the survey was structured how it was.
> >>>
> >>> Warren Togami has it right - focus on upstream. See his blog:
> >>>  http://wtogami.livejournal.com/15101.html
> >> Realistically, if they rigorously tested against any major distribution,
> >> the rest would also work in most cases.
> >>
> >> Matthew Flaschen
> >>
> > 
> > Whereas if they tested against upstream, it would work in all
> > distributions, unless the distro broke something.
> 
> But every distro breaks/changes/customizes something upstream.

We are talking about kernels here.

Distros tend to cherry pick patches from newer kernels, and backport
security patches to their stable kernels.

In Dell's case, they want their hardware working and to stay working, so
however you look at it, they need it to work upstream.

> That's
> part of why people choose the distro they do.  Moreover, testing against
> upstream is more work for Dell then testing against a single distro.

It should be less work, because they don't have to keep checking that
their non-upstreamed changes still work.

Why do you say it would be more work?