> > Uh, we're already doing that. Very few Debian resources are spent on > > non-DFSG-free stuff. A single day's uploads takes more disk space and > > bandwidth than the entirety of non-free. None of the regular maintenance > > work is non-free specific.
On Mon, Jan 05, 2004 at 10:58:57AM -0500, Dale E Martin wrote: > Yes, my point was that there would be little practical impact and you seem > to agree with that in part. There would be very little practical benefit to dropping non free. That's not the same as saying that there would be very little practical impact [unless you ignore the impact on the users, and on the people who support non-free]. > > Which means the only resources we can "concentrate" are our servers, not > > our developers' time, which means we get _no_ benefit from this, as far > > as I can see. > > The only benefit anyone can argue is philosophical. (Well, see below for > an actual practical benefit.) We have something called the DFSG, and we (as > an organization, not as individuals necessarily) will only support software > that conforms to the DFSG if we drop non-free. Oddly enough, the DFSG was originally a part of the social contract. Fundamentally: the social contract says that we're interested in supporting users of free software, and the DFSG is the part which defines what we mean by "free". The philosophical benefit, I think, lies not in better understanding the DFSG, but in other influences (perhaps, as someone else has indicated, this is based on some of the writings published by the FSF -- but even there, what people have been proposing is not very consistent with those writings). > > non-free maintainers have to setup their own archives, > > There are tons of those already, many in wide use. download.kde.org, > backports.org, and the bunk backport collection all come to mind. I don't > see this as big stumbling block. Both of which represent far larger efforts than non-free. Or is your point that you think non-free should be a larger effort? > > contrib becomes at best much harder to support well and at worst > > unsupported. > > The contrib issue is one that I have not mulled over too much. If we're > interested in "moral purity" (which really is what this is about I think - > as a project are we interested in it?), I suppose we'd ditch contrib too? > That's a real sticky one. Well... taking a statement written to be a practical best-effort at describing how we work and turning it into a moral code and then using that moral code to justify changing that original statement *is* sticky. -- Raul
