Re: How does a port become real?
Josip Rodin <joy@srce.hr> writes:
> On Mon, Feb 16, 2004 at 12:19:18AM +0100, Goswin von Brederlow wrote:
> > > > > > > We haven't done anything like this, and I would expect that such an
> > > > > > > undertaking (the accompanying conversion and accomodating changes)
> > > > > > > would take years to get done properly...
> > > > > >
> > > > > > Woody did it before, converting to pools. The transition was finalized
> > > > > > end of last year when the last of potato was removed I think.
> > > > >
> > > > > No, it didn't, the Packages files changed paths but no data was moved out of
> > > > > the same toplevel directory.
> > > >
> > > > I don't thing anything else is planed for "segmenting" the mirror.
> > >
> > > Well, it better be, because it's silly to believe that the mirrors
> > > should just keep updating their exclusion lists or available space
> > > whenever we want to add an architecture... the ftpmasters are quite
> > > correct to be hesitant of adding new architectures because we don't
> > > handle the space issue well enough with the current set, let alone
> > > with new stuff being added.
> >
> > Pool could be split up into pool/all/, pool/<arch>/. Doesn't mean the
> > Packages files mechanism has to change. Thats what I ment.
>
> Mirror scripts would still need adjustment to change the exclusion
> procedure.
debmirror probably not :)
But thats why I want details.
MfG
Goswin
Reply to: