Re: Status of 'sarge' for the amd64 architecture
On Mon, Apr 25, 2005 at 09:24:28AM -0500, Steve Greenland wrote:
> On 23-Apr-05, 17:24 (CDT), Andreas Barth <aba@not.so.argh.org> wrote:
> > Beyond the fact that it is too late to add another architecture for
> > sarge, removing arm from sarge does not make the mirror pulses much
> > smaller - and AFAIK the size of the mirror pulses is the main problem.
>
> See, that just makes no sense whatsover. You can claim either:
>
> 1) Adding AMD64 would increase the mirror load unacceptably
>
> OR
>
> 2) Removing ARM would not have a significant effect on the mirror load
>
>
> but not both at the same time. (Yes, I realize that two different people
> made these claims, but which one am I supposed to believe?)
Sure one can: removing arm would be removing arm from testing, not from
unstable. Currently around 800 (of the > 8000) source packages in
testing have a different version compared to unstable, even less so
different arm packages (think big arch:all packages, or arch:i386
specific packages). This means that removing arm from testing would
only save about 1GB of space. That's only half of what today's mirror
pulse will be, for example. Adding amd64 would however require all the
amd64 .deb's to be added, I don't know *how* many it is, but probably
about 10 times as much, certainly more than the space saving by removing
arm.
--Jeroen
--
Jeroen van Wolffelaar
Jeroen@wolffelaar.nl (also for Jabber & MSN; ICQ: 33944357)
http://Jeroen.A-Eskwadraat.nl
Reply to: