Re: debian archive disk space requirements.
- To: "Marcelo E. Magallon" <mmagallo@debian.org>
- Cc: debian-devel@lists.debian.org, joshk@triplehelix.org
- Subject: Re: debian archive disk space requirements.
- From: Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de>
- Date: 01 Sep 2003 12:19:36 +0200
- Message-id: <[🔎] 87bru4n747.fsf@mrvn.homelinux.org>
- In-reply-to: <20030831201230.GA24479@informatik.uni-stuttgart.de>
- References: <20030607003824.GA14627@daedalus.andrew.net.au> <20030607183328.GA16088@prvidomaci.srce.hr> <20030607233157.GA7432@snoopy.apana.org.au> <20030608101238.GB589@prvidomaci.srce.hr> <Pine.GSO.4.50.0308301136460.12344-100000@luna.rtfmconsult.com> <20030830101850.GD27871@prvidomaci.srce.hr> <20030830174147.GA10739@informatik.uni-stuttgart.de> <87znhpq4uq.fsf@mrvn.homelinux.org> <20030831190524.GA24126@informatik.uni-stuttgart.de> <20030831195619.GL2718@triplehelix.org> <20030831201230.GA24479@informatik.uni-stuttgart.de>
"Marcelo E. Magallon" <mmagallo@debian.org> writes:
> On Sun, Aug 31, 2003 at 12:56:19PM -0700, Joshua Kwan wrote:
>
> > > [alpha] 7283
> > > [i386] 7889
> >
> > I'd not include non-free in the batch because many non-free on i386 are
> > i386 only, i.e. binary driver installers.
>
> Ah, well spotted, but still there's some significant difference:
>
> mmagallo@auric:~/ftp/dists$ zcat testing{,-proposed-updates}/main/binary-alpha/Packages.gz | grep-dctrl -s Package -F Architecture -v all | wc -l
> 7105
> mmagallo@auric:~/ftp/dists$ zcat testing{,-proposed-updates}/main/binary-i386/Packages.gz | grep-dctrl -s Package -F Architecture -v all | wc -l
> 7628
>
> A lot of that is due to the gazillion kernel flavors and their modules,
> some compilers, some sensors and sensor-related stuff, a few acpi
> packages, old libc5 stuff, and the good old "uh?" stuff (firebird for
> example).
How much space do those 500 odd packages take up?
Given a 50% sizce increase on binaries alpha should have another 1.8G
of debs. If those 500 packages make up 1.2G (+50%=1.8G) then the 50%
claim would be right.
Slightly less for less % increases, you do the math.
MfG
Goswin
Reply to: