Re: Generally accepted cut-off limit for -doc packages
On Tue, 6 May 2003, Martin Schulze wrote:
> It seems to me that this limit is not well documented. However, since
> James in his incarnation of the ftpmaster has the final say, I'd
> suppose to adjust both the developers reference and the
> new-maintainers guide to mention what Oliver Elphick said
>
> The point of splitting out a documentation package is to reduce
> download time for people who want to install a package without its
> documentation. This has to be balanced against the load on the
> archive of having extra packages, not to mention the frustration
> experienced by those who discover that they have got to load another
> package if they want the documentation.
>
> and the "generally accepted cut-off limit for -doc packages is ~500k
> of installed data" that ftpmasters implement.
But this is not true. More than 100 times ftpmaster implemented even much
smaller packages:
#!/bin/sh
grep -A 9 "Package:.*-doc" /var/lib/dpkg/available | \
grep "^[PI][an].*" | \
sed "/Package:/{;N;s/Package: \(.*\)\nInstalled-Size\(: .*\)/\1\2/;}" | \
grep -v "[0-9][0-9][0-9][0-9]" | \
grep -v "[4-9][0-9][0-9]" | \
cat > inst-size-doc
wc -l inst-size-doc
130 inst-size-doc
I can clearly understand that ftpmaster wants to prevent packages like dpsyco-doc,
but there are more examples for small documentation packages than I think the
term "generally accepted cut-off limit for -doc packages is ~500k of installed
data" would fit.
Or should we start mass-filing bug reports against those low volume doc packages?
> Somebody already voluntered for the developers reference and one of
> the editors of the new-maintainers guide also dropped in. That should
> help other people.
While I repeat that I would favour the documentation of this limit and I would
advise the maintainer who is sponsored by me to reintegrate the docs into the
package I think we should stall this effort until ftpmaster (now in CC because
he might perhaps not noticed the thread) speaks a clear word here.
Kind regards
Andreas.
Reply to: