Re: Misclassification of packages; "libs" and "doc" sections
> > And BTW, about this `X axis'... You need some kind of qualification. It's
> > true, shadows of gray prevent us from drawing perfect lines that would
> > divide among packages... but we need those lines somewhere though.
>
> Why do we need those lines? When was the last time sections helped YOU?
> When was the last time they made your life more difficult?
So you propose a big bag of unclassified packages. That would be worse than
what we have now... I guess that for new users, who go thtough the package
web pages, or through dselect, or through the ftp sites, sections are
useful. Or you are in favor of a pure "set of package properties" approach?
This could go hand in hand with a "canonical" qualification...
I think the current problem is just that the sections are not well defined,
e.g.: X11 should be for packages related to X11/gnome/kde infrastructure,
not for packages that happens to have X as their interface.. look at this...
Section: x11
Package: njplot
Description: [Biology] A tree drawing program
NJplot is able to draw any binary tree expressed
in the standard phylogenetic tree format (e.g., the format used by the
Phylip package). NJplot is especially convenient for rooting the unrooted
trees obtained from parsimony, distance or maximum
likelihood tree-building methods.
... this would go in the "science" section.
Reply to: