[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#144046: general: Sections are not finely grained

> > I think it would be better to drop the sections altogether and use a
> > keyword-based system someone suggested a few months ago. The advantages

i suggested this a few months ago. unfortunately i havn't reworked my
proposal yet, nor did i make a proof of concept especially for my new
enhancements. (a kind of proof-of-concept can be seen in aptitude, as
well as on  http://people.debian.org/~erich/packagebrowser/
which basically has the algorithms required for the first stage.
(second stage will need a useful amount of "tags" added to packages)

> >  - ultimate fine-grainedness (?)

indeed. like selecting gpl licenced apps only, gtk apps only etc.

> >  - no dillemas about where to put packages which fit in more than
> >    section (like x11 net-related programs)


> 	Users need a hierachical layout in order to find software. Keyword

correct, but the keywords don't need to be hierarchically themselves.
You just have to give the user a way to browse the keywords
hierarchically. (that way different users can even have different trees,
which is a pro against hardwired hierarchies)

> by themselves are not that much useful since they would be only appropiate
> to the language used. Several disadvantages:

> 1.- more difficult to translate than sections

Don't think so. I think they'll be much simpler, as you don't have to
say which mail tools go in there and which not, these keyword-tags are
much simpler than the categories.

> 2.- are not organised hierarchicaly (sp?)

external hierarchies have always been included in my concept.

> 3.- difficult to represent graphically in a package-administration gui
> (sections are easily represented as trees).

you can represent this as tree easily. Have a look at the url i posted
above. Packages (and even sub-trees) appear multiple times.
For example
  Programming -> IDE -> VI
  Editors -> VI
  Editors -> IDE -> VI
are all the same.
Definitely a pro with such flexible tree hierarchies.

> 	If you want to have a keyword-based system I would suggest you
> take a look at dpkg-iasearch (yes, not documented, but it's a proof of

As i understand this works by kind of indexing the descriptions.
Which is by-design inferior to hand-tuned keywords imho.

what i'm planning to add to my proposal is the use of weighted keywords.
Such as "progress 0.8" and "licence:free 1.0", so i could select only
apps that are considered to be quite useable by the maintainer and
dfsg-free.  unfortunately this will lead to further bloat of the
packages files. :-(

The most difficult part will be an intuitive user interface. But i think
that a user interface doesn't need to implement all features directly.
(for example these weights could be in some extended selection menu
only, and influence only the sorting by default)

Erich Schubert

erich@(mucl.de|debian.org)        --        GPG Key ID: 4B3A135C
A polar bear is a rectangular bear after a coordinate transform.
Die kürzeste Verbindung zwischen zwei Menschen ist ein Lächeln.

To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Reply to: