Enrico Zini wrote: >On Wed, Jul 27, 2005 at 11:26:06AM +0200, Steffen Moeller wrote: > > >>distinction should be made on a higher level. And should it be >>field::medicine::biology or rather field::biology::medicine? The latter >>would make no sense in the way the CDD is currently been set up. >> >> >This is difficult and interesting: anyone minds if I forward the whole >thread to [EMAIL PROTECTED] > > It is a perfect topic for an everlasting flame bait. Please spread it.
I think we can well decide if a single package has something to do with genetics or bioinformatics or with physics, though we cannot decide, from the user's perspective, how one domain is related to the other. Such relationships are however determined by the tools suggesting, recommending or extending each other. Again, these links will be bidirectinal no matter what fields:: are being scetched. Hence, the specialisation of the fields:: classification we should strive to make irrelevant, creating tools the present the world to the user as he is likely to perceive it, with dynamic interdependencies on the basis of the tools that are installed already. Like, a matrix of all-against-all in terms of debian/control-based interpackage links and order the tools or the fields 2 or 3dimensionally after principal component analysis. Sounds like an interesting and fairly quick research project to me. Many greetings Steffen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
