Among other things I want to prevent the crawling of feeds and searches. Also we recently decided to remove all tags from our pages. Crawlers tend to have long memories so they keep on trying to crawl now not existing pages and I want to prevent that too.
In most of these cases the URL's refer to another page (the page without /feed or the searched page) or to pages that do not exist (the tag pages). I do not think that in these cases the noindex directive can be used.
In any case the use of the noindex directive seems to be unpractical as it seems to involve a lot of manual HTML coding.