Use key().name() instead of link_id This is now possible because key_name is constructed purely from the key fields of an entity. It is not sufficient to use just link_id, that works only for single-scoped entities (e.g., those that either do not have a scope, or that have a scope which itself does not have a scope). It would break if there was an entity that has a scoped scope (it would only include the scope's link_id in the url, which made it impossible to look up the scope as we missed the link_id of the scope's scope). Patch by: Sverre Rabbelier

     1

# Directions for web crawlers.

     2

# See http://www.robotstxt.org/wc/norobots.html.

     3

     4

User-agent: HTTrack

     5

User-agent: puf

     6

User-agent: MSIECrawler

     7

User-agent: Nutch

     8

Disallow: /