Wikimedia Bugzilla is closed!

Now XML-Dumps include only <namespaces>. They should also include <namespacealiases>, because it is a necessary information to work with the file.

There is no knowledge within a dump file which <namespacealiases> are in the system. To resolve this someone must have a online connection to the server to ask the api or ask the administrator.

i.e. The German Wiktionary changed 2009 the name space 104 from "WikiSaurus" to "Thesaurus" and made an namespacealias. Many links go today to WikiSaurus. In the system there is the alias and all is OK. But if you download the dump and after this have no internet connection / or you get the dump file per harddrive, you have no idea how to handle links to the namespace "WikiSaurus". Or with an offline-reader which import the dumpfile (like Wikitaxi).  There are many links to this namespace and all articles to which the link should go are there. But they can not come together. 

If there is the information of <namespacealiases> within the XML-Dump, the interpretation software can read this information and remap the namespace in their system.