Abstract : W3C XML Schemas specifications were published in 2001, but a large community still uses DTDs. Perhaps they didn't adopt new schema technologies because they are still awaiting some missing features ? What kind of feature would be useful to DTD users and schema practitioners ? Let's transpose the buzzword "mashups" to schema technologies: mashups of several schema languages and mashups of schema languages with non-schema languages. From this bazaar, we'll discuss some of the emerging features that could make what would be the next generation of schema languages: dynamicity, semantic, mixins, and hyperschemas.