The limit for a languagepack's name is 100 characters. However, the limit for the base_image_id of an assembly is 36, so it's possible to create a languagepack that cannot be used by an app at deployment.
ERROR: (DataError) (1406, "Data too long for column 'base_image_id' at row 1") 'INSERT INTO image (created_at, updated_at, uuid, name, source_uri, source_format, description, project_id, user_id, tags, state, base_image_id, created_image_id, image_format, artifact_type, external_ref) VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)' (datetime.datetime(2015, 3, 5, 22, 21, 8, 473721), None, '39fcb449-d356-4bcc-a2a4-92f9ecd960e8', 'test_deploy_w_java_language_pack_6568bbd4', 'https://github.com/arbor-test/tomcat-app', 'heroku', None, '938200', 'c37655011d28466989ec4ad4d80993af', None, 'PENDING', 'test_deploy_w_java_language_pack_6568bbd4', None, 'vm', None, None)