[Virtuoso-announce] Virtuoso Open-Source Edition, Version 6.1.3

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 454-5900

Hi,

OpenLink Software is pleased to announce the official release of 
Virtuoso Open-Source Edition, Version 6.1.3:

New product features as of March 30, 2011, V6.1.3, include:

  * Core Database Engine
    - Added configurable normalization of accented characters
    - Added global X509 CA repository
    - Added IMAP, IMAP-SSL and POP3-SSL  client support
    - Added http maintenance functions
    - Added optimization for case when there is no locality
    - Added optimization to look at GS index only
    - Added optimization to boolean expression evaluation
    - Added read-ahead on extent
    - Added retry deadlock if http session is not flushed
    - Added retry on deadlock when sponging on multiple threads
    - Added support for CORS on http endpoints
    - Added support for passing NVARCHAR arguments to thread-safe dicts
    - Added S/MIME encrypt and decrypt functions
    - Added SSL const decoration in explain
    - Added new aggregate functions primarily for use with SPARQL-BI
    - Added transitivity cache
    - Fixed busy loop when no free remaps
    - Fixed check for 5xx or 4xx status codes
    - Fixed check if subselect has a recursive ref
    - Fixed cost model to count inx cond and scale card at end
    - Fixed error on INSERT DATA with literals of non-default types
    - Fixed following checkpoint kill recovery remap extents may
      have allocated pages with no corresponding logical page because
      recovery performs unremap. Therefor if em remap page is allocated
      in em but free in dbs freemap, disregards and reset em allocation,
      if no logical page. If logical page, concider the dbs freeset
      in error and mark page as allocated.
    - Fixed increase delay on sequential deadlock
    - Fixed index_only on partial can return has data even if main row does not
    - Fixed infinite loop when dealing with string like abc&def and no token
    - Fixed issue parsing sparql function calls with zero arguments
    - Fixed issue with aggregates, distinct and ORDER BY
    - Fixed issue with closing XAConnection and connectionpool
    - Fixed issue with combination of UNION in the argument of IN
      causing General internal Optimized compiler error
    - Fixed issue with partial index when no path is followed
    - Fixed JDBC driver lost prepared data after execution query,
      so next execute of prepared query throws exception
    - Fixed memleaks
    - Fixed moved mutex lock/unlock calls to avoid deadlock
    - Fixed recovery mode
    - Fixed send body of 509 bandwidth error
    - Fixed set HTTP status to 503 when showing maintenance page
      and added a Retry-After: 1800 (seconds) header as a hint to the
      remote
    - Fixed set local timezone
    - Fixed size of input buffer of an XML parser is now adaptive
    - Updated documentation

  * SPARQL and RDF related
    - Added attribution ontology
    - Added Evri ontology
    - Added new Sponger Cartridges for EOL, Evri, Google Books, Groupon, hNews,
        hRecipe, ProductWiki, RDFa, RDFohloh, and many others
    - Added checking for RDFa DOCTYPE declaration
    - Added automatic wdrs:describedby relations for ingested RDF resources
    - Added expiration header handling re. Sponger cache invalidation scheme enhancements 
    - Added SPARQL-FED for Linked Data virtualization across 3rd party SPARQL endpoints
    - Added SPARQL, SPARQL-BI, and SPARQL-FED invocation from XSLT 
    - Added index to search over label text
    - Added initial codegen for ASK subquery
    - Added lax mode in csv import
    - Added MinExpiration param for Sponger
    - Added more prefixes
    - Added namespace prefix 'fn' for for XPATH 2.0 and XQuery 1.0
      Core Functions library
    - Added new SPARQL-to-SQL codegen for type cast functions
    - Added new SQL optimization to convert 
        sprintf ('format', col0, col1, ..., colN) = exnp 
	into more index friendly: 
	( col0 = sprintf_inverse ('format', expn, 2)[0] and 
	  col1 = sprintf_inverse ('format', expn, 2)[1] and ...  
	  colN = sprintf_inverse ('format', expn, 2)[N] )
    - Added optimization of SPARQL arithmatic and numerical comparisons
    - Added RDFa W3c validation button
    - Added rdfs schema
    - Added refresh=clean mode, to drop cache even if it is in the
      middle of sponging on another threads
    - Added scheduler task to keep volume under some limit
    - Added special immutable setting for inf rules
    - Added support for CSV as an output format
    - Added support for define input:inference in SPARQL DESCRIBE
    - Added support for define output:format "NT" for N-Triples
    - Added support for FILTER EXISTS and FILTER NOT EXISTS
    - Added support for URL template discovery via http://{virtuoso-instance-cname}/.well-known/host-meta
    - Added support for Insert, Update, Delete operations against XMLA endpoint
    - Added support for MINUS
    - Added support for posh triples
    - Added support for RDB2RDF
    - Added support of "rdfs:seeAlso" predicate in case of recursive sponging
    - Added WebID Protocol based ACL support across all SSL/TLS endpoints e.g. WebDAV, SPARQL, Sponger, other Virtuoso deployed Web Services
    - Fixed datatype for twitter
    - Fixed datetime format
    - Fixed do not use triple dicts on long files
    - Fixed do not check immutable graph when dba is running
    - Fixed DOI and RDFa handling in Sponger
    - Fixed enable xlat IRIs
    - Fixed hreview aggregates handling in Sponger Cartridge
    - Fixed issue with SPARQL CONSTRUCT/INSERT/DELETE/MODIFY queries
    - Fixed issue with sponging options and DESCRIBE on /sparql endpoint
    - Fixed issue with sparql FILTER with two OPTIONALS
    - Fixed issues reported by W3C Validator
    - Fixed issues with SPARQL-FED i.e., made it capable of accommodating rudimentary 3rd party SPARQL endpoints
    - Fixed missing column error in SPARQL-to-SQL preprocessor
    - Fixed missing namespace in </results>
    - Fixed security issue in SPARQL
    - Fixed sparql parse bug with NOT FROM NAMED queries like 
      select distinct ?g ?p not from named virtrdf: 
      where { graph ?g { ?s ?p ?o }}
    - Fixed sqltype and display size of DV_IRI_ID column
    - Fixed use common API to delete target
    - Fixed use doi proxy if no hslookup plugin
    - Fixed use dsn from content-location when it exists
    - Fixed use exclusive locks when doing SPARUL ops
    - Fixed when selection list contains expressions with constants
      do not inline OJ
    - Fixed small bugs

  * ODS Applications
    - Added ACL Inheritance
    - Added ACL semantic ping
    - Added group sharing
    - Added IMAP and IMAPS support
    - Added JSON format for WebFinger
    - Added language and type options for offers
    - Added 'Likes', 'Dislikes', 'Owns', 'Knows' and 'Social Network'
    - Added linkblog pages
    - Added login with Google account
    - Added microblog template
    - Added new index on OL accounts
    - Added new login and register library
    - Added profile security for offers, likes etc
    - Added progress bar, ping and ACL
    - Added RDF graph property for links
    - Added RDFa info
    - Added RSS and Atom imports
    - Added signed and/or encrypted mails
    - Added support for LinkedIn
    - Added support for Salmon Slap protocol re. notification services option
    - Added support for SWD
    - Added support for Twitter
    - Added UI for Login Authentication Keys
    - Added WebID to 'My Profile'
    - Fixed ACL handling
    - Fixed HTTPS handling
    - Fixed certificates
    - Fixed hyperlink session expire issue
    - Fixed openID server parameters extraction
    - Fixed RDF object properties
    - Fixed Twitter and LinkedIn registration/login
    - Fixed WebID for Briefcase instances
    - Fixed X.509 certificates in user pages
    - Fixed XRDS
    - Updated to CKeditor version 3.5.2