StreamBase New and Noteworthy Archives

The JMS Adapter was updated to support the Simple message type. The JMS Configuration File Editor was updated to support Simple messages, and the adapter's documentation was updated to match.

The StreamBase Exegy adapters now rely on version 3.5.0 or later of the JAR file that implements the Exegy Client API, XCAPI.jar. As before, this JAR file is provided as part of your Exegy installation and is not shipped with StreamBase.

As of TIBCO StreamBase release 7.4.5:

When a custom function is called with null values for arguments that expect a primitive type, execution continues, but now a console warning is issued and a tuple is emitted on the error port of the operator that invoked the function. Formerly, when given null values for such arguments, custom functions silently returned nulls.

The TIBCO StreamBase® 29West LBM Subscribe ACK Output Adapter enables a StreamBase application to acknowledge incoming LBM messages based on a topic and sequence number by performing an ACK operation when explicit ACK is enabled.

The TIBCO StreamBase® Adapter for JMS reader now partially supports CLIENT_ACKNOWLEDGE, TIBCO EMS' EXPLICIT_CLIENT_ACKNOWLEDGE and Apache ActiveMQ's INDIVIDUAL_ACKNOWLEDGE. If one of these modes is selected, the adapter will wait until the tuple is fully processed by the application before acknowledging the message. The StreamBase application cannot trigger an explicit acknowledgment of the message.

You can now provide an optional third argument to format_time() that specifies a time zone location (such as "US/Hawaii") or abbreviation ("PST"), as was already possible for the strftime function. For details, see the documentation for format_time.

User-defined aggregate functions can now accept array, list, and variable-length inputs.

The new property Set undefined numeric values to zero allows the TIBCO StreamBase® Adapter for Activ Financial to disambiguate unchanged data from missing data.

Previously when the value of a field in an incoming update message changed from defined to undefined, the adapter set that field's value in the resulting tuple to null. Fields missing in the update message are also set to null to indicate that the field's value has not changed since the last update. To enable applications to differentiate between undefined fields and unchanged fields, the new property has been added to the adapter to cause it to set undefined numeric fields to 0 instead of null. By default, the property is not selected, preserving the adapter's legacy behavior.

The default time format for the CSV File Reader and CSV Socket Reader input adapters was changed to use 24 hour time. Formerly, the default format was:

yyyy-MM-dd hh:mm:ss.SSSZ

The new default format is the following, with the hours field now uppercase:

yyyy-MM-dd HH:mm:ss.SSSZ

The StreamBase documentation for release 7.4.3 included all the documentation updates reported for release 7.4.2. Specific updates include:

This release includes a new graphical editor for configuring the properties of JMS and EMS adapters that makes it easier to specify and reuse name maps and other configuration elements. The editor can also import settings from TIBCO Businessworks™ applications. The new UI eliminates the need to hand-edit .jmsconf XML files under most circumstances. For more information, see StreamBase JMS Configuration File Editor in the StreamBase Studio Reference Guide.

To more easily reuse Java operators and adapters across application contexts, StreamBase Studio now allows you to enter parameters and expressions for Java Adapter and Operator properties that do not use text boxes. In such components, all check boxes, radio buttons, and drop-down menus now optionally display a small T icon . Clicking the icon converts the control to a text field where you can enter the name of a module parameter in ${param} format or a StreamBase expression in =expr format.

By default, the icons are hidden. To display them, clear a new preference check box, Hide Control to toggle to/from text fields in Java Operator Properties, found under Window → Preferences → StreamBase Studio → Authoring. A keyboard shortcut is also available.

Property views of standard StreamBase operators and data constructs do not have this feature. For further information, see Parameterizing Java Operator and Adapter Properties.

To provide complete access to an sbd process by a remote debugger, a launch configuration can now specify that the server's JVM be suspended when the process is launched. The JVM then resumes once the remote debugger attaches to the process. To enable this, use the new Suspend JVM Initially check box in conjunction with the Allow Remote Java Debugging check box when creating a launch configuration. For more information about how this works, see Editing Launch Configurations.

As of this release, the Pattern operator's polling-interval parameter is no longer supported, and causes an error if it is present. To correct your applications, open your EventFlow module's .sbapp file in a text editor and search for the string polling-interval. If found, remove the entire <param> element that contains that string.

The simple function nan() returns a NaN literal as type double. This is the same value returned from the expression sqrt(-1).

The QueryTopN sample was revised and made easier to use by giving its Query Table a simpler schema and making the table self-loading. The sample lets you specify how many top values to output and formats them both as tuples and as a list of tuples.

Two new Java functions were added to the Java Simple Function sample to demonstrate how to handle variable-length argument lists.

Eclipse users who wish to incorporate StreamBase Studio as an Eclipse plug-in can now obtain the latest Studio Eclipse update site as a downloadable zip archive available from the TIBCO distribution site. For more information, see StreamBase Studio Eclipse Update Site. The StreamBase Studio plug-ins can now be used with Eclipse versions 3.7, 3.8, 4.3, and 4.4.

The TIBCO StreamBase® Adapter for CitiFX Options FIX allows a StreamBase application to connect to and exchange FIX messages with the following Citi Options product. See its documentation here.

An option was added to The TIBCO StreamBase File Reader adapter to read a file as binary data into a single tuple as a blob field.

You can now provide an attachment to the Email Sender as a blob, which can be simpler than providing a local file name. In addition, you can also now specify each attachment's content type and encoding.

The following updates were added to the TIBCO StreamBase® Adapter for Bloomberg B-Pipe:

The TIBCO StreamBase® Adapter for ICAP/EBS was updated to use version 1.6 of the EBS FIX specification.

The time-getting functions get_second() ... get_year() now take an optional second argument, a string literal, for example "UTC" or "Europe/Zurich". When called with one (timestamp) argument, the functions return local time. Providing a recognized time zone ID as the second argument returns the requested unit of time for the specified time zone. Use the Java function TimeZone.getAvailableIDs() to get a list of all time zone IDs. For more information, see Specifying Time Zones.

In addition, the strftime() function has a new optional argument to format timestamps for a specified time zone instead of local time.

Like aggregatelist(), the new aggregatelistnonnull() function creates lists of expressions in the aggregate's window, but excludes nulls, emulating the syntax filternull(aggregatelist(expr)).

The new version() function returns the version of the currently executing StreamBase CEP as a string.

The TIBCO StreamBase® Twitter input and output adapters enable a StreamBase application to interact with the Twitter service using the twitter4j library. The adapters receive, send and retweet tweets and issue other commands, such as searching and filtering.

The new TIBCO StreamBase input and output adapters for KCG Hotspot FX allows a StreamBase application to connect to and exchange FIX messages with KCG Hotspot BookFeed Market Data and Order Entry And Execution services. See the Adapters Guide for more information.

To provide more useful diagnostics, SBUnit test failures with multiple missing/unexpected tuples now throw a ComparisonFailure instead of an AssertionFailedError.

StreamBase kits now install runtime components of JDK 1.7_55 (also called 7u55).

Running the Type Coercion Sample shows how same-named numeric field values with certain compatible types can all appear to be carried by a new field with a type compatible with all original values.

The DebuggingBreakpoint sample illustrates setting unconditional and conditional breakpoints on data arcs when debugging applications in StreamBase Studio.

The StreamBase sample explaining how to use JDBC data sources has been completely revised to be more informative. It now describes using JDBC tables in a trading context, illustrates the use of stored procedures, and provides templates for connecting to several databases. As before, you must configure a database in order to run the sample application.

The StreamBase documentation for release 7.4.1 includes all the documentation updates reported for release 7.3.11.

Like aggregatelist(), the new aggregatelistnonnull() function creates lists of expressions in the aggregate's window, but excludes nulls, emulating the syntax filternull(aggregatelist(expr)).

Eclipse users who wish to incorporate StreamBase Studio as an Eclipse plug-in can now obtain the latest Studio Eclipse update site online, as a downloadable zip archive available from the TIBCO distribution site. For more information, see StreamBase Studio Eclipse Update Site. Installing Studio 7.3.12 as an Eclipse plug-in requires Eclipse versions 3.7 or 3.8.

A reference data output port and schema specification were added to the TIBCO StreamBase® Adapter for Bloomberg B-Pipe to enable it to respond to reference data requests. See the adapter's page in the Adapters Guide for details.

Users of the TIBCO StreamBase® Adapter for Bloomberg B-Pipe can now specify which of five types of authentication to use. See the the adapter's page in the Adapters Guide table for details.

The TIBCO StreamBase 29West LBM Subscribe ACK Output Adapter enables a StreamBase application to acknowledge incoming LBM messages based on a topic and sequence number by performing an ACK operation when explicit ACK is enabled.

An option was added to The TIBCO StreamBase File Reader adapter to read a file as binary data into a single tuple as a a blob field.

The default time format for the CSV File Reader and CSV Socket Reader input adapters was changed to use 24 hour time. Formerly, the default format was:

yyyy-MM-dd hh:mm:ss.SSSZ

The new default format is the following, with the hours field now uppercase:

yyyy-MM-dd HH:mm:ss.SSSZ

The Service Provider Status for the TIBCO StreamBase Adapter for Thomson Reuters Enterprise Platform for Real-Time subscribe has been updated to handle the new Thomson Reuters message structure while remaining compatible with the old structure.

The schema for market data requests in the UBS FIX adapter was updated to match version 1.19 of the UBS specification.

The ICAP/EBS adapter was updated to use version 1.6 of the EBS FIX specification.

The StreamBase JDBC Query sample explaining how to use JDBC data sources was completely revised to be more informative. It now describes using JDBC tables in a trading context, illustrates the use of stored procedures, and provides templates for connecting to several databases. As before, you must configure a database in order to run the sample application.

The StreamBase TopN Query sample was revised and made easier to use by making its Query Table self-loading and with a simpler schema. The sample lets you specify how many top values to output and formats them both as tuples and as a list of tuples.

The StreamBase documentation for release 7.3.12 includes all the documentation updates reported for release 7.2.15. Specific updates include:

Four related properties added to the XML-to-Tuple operator provide greater control over converting XML elements to fields:

See Using the XML to Tuple Operator in the Authoring Guide for more information.

Users of FIX Adapters and Trading Components Handlers now can set the log level of a venue at runtime by issuing setLogLevel commands to a control port. Tuple values for the optional Log input field can dynamically set a log level and also control filtering of incoming and outgoing messages. For details, see Adapter Ports And Schemas in the FIX Adapter documentation.

The TIBCO StreamBase® Adapter for TIBCO Enterprise Message Service™ allows a StreamBase application to connect to TIBCO EMS servers and exchange JMS messages. Note that unlike JMS adapters, which can connect to any number of JMS providers, the EMS adapters only enable connections to TIBCO EMS servers.

The TIBCO StreamBase® Email Sender adapter now supports attachments, identified by a static set of file names, a list of file names provided in a designated input field, or both.

The StreamBase Studio XML editor now provides syntax highlighting and autocompletion when editing JMSReader and JMSWriter adapter .jmsconf configuration files.

The following properties were added to the File Reader Input adapter:

The StreamBase documentation for release 7.3.11 includes all the documentation updates reported for release 7.2.15.

The StreamBase Admin operator now has an optional status port that can be enabled to allow the operator to output status tuples to give the user more information during processing. Status tuples for the operator will be emitted for errors and rejections of input tuples. A rejection status tuple will be emitted if there is a problem validating the input tuple. An error status tuple will be emitted if there is an error during processing. Note that errors will also be emitted from the Error Output port, if it is enabled.

The OSX shortcut to refresh a Studio project's Typecheck Environment has changed to Cmd+R, replacing the previous key binding of Cmd+F5, as Apple uses that to toggle VoiceOver.

Searching for appropriate functions in the StreamBase Studio Function Chooser yields more pertinent results because it now matches your search term with any sub-string of a function name, not just initial sub-strings.

The new parsejson_loose() function works similarly to the existing parsejson() function, but can (within reason) ignore JSON elements that do not conform to the schema definition provided.

StreamBase Server configuration files can now have an optional key-password child XML element under the ssl-authentication XML element. Call the new Java Client API method, StreamBaseURI.getKeyPassphrase(), to return the passphrase of the keystore key (defined by the keypass URI parameter). The existing StreamBaseURI.getKeystorePassphrase() method returns the passphrase of the keystore file (defined by the keystorepass URI parameter). For backward compatibility, the keystore password is used to access the key if the key password (the keystorepass parameter) is not present. In that case, StreamBaseURI.getKeystorePassphrase() returns value of the keypass URI parameter.

For details, see the StreamBase Server Configuration File XML Reference and the Java API Guide documentation.

You can now format timestamps output from the TIBCO StreamBase CSV File Writer. The new Timestamp Format field on the Record Formatting tab specifies the formatting of timestamps when Add Timestamp is set to Prepend or Append. The default format is yyyy-MM-dd hh:mm:ss.SSSZ.

The LDAP Authentication sample demonstrates using Active Directory to limit access to certain StreamBase commands to users who can provide authentication credentials associated with one of the StreamBase roles.

The TIBCO StreamBase® Database Change Data Capture Input Adapter allows a StreamBase application to monitor changes to an Oracle or SQL Server database. It detects when records are inserted, updated, or deleted in a database table and can be customized to report changes to particular table columns.

The new TIBCO StreamBase® File Writer Output Adapter writes tuples or single fields from a stream to a file. A sample application has been added that illustrates its use.

The TIBCO StreamBase® File Reader adapter now supports GZIP-compressed files. In addition, its Control and Status ports now have properties that let them be disabled. They are still enabled by default.

The TIBCO StreamBase® Web Reader input adapter has new functionality and associated properties:

The EBS Live XML input adapter now has an optional status port which, if enabled, emits tuples on connection state transitions. Reconnect logic is included in this update.

The TIBCO StreamBase® Log File Output Adapter has a new option allowing the end user to compress frequent messages. Activating the Compress Frequent Messages property compresses messages within a specified number of seconds (the Max window size property). If the Ignore Timestamp Fields option is enabled, compression will take place without regard to timestamp fields, which can change for each tuple.

In addition, new properties have been added for setting the logger name and for specifying a custom format using StreamBase expression language. For details, see the Log Output Adapter documentation.

The IBM WebSphere MQ Output Adapter now has a Queue When Connection Down property that causes it to queue rather than reject tuples when the connection to the MQ server is down. When the connection is restored, the queued tuples get processed.

The TIBCO StreamBase® Adapter for Thomson Reuters Enterprise Platform for Real-Time subscribe now has a Display Already Subscribed Messages At Debug Level property on its Debug tab that causes already-subscribed messages to be logged at DEBUG rather than INFO level.

The TIBCO StreamBase® StreamBase-to-StreamBase input and output adapter host and port properties now default to localhost and 10000, respectively.

Two optional special fields were added to the TIBCO StreamBase® ActivFeed input adapter: UPDATE_EVENT_TYPE and UPDATE_EVENT_TYPE_NAME, which receive the numeric value and name of the event type of update messages.

The TIBCO StreamBase® RSS Reader now has status and command ports. It can also now insert the RSS Feed link into each item output to allow tracing back to the feed it came from.

Support was removed for connectivity to Sybase RAP native databases.

The StreamBase documentation for release 7.3.10 includes all the documentation updates reported for release 7.2.15.

The CSV file reader adapter now has a Preserve Whitespace property, which, when enabled, preserves leading and trailing white space in string fields. For backward compatibility, the property is disabled by default.

The following pairs of adapters have been added to enable two-way socket communication for various data encodings:

See the Adapters Guide for details on the new adapters.

Support for the following adapters is now deprecated:

The StreamBase documentation for release 7.3.9 includes all the documentation updates reported for release 7.2.15.

The FTP Operator in StreamBase Studio now has equivalent functionality to the SBX FTP operator. The new features are:

When using JMS container connections, in the event the JMS server becomes unavailable the container will retry 10 times to reconnect, waiting 5 seconds between each try. You can override this behavior by setting the following values in sbd.sbconf:

<jms-connections-defaults> 
   <param name="server-num-retries" value="XX" /> 
   <param name="server-reconnect-interval" value="YY" /> 
</jms-connections-defaults> 

XX is the number of retries (0 means no retries, max value is 2147483647) and YY is the number of seconds to wait between retries.

If you want to disable the Connection Timeout property of a TIBCO StreamBase® Output Adapter for TIBCO Spotfire® Automation Services, You can now specify 0 for its value.

The File Reader input adapter has a new property, One Tuple Per Line, which is disabled by default. If you enable it, a tuple is emitted for each line read from the file. When disabled, the entire file contents is emitted in a single tuple.

The following pairs of adapters have been added to enable two-way socket communication for various data encodings:

See the Adapters Guide for details on the new adapters.

Support for the following adapters is now deprecated:

If you use Zing, run the sb-install-zing-system-support command to update to the new version of Zing System Tools (ZST).

Due to limitations in some implementations of the rpm command, the TIBCO StreamBase Zing RPM no longer performs a check of Zing system support during installation, because it fails on certain versions of Red Hat Linux. To be confident about your Zing system configuration after an installation or upgrade, run /opt/streambase/bin/sb-install-zing-system-support check and pay attention to the messages that it issues.

The StreamBase documentation for release 7.3.8 includes all the documentation updates reported for release 7.2.15.

StreamBase 7.3.7 delivers new operators:

StreamBase 7.3.7 delivers several new operators:

StreamBase 7.3.7 delivers new input and output adapters.

LiveView has adapters that simplify using LiveView Server with EventFlow applications.

The product sample demonstrating the adapters is available to customers who have LiveView 1.5.1 or higher installed.

When creating a feed simulation from a CSV file, the Feed Simulation Editor now can match column headers to field names even when they are ordered differently. Fields of type tuple can have subtuple values out of order in the CSV file as long as its column headers use fully qualified names (using dot notation, e.g. Price.Open and Price.Close) that match or closely approximate field names. You can still reassign columns to fields in the Data File Options Dialog if the editor makes matching errors.

When setting up a feed simulation from a CSV file, the option Map data file columns to sub-fields now only appears when the stream's schema includes sub-tuples and a CSV file is selected. By default, the option is enabled if the CSV file contains flattened data and not enabled if it contains hierarchical data. These changes decrease the chances of setting the option incorrectly.

The StreamBase Unit Test Wizard now has improved default settings.

See Creating and Running StreamBase JUnit Tests for more information on StreamBase JUnit tests.

In the .NET Client API, the Timestamp class has been changed to better handle conversions to and from .NET DateTime objects. Specifically:

The Runtime Schema Cast Operator in the Java Operators group, which casts its input schema to its configured target output schema at run time, now has its own icon, as shown here.

Placing a copy of a .sbconf file that you have edited to include customized defaults into your SBHOME/etc/ folder and renaming it to studio-sbd.conf now enables the Configuration File Wizard to copy it into a project instead of generating a skeleton configuration file with standard defaults. For more information, see Cloning Customized Configuration File Default Contents in Studio.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_45. (which originally was incorrectly announced in release 7.3.4 as 1.6.0_43). As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

StreamBase Studio now ships with the default Java Compiler Compliance Level set to 1.6, the minimum level when developing Java code for StreamBase projects. If you install Studio as plug-ins to Eclipse, ensure that Java code developed for StreamBase server target Java 1.6. Users of Java 1.5 must upgrade to 1.6, as StreamBase no longer supports JDK 1.5.

StreamBase Java-based commands such as jsbc, jsbadmin, and sbprofile now have a default maximum heap size of 1 GB when launched on 64 bit systems.

A new system property, streambase.decouple.operator.log-level, enables operator and adapter logging levels to be set independently of the system logging level. Its default is false, which keeps the old behavior whereby operators and adapters use the more verbose setting of the component's or the system log-level property value. When set to true, operator and adapter logging verbosity can be set lower than the system logging level.

The log adapter now has an Emit Logged Tuples property, which is enabled by default. Disabling this property causes the adapter not to have an output port and not to emit logged tuples.

The TIBCO Rendezvous Subscribing Input Adapter now has a Use Unique Queue property, which, if enabled, causes the adapter to use a per-adapter-instance event queue rather than the default Tibrv event queue. Using per-instance queues can lower latency in some situations.

The Use Formatted Timestamps checkbox has been deprecated.

A new Timestamp option has been added as a drop-down list box, with choices Prepend, Append and None (default). Use this control to optionally add a timestamp to the beginning or end of each CSV row.

Adding timestamps enables applications to record data from any input or output stream in a top level application or exposed, in a format suitable for playback. Use the timestamped CSV files for feed simulations to play back streams you recorded in real time.

While an application is running, you can tell sbadmin to capture CSV files from input or output streams by issuing commands such as:

For more information, see CSV File Writer Output Adapter.

As an alternative to specifying fields in a Custom Message Format via a numeric index ({0}, {1}, etc.), you can now use field names ({price}, {symbol}, etc.) directly. Custom Message formatting supports escaped field name identifiers but not expressions.

The StreamBaseFIX configuration file now supports an entry named <dictionary-file> (which is only meaningful when <dictionary> is also specified) to allow users to specify a non-standard name or location for the dictionary file. Previously, StreamBase assumed that custom dictionaries were located in the fixaj-dictionaries folder.

The Trading Components DB-Rapid venue can now output missing depth tiers as zero price items as part of the DepthOfBook message when an item becomes stale. This functionality is disabled by default. Add the parameter OutputMissingTiersAsZeroPriceItems to the instance and set it to true to enable this feature.

A new option on the Adapter Properties tab of the TREP-RT Subscribing adapter, Include AttribInfo In All Updates, adds AttribInfo into all update and refresh messages sent to its subscribers. Receiving this data causes the publishing adapter to include AttribInfo in all messages.

The Tuple Properties tab has two new options:

When looking up fields, the Thomson Reuters Schema Designer now automatically detects fields with fractional pricing, and adds both the double and string fields to the available fields users can select to create a schema. For example, if the BID field indicates a price of 88 3/32, the field list will now contain a BID field of type double and also a BID_as_string field of type string. The Thomson Reuters subscribe adapter now also automatically maps any field in the schema with an "_as_string" suffix to the string representation of that field.

The Exegy library has been updated to version 3.1.11. You must upgrade your installation's Exegy library to this version in order to use this adapter.

A new property, Include Subscribe Pass-Thru Field, has been added. When enabled, the Subscribe stream schema is expected to include a field of any type named PassThru, which is is passed through to the Equities, PriceBook, and Commodities output ports.

A new property, Include Basket Pass-Thru Field, has been added. When enabled, the Basket stream schema is expected to include a field of any type named PassThru, which is is passed through to the BasketUpdates and BasketNav output ports.

For other Exegy input adapter updates, see Fixed in 7.2.15 release notes.

The IBM WebSphere Output Adapter has a new property, Quick Connection-Down Detection. If enabled, the adapter polls the base queue periodically to quickly detect connection-down conditions.

The Wombat Publishing adapter can now send messages using any of the ARRAY types, as well as the VECTOR_MSG, and VECTOR_STRING type. Types are mapped as follows:

As of this release, these FIX engines are no longer supported. Use StreamBaseFIX or QuickFIX/J instead.

The StreamBase documentation for release 7.3.7 includes all the documentation updates reported for release 7.2.15.

Java operators and adapters may now have properties of type string[][]. Only lists of pairs of strings are supported (that is, string[n][2]). Use this property type to allow a user to provide a list of string pairs without requiring unique key values, such as Map property types must have.

The following enhancements were made:

The Morningstar Quotes Feed Input Adapter has new properties that allow you to connect and disconnect from the Morningstar servers without stopping your application.

The StreamBase documentation for release 7.3.6 includes all the documentation updates reported for release 7.2.14.

A new Edit Dimension dialog for the Aggregate operator predicate dimension clarifies options for closing and emitting tuples for the current group and all groups of windows. In addition, all emit and close predicates can now be either simple or aggregate expressions. The new dialog is illustrated and described in the updated page Aggregate Operator: Predicate-Based Dimension Options in the StreamBase Studio Authoring Guide.

Beside pressing the zoom-to-fit icon in the StreamBase Studio editor to show the contents of an entire EventFlow diagram, you can now press Ctrl-0 (zero). On the Mac, press Command-0.

You can now display notes in EventFlow diagrams using any available font. The default is the font used to display text in the editor view. To change the font used for notes, go to Window → Preferences, then select General Appearance → Colors and Font → StreamBase → EventFlow Editor → EventFlow Editor Notes, and then click Edit. The example note displayed below uses the Segoe Script font.

This preference is global, so that all notes in all projects display the selected font until you change it.

The debugging Variables view includes a row for each tuple field. Previous to this release, the text appearing in a row's value cell was a complete representation of the data it contained, except for data in cells containing blobs, which was not displayed at all. Creating textual representations of large sub-tuples and lists could cause debugger performance problems, because the time to calculate strings was proportional to the size of the fields. Now, text for fields of type sub-tuple and list is intentionally abbreviated, and the code to calculate that text executes much faster, both in StreamBase Studio and StreamBase server. Also, cells containing blobs now indicate their content. You can control how that content displays by setting preferences. Go to Window → Preferences → StreamBase Studio → Test/Debug to set the Display blobs as hex or ASCII characters option and specify the Maximum blob characters to display.

You can always view the full values of abbreviated tuples by selecting a row in the Variables view. The Details pane at the bottom of the Variables view displays a textual representation of all of the selected row's data.

Version 7.3.5 ships with a local repository of Apache Maven artifacts for StreamBase. There are artifacts for the StreamBase Client library, the StreamBase Test Support library, and a third one for the embedded server. These artifacts are provided under your StreamBase installation directory, in STREAMBASE_HOME/sdk/maven/repo.

For information on using this repository with your own projects, see StreamBase Maven Artifacts in the API Guide.

In previous releases, the vwap function would return NaN if all volumes were zero or if there were no entries in the window. The function has been updated to return null, matching the behavior of avg.

The Log adapter has two new options:

The SNMP input adapter allows a StreamBase application to act as an SNMP manager for receiving SNMP v1 and v2c traps.

The Syslog input adapter allows a StreamBase application to act as a Syslog collector for receiving syslog messages.

The TREP-RT RMDS subscribe adapter now has a Download Service(s) property that lets users restrict the set of Thomson Reuters services from which data dictionaries are downloaded. In addition, Multiple instances of the adapter within a StreamBase application can now share downloaded and locally-loaded data dictionaries.

The RFA/Java library, rfa.jar, used by TREP adapters has been upgraded to version 7.2.1.

The Regex File Reader adapter now has an optional event port, enabled by a new Start Event Port property. It also has a Map Control Port to Event Port property, which can be used to include control port input tuples in event port tuples. The Start Control Port and Start Event Port properties must be enabled to enable the Map Control Port to Event Port property.

The embedded Web Reader input adapter now supports HTTP POST as well as HTTP GET. Two new examples were added to the Web Reader Samples illustrating how to handle POST and SOAP requests.

The XML to Tuple and Tuple to XML Operator Sample group has been reorganized as five sample applications, four of which have been redesigned:

The two -attribute variants illustrate generating and parsing XML attributes. The two -datatype variants illustrate using fields with a variety of StreamBase data types, including primitive types, lists, and nested tuples. xml2tuple-fixml.sbapp demonstrates how to parse a FIXML message.

The two tuple2xml samples now have two input streams each, one which runs through a map operator to supply default values for any tuple fields containing null, and a second which leaves the user's input unchanged. The three xml2tuple sample applications can now be fed from either new canned XML file read by the file reader adapter or from a user-supplied XML string through an alternate input stream (which is used by the automated tests as well). A note on each sample sbapp canvas describes what the sample does and how the canned data is supplied.

Calling an Embedded StreamBase Server from Java, has been added to the Client Libraries group of StreamBase samples that demonstrates how to run applications in an embedded StreamBase Server inside a JVM. Loading the sample opens a cheatsheet to guide you through it.

The functions getInputTuplesEnqueuedTotal() and getInputTuplesDequeuedTotal() in class OperatorInfo have been deprecated and will be removed in a future release. Their replacement functions, getTotalInputTuplesDequeued() and getTotalInputTuplesEnqueued(), provide the same information but return longs instead of ints. Functions have also been added for getting tuple counts for individual ports.

If you use the bundled Zing JVM from Azul Systems, you must upgrade the Zing system with the sb-install-zing-system-support install command, because the Zing license used with the version requires it. If you do not, you will get a license failure on StreamBase startup.

The Trading Components Framework now provides extra data for swaps and forwards. To support this change, the FX handlers of the Trading Components Framework have an additional schema field, SettlDateFar, for the settlement date for the far leg of swaps.

The StreamBase documentation for release 7.3.5 includes all the documentation updates reported for release 7.2.13, as well as the following:

A new set of functions and built-in language features ease the use of StreamBase blobs to store BSON (binary serializations of JSON-like documents) data. Within BSON blobs, StreamBase ints, longs, doubles, strings, timestamps, and blobs map to their respective BSON scalar types, nulls map to nulls, lists map to arrays, and tuples map to objects. The BSON functions consist of:

Associated language features include:

You can find the BSON specification at http://bsonspec.org.

The Gain GTX FIX adapter allows a StreamBase application to connect to the trading infrastructure provided by GAIN Capital, and to exchange FIX messages with it.

The File Reader input adapter reads a text file and emits a tuple with a string field containing the entire contents of the file. The adapter can be configured to read a file at start-up or in response to the receipt of a tuple on its control input port.

As deploying Microsoft .NET client applications from StreamBase Studio has been greatly simplified, it is no longer necessary to use the StreamBase Redistributable Installer. The simplified deployment process was introduced in StreamBase 7.3.0, and requires a .NET 4.0 or later framework on each client machine you deploy .NET applications to. As of this release, the Redistributable Installer has been eliminated and the .NET client documentation updated accordingly.

For information on .NET application deployment, see Creating StreamBase .NET Clients in the StreamBase documentation.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_45. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The StreamBase documentation for release 7.3.4 includeed all the documentation updates reported for release 7.2.11, as well as the following:

StreamBase Studio now has a redesigned splash screen and program icon.

The StreamBase documentation for release 7.3.3 includes all the documentation updates reported for release 7.2.11, as well as the following:

This release includes a preview of the .NET operator, which is expected to be added to a future release. StreamBase Studio projects can reference Microsoft .NET assemblies with the .NET operator on the Operators and Adapters palette (keyboard shortcut O.). It is a global Java operator that can load client code in the form of a .NET assembly, instantiate a class within that assembly, pass it incoming tuples, and optionally emit tuples provided by the .NET client code. For more information, see Using the Microsoft .NET Operator in the Authoring Guide. Contact StreamBase Support if you have questions or feedback regarding this feature. Until it is fully released, do not use this feature in production deployments.

In addition to being able to set the speed of a feed simulation with a slider, you can now type a speed factor in the text field to the right of the slider, which is now editable. This lets you simulate at any relative feed rate. The rate you specify can be higher than 25x, which is still the maximum that the slider enables.

The Manual Input view now has a round button at at the upper right of the toolbar for manually enqueuing JSON data. Switch to this mode to enter one or more tuples in JSON syntax. Send the data with the Send Data button or by pressing CTRL+Enter.

You can now control how the Log Output Adapter treats null values within string fields of tuples it logs to the Console. That is, a field of type string that contains a null character can be displayed in the Console as either null or "null". The quoted form is ambiguous, as it could either signify a string field has a null value or is a string containing the characters null. Now, in the Adapter Properties tab, when you select the Display Tuple Field Names check box, a new check box beneath it is enabled. This check box, Quote Null Strings, is by default selected, so as to preserve the adapter's existing behavior. To omit quotes around null strings, turn off the Quote Null Strings option by clearing the check box.

When comparing a timestamp with a string literal, the operators ==, !=, <=, >=, <, and > now attempt to interpret the string literal as a time. If this fails, the comparison fails to typecheck. If the string literal does not contain an explicit time zone, it is assumed to have the time zone set in the operating system of the computer compiling the application.

The tuple-to-XML operator now has an Attribute Fields property that can be used to specify a comma-delimited list of fully-qualified tuple field names that should generate XML attributes rather than sub-elements. For example, if the XML generated by the operator for a given input tuple is:

<Top><s>This value comes from tuple field Top.s of type string</s></Top>

Adding Top.s to the Attribute Fields property would result in the following XML:

<Top s="This value comes from tuple field Top.s of type string"></Top>

Aggregate now has predicates to emit and close all windows. You can use this feature in an application that groups tuples to clean up when a specified condition occurs (such as market close), by emitting and/or closing all the windows.

The following Boolean utility string functions have been added:

When you use sbbundle to create a bundle that includes projects in StreamBase Studio that use Eclipse Classpath Variables, you can identify one or more of them using the new --classpath-variable (-V) option. The option is described in Application Bundling.

The StreamBase documentation for release 7.3.2 includes all the documentation updates reported for release 7.2.11, as well as the following:

The Aggregate operator has two new features for this release. The first is that the input1. prefix is still accepted but no longer required when entering field names in an expression for a predicate-based dimension. The second feature is a new check box in the Edit Dimension dialog for predicate-based dimensions, Evaluate Close and Emit predicates before adding the new tuple to the window. The new check box is described on the Predicate-Based Dimension Options page.

In StreamBase 7.3 and later, when queues begin to grow, each additional buffer is twice the previous buffer size, until a default maximum of 1 GB per buffer is reached. After the limit is reached, each additional buffer requested is capped at the limit size. You can change the limit size by setting the following Java property:

streambase.runtime.stream-queue-max-buffer

See StreamBase Java System Properties for more information on this property.

The output from the sbprofile command now includes the number of tuples delivered for each input/output stream.

The StreamBase expression language includes a new aggregate function, lag(), which returns the next-to-last value for field f in an Aggregate operator's current window or in a Query Table column. If the window or table column has only one row or zero rows, lag() returns null. The function accepts all data types and returns the same type as its argument.

This release increments the Python language support in StreamBase in two ways:

See Creating StreamBase Python Clients for details.

New support was added to com.streambase.sb.operator.parameter.UIHints that allows adapter and operator writers to offer String value proposals to the user when typing into a text field. See the UIHints#setProposals() Javadoc for details.

The Saxo Bank FIX adapter allows a StreamBase application to connect to the trading infrastructure provided by Saxo Bank, and to exchange FIX messages with it.

Several standard adapters now have two new controls on the Adapter Properties tab of their Properties views. There is a new check box, Use Default Charset, which is selected by default (with one exception, see the XML File Writer adapter note in the next paragraph). When selected, the default behavior of the adapter is to use the JVM default character set. When you want to configure an application to use Unicode support, you need to configure the adapter in addition to configuring StreamBase Server and Studio. When the check box is cleared, you must specify a standard Java character set in the new Character Set field.

This applies to the following adapters: CSV File Reader, CSV Socket Reader, CSV File Writer, CSV Socket Writer, Regular Expressions File Reader, Regular Expressions Socket Reader, and the Web Reader. The exception is the XML File Writer adapter, which does have the new controls, but whose default remains the US-ASCII character set.

The CSV Socket Reader Input adapter was updated to provide a server mode where the adapter listens on a specified port for incoming connections. In addition, when acting as a client, the adapter now attempts to reconnect if connections fail, using a specified wait period between reconnection attempts. The adapter also has new properties for specifying how to handle incomplete records, and whether to log incomplete record events.

The sample for the CSV File Reader adapter was simplified for clarity. See CSV File Reader Input Adapter Sample.

The CSV File Reader Input adapter and CSV Socket Reader Input adapter have a new property, Lenient Parsing, whose default value is true.

The FIX Output adapter and all adapters that inherit from the FIX Output adapter were updated with the ability to track acknowledgment of the receipt of sent messages. This feature is described in Requesting Outgoing Message Send Confirmation on the FIX Output Adapter page.

The data dictionary for the Trading Technologies adapter was updated to implement version 7.8.1 of the Trading Technologies FIX specification.

The Thomson Reuters TREP-RT Subscribing Input adapter was updated such that CLOSED events, including access denied errors, are now logged at the INFO level, in addition to being sent to the adapter's event port.

For release 7.3.1, StreamBase Studio is based on Eclipse 3.8.1, which provides some underlying Eclipse bug fixes over Eclipse 3.8.0. When installing Studio in the form of plug-ins for a separate copy of Eclipse, you must start with Eclipse 3.8.1. Earlier Eclipse versions are no longer supported for that purpose.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_37. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The StreamBase documentation for release 7.3.1 includes all the documentation updates reported for release 7.2.10, as well as the following:

A new system property, streambase.decouple.operator.log-level, enables operator and adapter logging levels to be set independently of the system logging level. Its default is false, which keeps the old behavior whereby operators and adapters use the more verbose setting of the component's or the system log-level property value. When set to true, operator and adapter logging verbosity can be set lower than the system logging level.

In the .NET Client API, the Timestamp class has been changed to better handle conversions to and from .NET DateTime objects. Specific changes are:

The Use Formatted Timestamps checkbox has been deprecated.

A new Timestamp option has been added as a drop-down list box, with choices Prepend, Append and None (default). Use this control to optionally add a timestamp to the beginning or end of each CSV row.

Adding timestamps enables applications to record data from any input or output stream in a top level application or exposed, in a format suitable for playback. Use the timestamped CSV files for feed simulations to play back streams you recorded in real time.

While an application is running, you can tell sbadmin to capture CSV files from input or output streams by issuing commands such as:

For more information, see CSV File Writer Output Adapter.

A new option on the Adapter Properties tab of the TREP-RT Subscribing adapter, Include AttribInfo In All Updates, adds AttribInfo into all update and refresh messages sent to its subscribers. Receiving this data causes the publishing adapter to include AttribInfo in all messages.

The Tuple Properties tab has two new options:

The Trading Components DB-Rapid venue can now output missing depth tiers as zero price items as part of the DepthOfBook message when an item becomes stale. This functionality is disabled by default. Add the parameter OutputMissingTiersAsZeroPriceItems to the instance and set it to true to enable this feature.

When looking up fields, the Thomson Reuters Schema Designer now automatically detects fields with fractional pricing, and adds both the double and string fields to the available fields users can select to create a schema. For example, if the BID field indicates a price of 88 3/32, the field list will now contain a BID field of type double and also a BID_as_string field of type string. The Thomson Reuters subscribe adapter now also automatically maps any field in the schema with an "_as_string" suffix to the string representation of that field.

The Exegy library has been updated to version 3.1.11. You must upgrade your installation's Exegy library to this version in order to use this adapter.

A new property, Include Subscribe Pass-Thru Field, has been added. When enabled, the Subscribe stream schema is expected to include a field of any type named PassThru, which is is passed through to the Equities, PriceBook, and Commodities output ports.

A new property, Include Basket Pass-Thru Field, has been added. When enabled, the Basket stream schema is expected to include a field of any type named PassThru, which is is passed through to the BasketUpdates and BasketNav output ports.

For other Exegy input adapter updates, see Fixed in 7.2.15 release notes.

The Wombat Publishing adapter can now send messages using any of the ARRAY types, as well as the VECTOR_MSG, and VECTOR_STRING type. Types are mapped as follows:

The XML to Tuple and Tuple to XML Operator Sample group has been reorganized as five sample applications, four of which have been redesigned:

The two -attribute variants illustrate generating and parsing XML attributes. The two -datatype variants illustrate using fields with a variety of StreamBase data types, including primitive types, lists, and nested tuples. xml2tuple-fixml.sbapp demonstrates how to parse a FIXML message.

The two tuple2xml samples now have two input streams each, one which runs through a map operator to supply default values for any tuple fields containing null, and a second which leaves the user's input unchanged. The three xml2tuple sample applications can now be fed from either new canned XML file read by the file reader adapter or from a user-supplied XML string through an alternate input stream (which is used by the automated tests as well). A note on each sample sbapp canvas describes what the sample does and how the canned data is supplied.

The embedded Web Reader input adapter now supports HTTP POST as well as HTTP GET. Two new examples were added to the Web Reader Samples illustrating how to handle POST and SOAP requests.

The Trading Components Framework now provides extra data for swaps and forwards. To support this change, the FX handlers of the Trading Components Framework have an additional schema field, SettlDateFar, for the settlement date for the far leg of swaps.

The following enhancements were made:

StreamBase now has a Deutsche Bank AutobahnFX Classic FIX adapter, with accompanying market data and execution handlers. The Deutsche Bank AutobahnFX Rapid FIX adapter and handler, which was named Deutsche Bank FIX in prior StreamBase releases, is now called Deutsche Bank AutobahnFX Rapid FIX Adapter to distinguish it from the new Deutsche Bank Classic FIX adapter and handlers.

The Log output adapter now has an option to override the Log Level you set for the adapter on a per-tuple basis. In the Log Adapter Properties pane, use the Log Level Override Field to specify the name of a field in a logged stream that contains log level information. The field can contain any of the following values: Error, Warn, Info, Debug, or Trace. Any other value, including null, causes the adapter use the configured log level for that tuple.

The File Reader input adapter reads a text file and emits a tuple with a string field containing the entire contents of the file. The adapter can be configured to read a file at start-up or in response to the receipt of a tuple on its control input port.

The TREP-RT RMDS subscribe adapter now has a Download Service(s) property that lets users restrict the set of Thomson Reuters services from which data dictionaries are downloaded. In addition, multiple instances of the adapter within a StreamBase application can now share downloaded and locally-loaded data dictionaries.

A new command, ListInstruments, has been added to the FAST-based adapters, BM&FBovespa UMDF Input adapter and CME Market Data FIX/FAST adapter. The ListInstruments command will cause the adapter to list all the instruments available on a given channel by emitting one tuple per defined instrument on the Instrument Definitions output port.

The RFA/Java library, rfa.jar, used by TREP adapters has been upgraded to version 7.2.1.

The Gain GTX FIX adapter allows a StreamBase application to connect to the trading infrastructure provided by GAIN Capital, and to exchange FIX messages with it.

The File Reader input adapter reads a text file and emits a tuple with a string field containing the entire contents of the file. The adapter can be configured to read a file at start-up or in response to the receipt of a tuple on its control input port.

The new Bloomberg Server API input adapter is an embedded adapter that can connect to a Bloomberg Server API process to subscribe and get data from the process.

The new Bloomberg B-Pipe input adapter is an embedded adapter that can connect to a Bloomberg B-Pipe process to subscribe and get data from the process.

The new Deutsche Bank AutobahnFX FIX adapter allows a StreamBase application to connect to the Deutsche Bank AutobahnFX trading infrastructure and to exchange FIX messages with it.

Additional properties to support XML parsing were added to the IBM WebSphere MQ input adapter and Wall Street Systems input adapter. In addition, the existing and new XML parsing properties have been placed in a new XML Parsing tab. The properties on the new tab include:

Thomson Reuters Enterprise Platform for Real-Time (TREP-RT) subscribing adapter has two new properties:

The Saxo Bank FIX market data handler now has mass quote support.

The sbadmin operator now has username and password properties for use with local connections. See the sbadmin reference page for more information.

The tuple-to-XML operator now has an Attribute Fields property that can be used to specify a comma-delimited list of fully-qualified tuple field names that should generate XML attributes rather than sub-elements. For example, if the XML generated by the operator for a given input tuple is:

<Top><s>This value comes from tuple field Top.s of type string</s></Top>

Adding Top.s to the Attribute Fields property would result in the following XML:

<Top s="This value comes from tuple field Top.s of type string"></Top>

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_45. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. See Supported Configurations.

In previous releases, the option to have an adapter or operator start with an application was chosen with a checkbox. This option is now a field with allowed values yes, no, or a module parameter that resolves to either true or false.

Adapters in a application that have Start with application set to false now have a visual indicator () on the adapter icon. Input adapters display the decorator on the bottom-right corner of the adapter icon; output adapters display the decorator on the top-left corner of the adapter icon.

You can create a named schema from an existing schema directly from the application view by right-clicking any arc and selecting Refactor → Extract Named Schema.

Event Flow application streams that are exposed for enqueue or dequeue are now indicated by a decoration on the stream's icon:

The output from the sbprofile command now includes the number of tuples delivered for each input/output stream.

The Feed Simulation Data File Options dialog now has a vertical scroll bar to enable use on small displays.

Copying and pasting multiple tuples with the Copy as JSON option now formats multiple tuples as:

[tuple1,tuple2,...,tuplen]

In previous releases, multiple tuples were separated with carriage returns.

The drop-down menu near the upper right corner of the Variables view of StreamBase Studio's debug perspective gives you the option to hide null values in tuples fields, list elements, and dynamic variables. See Viewing Tuple Contents for more information.

The Barclays BARX FIX adapter allows a StreamBase application to connect to the BARX trading infrastructure provided by Barclays and exchange FIX messages with it.

The 29West LBM Subscribing Input Adapter now has the following new features:

To use the LBM Adapters from StreamBase 7.2.11 and above, ensure you have LBM version 6.0 or above. The StreamBase installation kit includes version 6.0 of the UMS_6.0_jdk1.5.0_12.jar JAR file.

The CSV File Reader Input adapter and CSV Socket Reader Input adapter have a new property, Lenient Parsing, whose default value is true.

The Trading Components framework now supports market data and execution handlers for UBS.

When a container, A, is added to a running StreamBase Server, and that container has a connection to the system.control stream, container A no longer receives the STARTING tuple on the control stream. However, other control-connected containers in the same Server instance do receive the STARTING tuple for container A. See The Container Subsystem.

This release increments the Python language support in StreamBase in two ways:

See Creating StreamBase Python Clients for details.

This release includes a change to the rolling file name feature used by the CSV File Writer and XML File Writer adapters, and by the sbprofile command. Before this change, the rolled-over file's name represented the beginning of the roll period for weekly rolling, but represented the end of the roll period for daily or hourly rolling. After this change, rolled-over file names represent the beginning of the roll period for all rollover intervals.

The CSV Socket Reader Input adapter was updated to provide a server mode where the adapter listens on a specified port for incoming connections. In addition, when acting as a client, the adapter now attempts to reconnect if connections fail, using a specified wait period between reconnection attempts. The adapter also has new properties for specifying how to handle incomplete records, and whether to log incomplete record events.

The CSV File Reader Input adapter was updated to provide new properties on its Parsing Options tab for specifying how to handle incomplete records, and whether to log incomplete record events.

The sample for the CSV File Reader adapter was simplified for clarity. See CSV File Reader Input Adapter Sample.

The tojson() function was updated to accept an optional second argument, boolean verbosity. By default, or with the second argument expressed as true, the function returns a string formatted as a JSON object, as defined on http://www.json.org. With the second argument expressed as false, the function returns a JSON array.

The new Loop sample in the Applications category of Studio's Load StreamBase Samples dialog demonstrates a looping arc in a simple EventFlow application.

The Saxo Bank FIX adapter allows a StreamBase application to connect to the trading infrastructure provided by Saxo Bank, and to exchange FIX messages with it.

The FIX Output adapter and all adapters that inherit from the FIX Output adapter were updated with the ability to track acknowledgment of the receipt of sent messages. This feature is described in Requesting Outgoing Message Send Confirmation on the FIX Output Adapter page.

The data dictionary for the Trading Technologies adapter was updated to implement version 7.8.1 of the Trading Technologies FIX specification.

The Thomson Reuters TREP-RT Subscribing Input adapter was updated such that CLOSED events, including access denied errors, are now logged at the INFO level, in addition to being sent to the adapter's event port.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_37. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The StreamBase documentation for release 7.2.10 included the following updates independent of the above changes:

When you drag the Java Operator icon from the Palette View to the canvas, the resulting dialog contains a new global Java operator, the Runtime Schema Cast operator. This operator transforms its input schema to its specified output schema, at the same time expanding capture fields in the input schema to the subfields of each capture field. Capture field expansion is configured in the operator to be performed with either Flatten or Nest transform strategies.

The new operator is used inside modules that use capture fields in their input schema or table definitions, when the module can be used generically to handle several similar incoming schemas, but also has special processing for one or more schema types, as described in the Runtime Schema Cast Operator's page in the Authoring Guide. The new operator has a sample, described in Runtime Schema Cast Operator Sample.

In the Properties view for a Query Table data construct, when pre-loading a Query Table with the Initial Contents tab, you can now use tab-separated value (TSV) files or TSV lines typed directly in the Initial Contents tab. TSV format joins the CSV and JSON formats already supported. See Properties: Initial Contents Tab on the Query Table Data Construct's page.

In the Simulation Streams section of the Feed Simulation Editor, you can now copy schemas from named schemas as well as from streams in modules or interfaces in your Studio workspace. The Copy Existing Stream button was relabeled Copy from Stream/Named Schema.

The EventFlow Debugger gained two new features in this release:

The expression language gained a new function, getAllHostIPs(), which returns as a list(string) all the non-loopback IP addresses assigned to the machine on which StreamBase Server is running.

The input and output adapters for IBM WebSphere MQ and Wall Street Systems were updated to use IBM MQ version 7.0. This update changed the JAR files you must obtain from IBM to use both adapter pairs. The list of required JAR files was updated in the documentation for both adapters and their samples.

The Alpha Trading Systems EMAPI Order Entry adapter was updated to support the EMAPI 5.1 specification.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_35. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The StreamBase documentation for release 7.2.8 included the following updates independent of the above changes:

When you select and right-click one or more tuples in the Application Input or Application Output views, the context menu now offers some new options designed to accelerate the creation of StreamBase JUnit tests.

The new Save-As options generate fragments that can be pasted into an existing StreamBase JUnit test's Java file. The options are described in Copy Tuples to Clipboard on the Application Output page.

In addition to the above, this release introduces a new feature that generates an entire working StreamBase JUnit Java test file from scratch, based on selected tuples in the Application Input and/or Application Output views. A new button now appears in the toolbar of the Application Output view:

and in the toolbar of the Application Input view:

Click either button to run the New StreamBase Unit Test Class wizard. This is the same wizard described in the Test/Debug Guide, with additions to specify the set of tuples on which to base the test, and to specify which tuple maker class to use.

Use the Selected Tuples option when you pre-select one or more tuple rows in the Application Input and/or Application Output views before starting the wizard. The Example code option generates a generic test as before.

The Tuple Maker Class drop-down selects CSV or JSON single quote tuple maker types. The JSON tuple maker is the default, but it does not handle tuples that contain slashes or double quotes in field values. Use the CSV tuple maker for those cases.

Starting with this release, you can select a Module Reference to a non-hygienic module, right-click, and from the context menu, select View Module with Overridden Schemas. Like using F3, this opens an EventFlow Editor canvas showing the referenced module. However, there are important differences compared to opening with F3:

While this feature is not an interactive debugging aid, it can nevertheless help you understand how a non-hygienic module is actually used at runtime. See EventFlow Editor Read-Only Canvas for more on this feature.

In the Project Properties dialog, the Module Search Path panel has a new feature. Below the Resolved Module Search Path grid, look for the new Generate configuration file snippet controls. These are links that create a fragment of a server configuration file ready for including into your project's top-level server configuration file, as described in Module Search Path Configuration File Snippets.

The same links appear in the Resource Search Path panel, in this case creating fragment files containing operator resource search path links. See Resource Search Path Configuration File Snippets.

StreamBase release 7.1.11 enhanced the X O keyboard shortcut so that if the most recently added component is a Module Reference, the automatically added output ports receive the names of the corresponding output streams in the referenced module.

This release adds the X I (ex eye) shortcut, which operates on input ports. For any component selected or recently added to the canvas, X I adds input streams to that component's input port or ports. If the newly added component is a Module Reference, the newly added input streams take on the names of the corresponding input streams in the referenced module.

Finally, this release adds the X A shortcut, which performs both X I and X O actions on the selected or most recently added component. This is especially useful right after adding an adapter or a Module Reference to the canvas.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_33. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

This release includes the new Web Reader adapter that reads from a specified URL using HTTP GET requests and emits the page contents in a string field of its Data output port. The adapter can be configured to read a web page on demand when receiving a tuple on its control input port, or to periodically poll the specified web page. The adapter is described in Web Reader Input Adapter in the Adapters Guide.

This adapter is especially useful for reading web pages that emit a pure XML structure containing ongoing market trade data. This adapter's sample demonstrates reading such a page, and uses the XML to Tuple operator to parse the received XML data. See Web Reader Input Adapter Sample in the Samples Guide.

The XML to Tuple operator was updated to eliminate unnecessary console messages when parsing XML comments and repeated elements that map to a list. The operator's sample was updated to include a new application that parses a complex, hierarchical XML input string, using a FIXML message as the example. See XML to Tuple and Tuple to XML Operator Sample.

The data dictionary for the Trading Technologies adapter was updated to implement version 7.8.0 of the Trading Technologies FIX specification.

The IBM WebSphere MQ Input adapter now has an Include JMS Correlation ID Field property, which is only activated when the Include Metadata Field is selected. If enabled, the adapter includes a JMSCorrelationID subfield in the schema metadata field that holds the JMS correlation ID received with the message.

The StreamBase documentation for release 7.2.7 included the following updates independent of the above changes:

Starting with this release, StreamBase Studio and StreamBase Server are supported using Oracle JDK 7, update 4 or later, as an alternative to the JDK 6 shipped with StreamBase. To use JDK 7, follow the instructions on Using an External JDK. That page also describes how to use JDK 7's G1 garbage collector.

To better support StreamBase projects saved in version control systems, this release provides a small behavior change when first opening EventFlow module files that do not have an associated layout file. In previous releases, under these circumstances, Studio would interpolate the layout information and save it back to the EventFlow file (to the .sbapp file). Now, Studio always creates a separate layout file (an .sblayout file) to contain the layout information.

Studio still honors the preference setting as before that determines whether or not to create a separate layout file when creating a new .sbapp file. The new behavior only affects the case when no matching .sblayout file exists when first opening an existing .sbapp file.

You can now select any two operators of the same type in the same EventFlow module, right-click and select Compare These Components from the context menu. This opens a side-by-side dialog that compares the properties of the selected operators. See Comparing Selected Components, a new page in the Authoring Guide.

If you place a trigger string in the text of a Note or in a Description field for an EventFlow component, the text of that note or field is placed in the Eclipse Tasks view as a task reminder note. Use this feature to place reminder notes about features of your application that need further development. The recognized trigger strings are TODO, FIXME, and XXX. See Using the Tasks View for more on this feature.

The icons for Query Tables and for Input and output stream on the EventFlow Editor canvas now show an overlay decoration to indicate two independent states. Streams that are used to update dynamic variables are shown with (x)= in the upper right of the icon. Streams and Query Tables that are implementations of an interface are shown with a boxed uppercase I in the lower right corner of the icon's tile.

The icons for input and output streams now show an overlay decoration on their upper back corner to indicate that this stream initiates a container connection to another stream by means of the Container connection field in the Advanced tab of the stream's Properties view.

Streams on the other side of the connection in the receiving container are not decorated. Decorations are not applied for container connections specified to connect at runtime in a deployment file, in a Studio launch configuration, in StreamBase Manager, or from the command line with the sbadmin command.

Studio has provided the X O keyboard shortcut for several releases: drag an operator to the canvas, then press X O, and the output port of the newly dropped component acquires an arc connecting to an output stream automatically named OutputStream. If the newly placed operator has more than one port, X O adds output streams to all ports. This keyboard shortcut performs the same function as right-clicking an output port and selecting Add Output Stream to Port.

This release adds special behavior to both the Add Output Stream context menu and the X O keyboard shortcut if the component dragged to the canvas is a module from the Package Explorer view. In this case, X O adds output streams as before, but it now names them the same as the corresponding streams in the referenced module.

When defining a Run, Debug, or Trace Configuration in Studio, and adding entries to the Environment tab, Studio has for many releases allowed the use of Eclipse variables to compose environment variables for a configuration. Starting with this release, the list of variables now includes ${sb_home}, which is defined as the absolute file system path of the StreamBase installation directory as currently set in Studio's Preferences dialog.

Studio was updated such that newly created Delta Stream icons are not subject to renaming, since Delta Streams must use their default names as created.

The Java and C++ Client Libraries were extended with new methods that allow you to specify a strategy (FLATTEN or NEST) for handling capture fields in streams accessed from a client application or from a custom embedded operator or adapter. The new APIs are described on a new page in the API Guide, Client Access to Capture Fields.

The Capture Fields for Generic Data Store sample now includes a Java operator that demonstrates the use of the new capture field strategy API methods.

The sbprofile command, in addition to the -c option, now has a -b outfile option. Use -b to specify that the generated output file is to be compressed with the bzip2 protocol, with the .bz2 extension added to the output file.

This release adds Vertica 5.1 to the versions supported by StreamBase. For Vertica 5.1 native connectivity, the StreamBase adapter requires the Vertica 4.1 Java client library (vertica_4.1.xx.jdk_5.jar). Vertica 5.1 is also supported for access through the Vertica JDBC driver.

The RSS Reader Input adapter was updated with three new properties: Use Proxy, Proxy Host, and Proxy Port. Use these properties to specify a proxy server through which the RSS feed of interest is visible.

The StreamBase documentation for release 7.2.6 included the following updates independent of the above changes:

This release extends and updates the Smart Merge feature for server configuration files, supplementing the support for standard XInclude with a new configuration file element <sb-include>, described in Using Modular Configuration Files in the Administration Guide.

This release also introduces a new utility, sbconfutil, that allows you to see as a single file the merged results of multiple inclusions. See the sbconfutil reference page, which is also available as a man page on UNIX platforms.

In previous releases, Query Tables were not subject to typechecking, because the schema of incoming data could not be anything other than the Query Table's schema. However, since release 7.2.0, Query Tables can have an initial load optionally specified. As of this release, individual Query Tables now participate in typechecking to make sure the table schema of associated Query operators are the same as the schema used for initial loading, if used.

In the Edit Dimension dialog that you open from the Dimensions tab of the Aggregate operator, the default size of the Open, Emit, and Close fields is now wider, and you can now use the F2 key to open a secondary Edit Expression pop-up dialog for each field. When you save a long expression in one or more of these fields, the dialog resizes when reopened to show as much of the expression in each field as possible.

In the Operation tab of the Query operator, the wording of the first entry in the If no row is found drop-down list was changed for clarity to Leave table unmodified, output nothing. The behavior of the operator is unchanged.

Releases 7.1.8 and 7.2.3 added the StreamBase → Clean File context menu option to clean and rebuild individual EventFlow modules. Starting with this release, you can use the Ctrl+Alt+T keyboard shortcut to initiate the same action.

The Open StreamBase Resource dialog, invoked with Ctrl+Shift+R, or from the Open StreamBase Resource toolbar button (), now includes any StreamBase LiveView™ configuration files with the .lvconf extension in the current Studio workspace.

Studio's Help menu now shows the Eclipse-standard Eclipse Marketplace entry, which gives Studio easier and more direct access to hundreds of third-party Eclipse plug-ins.

This release supports new Apache Flume External Input and Output adapters, which allow a StreamBase application to integrate with Apache Flume 1.1 or later as either a Flume sink or source. The Flume adapters are separately licensed, and, like other external adapters, this adapter pair has a separate installer downloadable from your usual StreamBase download site. See Apache Flume External Adapters for details.

The Feed Simulation Input adapter was updated to accept maxtuples and maxtime commands on its control input port, and to optionally report LASTNUMTUPLES actions on its status output port. See Feed Simulation Input Adapter for details.

A new property was added to the BM&FBovespa UMDF Input adapter and to the CME Market Data FIX/FAST adapter, which specifies a preference for the method to be used for recovering missing incoming FAST messages. The options are between either the TCP-based Historical FIX channel or the UDP-based Market FAST channel.

The Exegy Input adapter and Exegy Monitoring adapter were rewritten to take advantage of the Exegy Java API, resulting in the following changes:

The TableAccessor interface in the com.streambase.sb.operator package now has a size() method that returns the number of rows in the table.

For release 7.2.5 and later, StreamBase Studio is based on Eclipse 3.7.2. When installing Studio in the form of plug-ins for a separate copy of Eclipse, you must start with Eclipse 3.6 or 3.7. Eclipse 3.5 is no longer supported for that purpose. See Studio Eclipse Update Site.

For most purposes, StreamBase releases are well distinguished by their release numbers in three-position major.minor.maintenance format, such as 7.2.5. On rare occasions, you may need to report the fourth position build number. Starting with this release, the full four-position release number is visible from the command prompt in STREAMBASE_HOME/doc/VERSION.txt and in the documentation at the top of the Release Notes page.

The StreamBase documentation for release 7.2.5 included the following updates independent of the above changes:

In the Properties view for a Query Table, in the Table Settings tab, the radio button control to select In memory or On disk table types was replaced with a drop-down list. You can now write over the default selections to specify the name of a parameter in the form ${param-name}. This allows you to pass in the choice of table type at runtime. For example, if you set the Type control to the parameter ${tabletype}, then the module that contains the table must contain or be passed a parameter of that name with either disk or memory as its value.

StreamBase installation DVD-ROMs and installation download sites now include both 32-bit and 64-bit versions of the StreamBase .NET Redistributables kits. Unlike the 32-bit version, the 64-bit version includes only the required StreamBase assemblies and supporting DLLs, and does not attempt to install a .NET framework. Install either kit, but not both, to support running StreamBase .NET client programs on machines without StreamBase installed. See Redistributing the .NET Client Library.

When using the EventFlow Debugger, the red square Terminate button in the Debug view's toolbar () now performs the same clean shutdown of the running StreamBase Server instance as the Stop button () in the main Studio toolbar. In previous releases, the Debug view Terminate button stopped the debug session, then halted the Server instance without a clean shutdown.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_31. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

This release includes the Bloomberg Tradebook FX FIX adapter that allows a StreamBase application to connect to the trading infrastructure provided by Bloomberg Tradebook FX, and to exchange FIX messages with it. See Bloomberg Tradebook FX FIX Adapter.

The CME FIX/FAST adapter allows a StreamBase application to connect to CME's Market Data Platform to receive market data in the form of FIX/FAST messages. The adapter uses the high-speed, low-latency StreamBaseFIX engine to decode the incoming FAST messages. The adapter is described in CME Market Data Platform FIX/FAST Input Adapter and has a sample.

The EventFlow Debugger saw continuing improvements for 7.2.3, including the following features:

The default JVM memory settings were updated as of release 7.2.2 and are now documented in this release in Java VM Memory Settings and Garbage Collection Policy Settings in the Administration Guide. The following points summarize the changes:

The Query.sbapp sample in the Operators sample group was extended to add samples of Write-Update and Delete operations. See Query Operator Sample.

The implicit conversion of int to string for expression values in the Query operator, announced as deprecated in release 7.2.0, was removed as of this release for consistency with other StreamBase operators.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_30. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The sbd command has a new command-line option, --debug or -d. Use either option to prepare a run of StreamBase Server to accept remote debugging connections. Notice that the -d option was used in previous releases to enable intermediate stream dequeuing; that feature is now enabled with --intermediate-stream-dequeue.

The --debug option accepts one or more comma-separated key-value pairs as arguments that override default settings. For example, sbd --debug suspend=y causes the server to start and immediately suspend execution, awaiting connection from the debugger. See the sbd reference page for further details on overriding default settings.

StreamBase Studio has long supported attaching to a remotely running instance of StreamBase Server, as described in Attaching to a Running Application. Starting with this release, you can attach to a remote server in debug mode and use the EventFlow Debugger to debug both StreamBase and custom Java code. Follow the steps described in Attaching to a Running Server in Debug Mode.

The Thomson Reuters Enterprise Platform for Real-Time Subscribing Input Adapter was updated to support a new property, Disable Market Price View Requests. When enabled while using the Market Price message model, the OMM View Request optimization for subscribing to a selected set of fields is disabled.

This release adds the Feed Simulation Input adapter, which uses Feed Simulation configuration files to define the adapter's output ports and their schemas, and to determine what data is emitted from each adapter port. Feed simulation configuration files are created with the StreamBase Feed Simulation Editor as usual, and the same files can be used interchangeably with this Feed Simulation adapter and with the command line tool, sbfeedsim. Any data sources, generation methods, and emission rates definable with the Feed Simulation Editor can be used with this adapter. The adapter has a sample, described in Feed Simulation Input Adapter Sample.

In previous releases, the sbunit command was delivered as a 32-bit executable on both 32-bit and 64-bit Windows, which meant that StreamBase JUnit tests ran in a 32-bit JVM. Starting with release 7.2.1, the 64-bit Windows installation of StreamBase includes both 32-bit and 64-bit sbunit commands, so that 64-bit Studio now launches 64-bit StreamBase JUnit tests by default.

The PATH environment variable passed to StreamBase Server when launched from Studio now includes the streambase-install-dir/bin directory appropriate for the platform. This feature only affects the ability to locate StreamBase utility commands when run from within StreamBase Server by means of an External Process operator or the SB Admin operator. For Linux, a Studio-launched Server's PATH includes /opt/streambase/bin. For 32-bit Windows and for a 32-bit Server launched from 64-bit Windows, the Studio-launched Server PATH includes streambase-install-dir\bin. For 64-bit Windows, the PATH includes streambase-install-dir\bin64;streambase-install-dir\bin.

In addition, previous releases incorrectly appended the StreamBase home directory itself, streambase-install-dir, to the Studio-launched Server PATH. This was removed from the Studio-launched Server PATH as of this release, which may result in changed behavior if your application came to rely on this undocumented feature in some way.

This release adds the getClientIP() function.

In previous releases, the StreamBaseMonitor class always included the monitored details for each running operator in the monitored Server. Starting with 7.2.1, you must use the enableOperatorDetail() method to obtain operator details.

When a container, A, is added to a running StreamBase Server, and that container has a connection to the system.control stream, container A no longer receives the STARTING tuple on the control stream. However, other control-connected containers in the same Server instance do receive the STARTING tuple for container A. See The Container Subsystem.

The tojson() function was updated to accept an optional second argument, boolean verbosity. By default, or with the second argument expressed as true, the function returns a string formatted as a JSON object, as defined on http://www.json.org. With the second argument expressed as false, the function returns a JSON array.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_35. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The StreamBase documentation for release 7.1.12 included the following updates independent of the above changes:

To better support StreamBase projects saved in version control systems, this release provides a small behavior change when first opening EventFlow module files that do not have an associated layout file. In previous releases, under these circumstances, Studio would interpolate the layout information and save it back to the EventFlow file (to the .sbapp file). Now, Studio always creates a separate layout file (an .sblayout file) to contain the layout information.

Studio still honors the preference setting as before that determines whether or not to create a separate layout file when creating a new .sbapp file. The new behavior only affects the case when no matching .sblayout file exists when first opening an existing .sbapp file.

Studio has provided the X O keyboard shortcut for several releases: drag an operator to the canvas, then press X O, and the output port of the newly dropped component acquires an arc connecting to an output stream automatically named OutputStream. If the newly placed operator has more than one port, X O adds output streams to all ports. This keyboard shortcut performs the same function as right-clicking an output port and selecting Add Output Stream to Port.

This release adds special behavior to both the Add Output Stream context menu and the X O keyboard shortcut if the component dragged to the canvas is a module from the Package Explorer view. In this case, X O adds output streams as before, but it now names them the same as the corresponding streams in the referenced module.

The XML to Tuple operator was updated to eliminate unnecessary console messages when parsing XML comments and repeated elements that map to a list. The operator's sample was updated to include a new application that parses a complex, hierarchical XML input string, using a FIXML message as the example. See XML to Tuple and Tuple to XML Operator Sample.

The RSS Reader Input adapter was updated with three new properties: Use Proxy, Proxy Host, and Proxy Port. Use these properties to specify a proxy server through which the RSS feed of interest is visible.

The EBS adapter was updated with a new data dictionary that supports both EBS Ai FIX 1.3 and 1.4.

The StreamBase documentation for release 7.1.11 included the following updates independent of the above changes:

This release introduces two new Palette operators, the XML to Tuple Converter and the Tuple to XML Converter. These operators are found by dragging the Java Operators icon from the Palette to the canvas, selecting one of the new operators from the list, then pressing Enter.

As their names imply, the new operators are inverses of each other. In fact, the best way to learn the XML syntax expected by the XML to Tuple Converter is to run the same schema through the Tuple to XML Converter.

The XML to Tuple Converter takes a single string field as input, with the string expected to contain a well-formed, XML-encoded message. The operator emits one tuple for each input XML message. The operator has a standard Edit Schema tab in its Properties view in which you specify the schema of the tuple you want to extract from the input XML string. The tuple's schema does not need to describe the entire XML string, which lets you extract only the elements of interest, as long as the schema preserves the hierarchy of the incoming XML. The operator provides a method to extract attributes of incoming XML elements.

The Tuple to XML operator converts StreamBase tuples to XML-encoded messages. Each field in the operator's input port schema populates an element or attribute in the resulting XML message. Each tuple enqueued to the operator results in a single XML message emitted on its output port.

The new operators share a new sample that provides examples that convert both element-only XML and XML that has elements with attributes. The sample EventFlow modules use named schemas to define elements that may include attributes. See XML to Tuple and Tuple to XML Operator Sample.

The behavior of the Union operator was changed slightly as follows: when dragging an arc to a Union operator, the standard create-new-port behavior is only triggered when the Union operator has no available input ports.

StreamBase Studio now places entries in the Problems view when it detects semantic errors in the server configuration file that affect the current project's typechecking environment. For example, referring a custom function to a Java operator that does not exist now shows a Problems view entry with the name of the function that could not be configured. Resolve such errors early in the development process.

The Log Asynchronously property for new Log adapters is now enabled by default.

The IBM WebSphere MQ subscribe (input) and publish (output) adapters were updated to provide support for topics as well as queues. For the input adapter, the following new properties control the use of topics: Use Topics, Default Topic, Allow Dynamic Topics, and Max Topic Count. See IBM WebSphere MQ Input Adapter.

The output adapter has the following new properties: Use Topics, Default Topic, and Topic Field Name. See IBM WebSphere MQ Output Adapter.

The Thomson Reuters TREP-RT Subscribing Input adapter was updated to add a new property, Derive Update Type from Market Feed Data, which is active when the current message model is Market Feed (RMDS 5). When enabled, this property causes the adapter to derive the specific update message type from the raw market feed data. This might be used when the useMarketfeedUpdateType property in the RFA configuration file does not function as expected in certain Thomson Reuters environments. See Thomson Reuters Enterprise Platform for Real-Time Subscribing Input Adapter.

This release extends and updates the Smart Merge feature for server configuration files, supplementing the support for standard XInclude with a new configuration file element <sb-include>, described in Using Modular Configuration Files in the Administration Guide.

This release also introduces a new utility, sbconfutil, which allows you to see as a single file the merged results of multiple inclusions. See the sbconfutil reference page, which is also available as a man page on UNIX platforms.

StreamBase installation DVD-ROMs and installation download sites now include both 32-bit and 64-bit versions of the StreamBase .NET Redistributables kits. Unlike the 32-bit version, the 64-bit version includes only the required StreamBase assemblies and supporting DLLs, and does not attempt to install a .NET framework. Install either kit, but not both, to support running StreamBase .NET client programs on machines without StreamBase installed. See Redistributing the .NET Client Library.

In the Edit Dimension dialog that you open from the Dimensions tab of the Aggregate operator, the default size of the Open, Emit, and Close fields is now wider, and you can now use the F2 key to open a secondary Edit Expression pop-up dialog for each field. When you save a long expression in one or more of these fields, the dialog resizes when reopened to show as much of the expression in each field as possible.

This release includes the Bloomberg Tradebook FX FIX adapter that allows a StreamBase application to connect to the trading infrastructure provided by Bloomberg Tradebook FX, and to exchange FIX messages with it. See Bloomberg Tradebook FX FIX Adapter.

The CME Market Data FIX/FAST adapter allows a StreamBase application to connect to CME's Market Data Platform to receive market data in the form of FIX/FAST messages. The adapter uses the high-speed, low-latency StreamBaseFIX engine to decode the incoming FAST messages. The adapter is described in CME Market Data Platform FIX/FAST Input Adapter and has a sample.

This release supports new Apache Flume External Input and Output adapters, which allow a StreamBase application to integrate with Apache Flume 1.1 or later as either a Flume sink or source. The Flume adapters are separately licensed and, like other external adapters, this adapter pair has a separate installer downloadable from your usual download site. See Apache Flume External Adapters for details.

A new property was added to the BM&FBovespa UMDF Input adapter and to the CME Market Data FIX/FAST adapter, which specifies a preference for the method to be used for recovering missing incoming FAST messages. The options are between either the TCP-based Historical FIX channel or the UDP-based Market FAST channel.

The Exegy Input adapter and Exegy Monitoring adapter were rewritten to take advantage of the Exegy Java API, resulting in the following changes:

Release 7.1.8 added the StreamBase → Clean File context menu option to clean and rebuild individual EventFlow modules. Starting with this release, you can use the Ctrl+Alt+T keyboard shortcut to initiate the same action.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_31. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

The C++ Client Library was augmented with two methods added to the sb::ClientSettings class:

The associated setting controls the maximum number of tuples stored in the client's internal dequeuer queue before applying back pressure on the connected StreamBase Server. See the Doxygen documentation for details.

For most purposes, StreamBase releases are well distinguished by their release numbers in three-position major.minor.maintenance format, such as 7.1.9. On rare occasions, you may need to report the fourth position build number. Starting with this release, the full four-position release number is visible from the command prompt in STREAMBASE_HOME/doc/VERSION.txt and in the documentation at the top of the Release Notes page.

The StreamBase documentation for release 7.1.9 included the following updates independent of the above changes:

Release 7.0 added the ability to use the XInclude standard to break server configuration files into reusable pieces. This release extends that support with a new smart merge feature that allows you to XInclude an entire server configuration file into another, including the root element, with any duplicated elements merged together into one element in the resolved file. Child elements that do not conflict are simply merged, while child elements with conflicting definitions are resolved, usually with the last definition winning. See the revamped Using Modular Configuration Files page for details.

This release includes a preview of the module signature feature, which is expected to be added to a future release. To enable, use Window → Preferences, open the StreamBase Studio → Typechecking panel, and select the check box labeled Enable module signature change analysis during builds. This directs Studio to store a hash of the signature for all modules and all intermediate stream schemas in the metadata of each module file. This in turn allows Studio to build dependent modules only when the hash value changes, which means only when the module itself has changed. For large, complex applications with many modules, this setting speeds up building of the application as a whole.

In the Properties view for a Query Table, in the Table Settings tab, the radio button control to select In memory or On disk table types was replaced with a drop-down list whose third entry is a blank line. Select the blank line to enter the name of a parameter in the form ${param-name}, which allows you to specify the choice of table type at runtime. For example, if you set the Type control to the parameter ${tabletype}, then the module that contains the table must contain or be passed a parameter of that name with either disk or memory as its value.

This release introduces several new features in StreamBase Studio that help you detect and eliminate any schemas or constants imported into a module that are not currently being used by that module. Eliminating unused imports speeds up the build times for large StreamBase applications. These features are discussed in a new page in the Authoring Guide, Minimizing Module Imports.

In large StreamBase applications with many module references, building the entire application can take a noticeable amount of time. Starting with this release, you can specify building of one module at a time instead of building the entire application. To do this, select an individual EventFlow module's name in the Package Explorer view, right-click, and select StreamBase → Clean File. This option does not appear for StreamSQL modules.

In previous releases, if you changed anything in a project's server configuration file, Studio would re-typecheck the build environment for the entire project. As of this release, Studio detects the type of changes made to server configuration files, and only rebuilds when necessary. For example, a change in XML comments no longer rebuilds the project.

A new preference setting, Ignore label widths during auto-layout was added to the EventFlow Rendering page of StreamBase Studio preferences. This setting is disabled by default; enable it to direct Studio to ignore the width of component labels when running Layout Application (Ctrl+L), which allows long labels to overlap on the canvas. For the most compact canvas display for large modules, enable this option and disable the slanted arcs preference.

The sb-config command was updated to provide better support on UNIX for directories named with spaces. The Windows version of the command was not changed, and on UNIX, the Java-related command options were not changed. See the sb-config reference page for details.

The Exegy input adapter now has an Include Receipt Nanotime property, which, when enabled, adds an XC_ADAPTER_INITIAL_NANOTIME field to the Equities, Commodities, and PriceBook output ports. See Exegy Input Adapter.

The Thomson Reuters Enterprise Platform for Real-Time Subscribing Input Adapter was updated to support a new property, Disable Market Price View Requests. When enabled while using the Market Price message model, the OMM View Request optimization for subscribing to a selected set of fields is disabled.

The Oracle JDK installed for private use by StreamBase was updated to version 1.6.0_30. As before, you can override this installed default to use a different supported JDK as described in Using an External JDK. The supported JDK versions are listed in Supported Configurations.

This release includes two adapters contributed by Thomson Reuters for connecting to the Thomson Reuters Enterprise Platform for Velocity Analytics. These are the Velocity Analytics Broadcast Server Input Adapter and the Velocity Analytics UTSS Output Adapter. Both adapters are illustrated in a new set of samples, described in Thomson Reuters Velocity Analytics Samples.

Release 7.1.6.2 included a new adapter for the Exegy Ticker Plant data feed. The adapter is described in Exegy Input Adapter in the Adapters Guide, and has a sample, described in Exegy Adapter Sample in the Samples Guide.

In addition, there is a new Exegy Monitoring adapter described in Exegy Monitoring Adapter. The Exegy sample includes modules that illustrate both the Exegy Input and Exegy Monitoring adapters.

Certain typecheck errors in a server configuration file can result in the entire file considered invalid and discarded for a particular run of StreamBase Server. The handling of these errors was changed as of this release.

In previous releases, an error such as a failure to locate the files that support a custom function class would result in logging a warning to the Error Log or Problems view. To ensure that errors of this type are not missed (since they are often the root cause of other application typecheck errors), Studio now shows these errors as an error against the server configuration file itself.

The PATH environment variable passed to StreamBase Server when launched from Studio now includes the streambase-install-dir/bin directory appropriate for the platform. This feature only affects the ability to locate StreamBase utility commands when run from within StreamBase Server by means of an External Process operator or the SB Admin operator. For Linux, a Studio-launched Server's PATH includes /opt/streambase/bin. For 32-bit Windows and for a 32-bit Server launched from 64-bit Windows, the Studio-launched Server PATH includes streambase-install-dir\bin. For 64-bit Windows, the PATH includes streambase-install-dir\bin64;streambase-install-dir\bin.

As of this release, there is a new preference setting in the Window → Preferences → StreamBase Studio → Authoring → EventFlow Rendering panel, Use slanted arcs during arc routing and auto-layout.

When enabled (the default), arcs placed by the Layout Application (Ctrl+L) and Reroute Arcs (Ctrl+Alt+L) functions are shown with an angled slant between components. When disabled, arcs are only placed horizontally or vertically, with right angles for every direction change. This feature restores to release 7.1.7 and later the layout cosmetics of StreamBase releases 6.6 and earlier. In addition, an automatically laid out module is now slightly denser than in earlier 7.x releases.

This release adds the getParallelRoot() function.

In the Pattern operator, the Window Size field in the Pattern Settings tab now accepts a module parameter in the form ${ParamName}.

The FIX Schema Designer (FSD), added in release 7.1.3, now supports the FIX/FAST protocol. FAST stands for FIX Adapted for STreaming, and is a FIX variation used by some trading venues. To design a schema to communicate with such a venue, select FAST Data Dictionary File in the FSD's Select Data Dictionary dialog, then navigate to select the dictionary file provided by the venue. In all other ways, the FSD works the same as before.

The input adapters for IBM WebSphere MQ and Wall Street Systems were updated with two new properties: Include Depth Field and Syncpoint on Get. These properties are enabled by default for compatibility with previous releases of these adapters. Disabling both settings significantly improves performance of these adapters. See IBM WebSphere MQ Input Adapter and Wall Street Systems Input Adapter.

The Thomson Reuters Enterprise Platform for Real-Time (TREP-RT) Subscribing Input adapter (formerly known as the Thomson Reuters RMDS Subscribing Input adapter) now supports thin tuples for the MarketPrice message model. Previously, thin tuples were supported for the MarketFeed message model only, although the adapter incorrectly allowed thin tuples to be configured for other message models.

In addition, the adapter gained a new property, Include Subscribe Pass-Thru Field property on the Tuple Properties tab of its Properties view. When selected, this property requires a field named PassThru to be added to the schema of the DynamicSubscribe input port. There are no constraints on this field's data type. The adapter copies the value from this field to a corresponding field of each each market data tuple emitted as a result of the subscription.

The TIBCO Rendezvous Subscribing Input Adapter has a new property, Ignore Advisory Messages. When enabled, the adapter discards messages with subjects that begin with _RV.

The E-mail Sender Output Adapter has two new properties, Enable Event Port and Event on Send Success. When the first is enabled, the adapter gains an event output port that emits event tuples when errors occur sending e-mail. When the second property is enabled, the event port also emits a tuple when e-mail is successfully sent.

This release adds support for Vertica 5.0, both for JDBC access and for high-performance bulk loading with native connectivity, and drops support for Vertica 3.5.

The JAR files supporting the StreamBaseFIX FIX engine were updated to version 2.10.7. The JAR files supporting FIX/FAST were updated to version 2.11.8.

This release includes a new adapter for the Exegy Ticker Plant data feed. The adapter is described in Exegy Input Adapter in the Adapters Guide, and has a sample, described in Exegy Adapter Sample in the Samples Guide.

The three Thomson Reuters RMDS adapters were renamed to reflect the new name of the protocol formerly known as RMDS: Thomson Reuters Enterprise Platform for Real-Time, or TREP-RT. The adapters are otherwise unchanged. See:

The Results grid in the Thomson Reuters Schema Designer view now has an In Summary column, which for each FID row is either blank or contains Yes. Use this column to help select fields when generating a schema for a summary-only tuple. See Thomson Reuters Schema Designer View.

The StreamBase utilities jsbadmin, jsbc, sbbundle, and sbfeedsim were updated to accept the -J option anywhere on the command line instead of its former required first position. You can now specify -J options with or without a following space.

The expression language's eval() function gained an optional fourth parameter to specify the evaluation cache size. See eval().

The .NET Client Library was rewritten to improve the performance of creating tuples and accessing tuple field values, with initial tests showing a speedup of 10 to 40 times from the previous API. Most of the .NET API for building StreamBase clients is now a pure .NET implementation, although some classes (such as StreamBaseClient) remain based on the C++ API. (The .NET API for building StreamBase Server monitor applications is unchanged.)

Rewriting most of the API as pure .NET forced some changes in the interface, as described in Migrating .NET Clients in the API Guide. You may need to make source code changes for existing .NET client applications and recompile them in order to accommodate the new API.

The Query operator associated with a JDBC data construct has a new property, Timeout, on the Query Settings tab of the Properties view. Use this field to specify an integer time in milliseconds after which this particular query times out. You can still use the jdbc-query-timeout option in the <data-source> element of the server configuration file to specify a default timeout for all queries to that data source.

This release adds the Java system property streambase.jdbc.max-repeat-exceptions. The default value of 1 tells StreamBase Server that a single instance of a JDBC exception is regarded as an exception. Specify zero to mean that JDBC exceptions can be repeated without limit. Use an integer, n, greater than 1 to specify that JDBC exceptions might occur n times before succeeding.

The Timestamp class in the C++ Client Library now uses snap to end of month semantics for the setMonth() method. The following code fragment illustrates how this works:

Timestamp tm;
tm.setYear(2011);
tm.setMonth(10);
tm.setDayOfMonth(31);
// tm now represents October 31, 2011
tm.setMonth(2);
// tm now represents February 28, 2011

The sbd command has a new option, --override-version-check. By default, when you ask StreamBase Server to run a bundle file, it checks the StreamBase version that created the bundle and refuses to run a bundle not created with the same major.minor release number. You can use the new option to direct the server to override this check and try to run the bundle file anyway. Incompatibilities between the bundled application and the server may still prevent the bundle from running.

The --if-modified option of the sbargen command, when determining whether an application needs to be re-archived, now ignores changes in files whose names match the following default regular expression patterns:  .*\.log,  .*\.bak,  .*~, and  #.*#

You can customize the list of exclusion patterns with the new --exclude-pattern option, which can be specified multiple times.

The expression language gained the join() function, the opposite of split(). Use join() to return a string formed by concatenating the members of a list of any type, and interposing a specified character or string between each element.

The Excel External adapter was updated such that the buffer-size configuration parameter and TUPLENUM tag are now limited to values between 1 and 1000. Prior to this change, there was no limit on buffer-size, and TUPLENUM was limited to values between 1 and 999.

The embedded JMS Input and Output adapters were updated such that the following events are now sent to the Status output port:

In addition, both adapters have a new property Log Full Call Stacks. When selected, exception call stacks are shown independent of the adapter's Log Level setting.

The 29West LBM Subscribing and Publishing adapters were updated to support pre-defined messages. Enable support for this feature with the new Enable Pre-defined Messages property. The adapters have several other new properties that control how a publish and subscribe adapter pair agree on message definitions in cooperation with the StreamBase schema.

The StreamBase documentation for release 7.1.5 included the following changes independent of the above updates: