Cradle-5.2 Features

The Cradle-5.2 release provides major new features, bug fixes, and some general improvements. The most significant changes are listed below, grouped by Cradle module. For a full list of all enhancements and bug fixes, consult:

The Cradle-5.2 Release Notes

The What’s New document in the Help menus of the Cradle clients

UNIX Installation Changes

1. The set_access script has new options to control user and group owners, and the access modes of installation files
2. Installation scripts correctly recognise users and groups on all platforms
3. server option to enable and disable CDS and CWS setuid-on-execute rights
4. A new UNIX system start-up script c_start that can be run to start, stop, restart and test whether the CDS and/or CWS are running. It can be used as a start or stop (S or K) script in a UNIX system’s automatic startup and shutdown directories /etc/rc*.d and so on.

Windows Installation Changes

1. InstallShield has been changed so it is possible to install in Standalone mode in which Cradle clients and servers only communicate through the TCP/IP software loopback address 127.0.0.1

Communications Changes

1. The Menu and ping_cds and test_login utilities support –standalone and –cds command-line arguments and the CRADLE_CDS_HOST environment variable to specify the hostname or IP address of a machine running the CDS. This allows them to link to a CDS outside the subnet where they are being run.
2. The ping_cds and test_login diagnostic utilities now have confirmations on Windows so their last output can be read before their windows disappear
3. Fixed problem connecting to CDS when the CDS could use a wrong IP address for a client, preventing the client from connecting. Examples:

When a client connects over VPN, the CDS used the client’s IP address, not its address in the VPN network

When a laptop has multiple interfaces, the CDS might use the wireless IP address even if this interface was inactive on the PC

A new message protocol has been introduced in Cradle-5.2 that:

Produces a 10-25% reduction in the data volumes sent between Cradle clients and servers, with the associated performance improvement

Makes Cradle considerably more resistant (or immune) to the side-effects of port scanners and Denial of Service (DOS) attacks by malicious software

Ensures that Cradle clients of one software release cannot accidentally be used with Cradle servers of another release

Allows detailed communications statistics to be obtained from Toolset and WorkBench

Cradle-PDM (Product Data Management) Changes

1. Project Unique IDs (PUIDs) are now supported in project databases and are discussed in a separate section of this newsletter.
2. In Hierarchy Diagrams (HIDs), you can set the To type as a number of item types from a selection list. Previously you could only select All or None.
3. The arrows in Hierarchy Diagrams could occasionally be drawn the wrong way
4. The print quality of large Hierarchy Diagrams (HIDs) has been improved in Word and SVG output:
   • Text is more readable
   • Line thicknesses have been reduced
   • Misalignment of separators and shading in very large HIDs has been fixed
5. A $ACROBAT command directive is now provided which automatically runs whichever version of Acrobat Reader (first) or Acrobat itself (second), version 4 or later, that it finds. The default PDF 3 frame type has also changed so that it is just a PDF frame type and now uses the $ACROBAT command directive.
6. A bug whereby the LATEST ownership and Middle, Bottom or Non-Bottom link selection in a query could cause Cradle to hang, has been fixed
7. The iconv utility is distributed in all UNIX and Windows filesets to convert the character encodings for special characters (such as less than or equal) and Windows code pages into UTF-8 to be brought into Cradle. It can be used to preprocess data files from tools that do not use standard character encodings (such as CORE or RDD-100 or DOORS) and is automatically run by the Cradle converters in some cases.
8. Table titles did not support international languages, and now do
9. Default TEXT and NOTES frames are now loaded when project-specific items are imported that are not part of the project schema at the time of the import
10. A <null> value for empty attributes and !<null> for non-empty attributes is now supported in queries. Previously, it was difficult (but possible) to select items by empty category or frame values, but not possible to select by empty attributes for Identity, Name, Key, Group, Comment or Description.
11. Queries returning results pre-sorted by Key or Group failed in searches of system notes or Implementation domain data definitions. This has been fixed.
12. The CSV merge options have been extended to include the merging of Key, Group, Comment and Description fields for DD entries, events, requirements, system notes and specifications
13. Some issues with CSV imports and exports have been fixed, such as problems with empty slots in the CSV file field mappings, and an inability to import or export specifications’ symbol categories (such as function, process or class)
14. An internal error during some CSV imports on Linux has been fixed
15. The import code detects diagrams with symbols whose numbers have not been supplied, and now automatically sets numbers into such symbols
16. It is now possible to setup automatic numbering for requirements and system notes within Project Setup
17. Query phase nodes have been extended to display the items that they find as a further option for such nodes, in addition to running the query and displaying it, or opening the items that it finds
18. New phase node types Document, Command and Folder have been added to the Phases section of Project Setup. These allows a command to be run or a document to be launched, and allow explicit grouping of phase nodes.
19. The Project Setup tools in Toolset and WorkBench now cannot be used if any items are being edited. This is to avoid potential problems that could occur if frame type definitions were changed whilst these items are open.
20. An inability to launch external applications on frames stored As Reference or As File with pathnames or URLs of over 256 characters has been fixed
21. A directory chooser has been added to the Create Project dialogue in the Cradle Menu on both Windows and UNIX
22. When creating a project on Windows or UNIX, you can choose which schema to use, to determine the prj_params file and the initial set of project definitions
23. A facility has been added to the Menu whereby a user can view and/or set and/or remove the current value of the CRADLE_CDS_HOST environment variable or to create it if it doesn’t exist

SpellChecker

1. The spell checker now checks all item attributes

CORE-to-Cradle Converter

• The CORE-to-Cradle converter has been updated to support changes made by Vitech Corporation in the format of .rdt files produced by versions 4 and 5 of its CORE tool. This means that you can migrate requirements, models and other data from CORE into Cradle.

Teamwork-to-Cradle Converter

1. The Teamwork-to-Cradle converter has been updated to support the parsing of Software Architecture Diagrams (SADs)
2. The converter has also been amended to refine the algorithms by which excess routepoints are removed and by which disconnected flows are detected inside source or destination symbols and reconnected to these symbols

Cradle Toolset Changes

1. The Delete key is now supported to delete selected items
2. Items can now be exported based on a query or a set of queries
3. CASE and Viewer UI styles have been removed
4. Removed the need for team leaders to have APPROVAL privilege to register items in the Configuration Management System (CMS). This stops the need for all team leaders to review one team’s work.
5. A Change/Reset Counts… option has been added to the Admin pulldown menu in the Toolset’s main screen. This allows you to set the value of the auto numbering counts file for each item that supports auto numbering.
6. An option has been added to the Admin -> Utilities menu, Change/Reset PDUID count…. This option allows you to change the unique ID of the next PDUID to be assigned.

Cradle-API Changes

1. The Cradle-API on Linux has been rebuilt with GNU gcc, version 3.3.2 so it is easier to link it with current versions of scripting and other tools
2. A new routine CAPIitem_createxref has been added to the Cradle API so cross references can be created between items in the database
3. The API now supports PUID attributes and auto-numbering
4. The API now supports querying by symbol category for specifications and data definitions

Cradle-REQ (Requirements Management) Changes

1. The Key, Group, Comment and Description attributes are now supported in source document capture operations
2. When capturing information from a source document using the Source Document Manager, a paragraph acts as a buffer for the source document text from which capture operations set attributes in the item being captured. If a capture operation captures all of a line, then that line is now deleted from the buffer. This now means that successive lines from the buffer can be processed, greatly increasing the flexibility of the available captures that can be performed.

Cradle-MET (Metrics) Changes

1. The Metric tool’s UI has been changed so the contents of a metric are shown in a scrollable list of metric elements. Added the means to copy, delete and edit metric elements, to define new elements and to move elements up and down in the ordered list of metric elements. The number of metric elements supported has increased from 32 to 100.
2. The range of calculations that can be performed in metric elements has been extended to include Minimum, Maximum, Average and Total, with optional grouping and sub-grouping by one or two attributes.
3. The Count option has been extended to allow optional grouping by another attribute and further optional sub-grouping by a second attribute.

Cradle-SYS (System Modelling) Changes

1. When copying and pasting in a selected timeline, you can specify the scope of the operation to include definitions and hierarchies.
2. If the move-and-reconnect operation was used on diagrams with gaps in the symbol numbering, then the numbers of the reconnected symbols could be changed. This has been fixed to guarantee that reconnected symbols keep their numbers.
3. Problems in the renumbering and reordering of diagram and specification hierarchies have been fixed. In particular, all hierarchical operations now work correctly irrespective of whether equivalent diagrams and specifications exist or if the specifications are linked in hierarchies by cross references, or both.
4. The Key attribute was not preserved when specifications were copied, either directly or as part of a larger operation, such as reordering functions
5. When adding new branches to an existing eFFBD function (PFD operation) that are to end in a particular node, the topmost existing branch is searched (skipping through symbols with one input timeline and one output timeline) to find an existing node of the required type and if one is found it is used else a new node is created. This behaviour is particularly required for J nodes.
6. Inserting new branches from an eFFBD function or PFD operation could incorrectly reuse an existing J (join) node instead of creating a new J node
7. The carry down of timelines in the Expand operation has been improved. If the input timeline has routepoints, the last routepoint is used for the position of the timeline start symbol. If the symbol has a single output timeline with routepoints then the first routepoint is used for the position of the timeline end symbol.
8. Project-specific data definition (DD entry) frames were not shown in the Data Dictionary Editor (DDE) picture bar. They are now.
9. Toolset now recognises JPEG images and ignores unsupported image types
10. Cutting and pasting eFFBD functions will now carry the I/O
11. eFFBD functions, PFD operations and control nodes can be reconnected into other timelines by moving them. As symbols are moved over timelines where they could be inserted, the timeline highlights. Releasing the mouse button disconnects the functions/operations/nodes from their current timeline and reconnects them into the new timeline, retaining links to all of their I/O.
12. Fixed error when attempting to move eFFBD or PFD timeline loop symbols
13. Adding new exit branches from eFFBD function/PFD operations or control nodes could cause these to overlay existing branches, this has been fixed
14. Selected symbols’ sequence numbers are now shown in the DGE status bar
15. Pressing the ESC (Escape) key will cancel any DGE operation
16. The cursor is now positioned at the bottom-right corner of rectangles marking move, paste or resize operations, instead of the bottom-left corner
17. Saving a diagram containing symbols that should be numbered but which had no numbers (due to the above import problem) could damage the names of definitions if the Name Synchronisation feature was enabled. This is fixed.
18. Software Architecture Diagrams (SADs) diagrams have been implemented in the Implementation Domain
19. New DGE Symbol Properties dialogue to control per-symbol characteristics
20. Added navigation from a PFD system operation to its referenced specification
21. The user can now specify scope options when inserting into a timeline. A new Insert Symbols dialogue is produced when pasting into a selected timeline. This dialogue has the same options as the Paste Symbols dialogue and allows the user to control the scope of the symbol insert which includes With Defs and Hierarchy and inheriting cross references.
22. The names of data links in Architecture Interconnect Diagrams (AIDs) can now be optionally hidden on a per-symbol basis by the new Symbol Properties dialogue in the Toolset’s Diagram Editor (DGE)
23. Consistency checks of eFFBDs did not correctly allow for inputs to, and outputs from, environment symbols
24. Consistency check reports use symbol sequence numbers as well as existing symbol details. Used with the Diagram Editor’s facility to display such symbol numbers, this prevents ambiguity as to which symbol has a particular error.
25. Consistency checks of eFFBDs and PFDs could report timelines as being un-necessarily named where names are needed. For example, an input timeline must be named if it is one of many output timelines from another function.

Cradle-DOC (Document Publisher) Changes

1. Several significant improvements have been added, notably the ability to control the sort order of reported items, and cross referenced items
2. -login and -file command line parameters have been added to the Document Publisher to allow it to be run directly from a project phase hierarchy

Cradle-WRK (WorkBench) Changes

1. WorkBench (and Cradle Menu) will now detect the user’s current UI style on Windows, and will adjust their UIs to match
2. The Find operation now finds the user’s text in attributes of linked items shown in a view, for example within the cells of nested table rows
3. Diagrams were not properly centred when initially displayed
4. Navigating through model elements in any WorkBench tree could cause errors in the CDS error log, even though the operations were completed successfully. This could produce very large log files in large Cradle systems.
5. Extended the matrix generation facility to support matrices with specifications (such as functions, architecture components, activities) as rows and columns, and details of their interconnecting data I/Os in the matrix cells
6. Fixed problem replying to alerts where the Subject line could be incorrect
7. Fixed problem where deleting an alert did not select the next alert
8. Multiple Ctrl-Z (undo) and Ctrl-Y (redo) are now supported in text areas
9. Items can be deleted by selecting them and pressing the Delete key
10. WorkBench has been extended to fully support PUIDs and auto numbering, in views, forms, queries and in Project Setup
11. New Utilities item in the Admin menu for PUIDs and auto numbering
12. Reports have been extended to support metrics. Metrics are generated using the Cradle metrics subsystem
13. A default view, style and navigation can now be specified for a query
14. The full query name is now displayed in a selected tab in the query pane label and in the item type and the full identity of items is shown in the item pane label
15. When a new item/child/sibling is created, no queries are updated, the new item is added to the end of the active query that it matches. This means that creating the new item is quicker as all queries are not being updated.
16. Ctrl+C can now be used to copy text in read-only items
17. The lists of definition files are now cached so they are only retrieved from the CDS when needed, improving responsiveness for the user

Cradle-WEBA (Web Access) Changes

1. Once the Login button has been pressed, it is then disabled to prevent users pressing it again during the login sequence and causing multiple logins
2. Web Access now supports HTTPS
3. Unsupported browsers cannot login because when they do Web Access licences are locked
4. Full support for PUIDs and auto-numbering has been added