04-2012

Owned by Laura White (Unlicensed)
Sept 10, 2015

New Features

  • Workflow Improvements
    • Pre-translation has been added to the Lingotek workflow engine and completely revamped to provide an array of translation memory leveraging options
    • Pre-fill TM is triggered by the "pre-fill translation memory" action, which can be added to any phase within a workflow; its settings include:
      • TM vault sequencing - determines tie breakers in the case where two vaults produce an equal match
      • Match minimums - determine the percentage threshold required for pre-filling a segment with TM
      • Segment locking - determines whether or not to lock exact matches from editing within the workbench
    • The "propagate locked segments" action can be added to phases following a phase with a "pre-fill translation memory" action
      • Phases to which this action is added will keep locked any segments that were pre-filled and locked with an exact match in a previous phase
      • The propagation of segment locks must be continuous (i.e. you cannot skip phases)
    • These settings can be individually set for each TM vault and glossary being searched for matches to pre-fill
    • All of these configurations can be defined as part of your workflow templates to further reduce project setup time, while ensuring consistent leveraging of your TM assets in accordance with your TM strategy
    • Machine translation can now also be configured as part of your workflow templates, giving you full control over when MT is ran and in which phase(s) it gets ran
      • Machine translation is triggered by the "machine translate" action, which can be added to any phase within a workflow
      • Note that including the "machine translate" action in a mid-workflow phase will only fill in those segments that have not been edited in that phase; thus overwriting un-edited translations that were propagated forward from a previous page
    • Added a "send an HTTP notification" action to the rules engine that allows for defining a URL and BODY for either a GET or POST notification, which can then be triggered by workflow
  • Dashboard Improvements
    • Translation memory analysis can now be ran across all documents within a project, decreasing translation project costs by identifying intra-project repetitions and matches
    • Historical analysis results (including the settings used to run the analysis report) are now stored for future reference and quick retrieval
      • Analyze no longer implicitly pre-fills exact matches from TM (pre-fill TM is now exclusively processed via workflow)
      • Analyze and Pre-fill TM results are stored under the name specified in the settings
      • Analyze no longer implicitly searches the "save to" vault
      • Project managers can now specify which glossaries and TM vaults to search (and in which order) when running Analyze
    • Analyze can now be ran for a sub-set of documents within a project
    • Translators and reviewers can now find and replace text across all documents assigned to them within a project
      • The specified term can be searched for in either source or target (replace is only supported for target text)
      • You can specify whether or not the search should be case-sensitive
      • All instances of the phrase can be replaced, or you can specify which segment to replace the phrase in
    • The "Paste MT" action has been renamed to "Machine Translate"
      • Running this action now allows for specify in which phase the machine translated text should be added
      • Note that running Machine Translate in a mid-workflow phase will only fill in those segments that have not been edited in that phase; thus overwriting un-edited translations that were propagated forward from a previous page
    • Support for the SRX (Segmentation Rules eXchange) standard has been added for Microsoft Office (.doc, .xls, .ppt) and Java properties file types
    • Support for the ITS (Internationalization Tag Set) standard has been added for .xml files
  • Workbench Improvements
    • Match percentage is now displayed in the workbench for TM hits (New & Classic)
      • Exact matches are shown in green
      • Fuzzy matches are shown in yellow
    • Match percentage and TM Vault is now displayed in the workbench for pre-filled TM (New & Classic)
      • The data labels display adjacent to the input box
      • The labels are cleared if the pre-filled text is edited
    • Segments pre-filled with an exact translation memory match can now be locked by the workflow (New & Classic)
      • Locked segments cannot be edited
      • Phase rules can be added as part of the workflow to control in which phases segments remain locked
    • The Next button will skip past segments that have been pre-filled with an exact match and locked (New & Classic)
    • TM hits and Glossary hits no longer implicitly search the logged-in user's personal TM Vaults and Glossaries (New & Classic)
      • TM hits also no longer implicitly search the project's "save to" vault (New & Classic)
      • TM hits and Glossary hits can be configured to search the logged-in user's personal TM Vaults via the "Show Options" panel (Classic only)
    • Added the "Accept All Segments" function to the Classic Workbench (was already available in the New Workbench)
    • Source text editing can be enabled in the workbench (Classic only), allowing for the correction of OCRed text from within the workbench
      • Feature must be enabled for the community by the system admin
      • Editing the source text for a segment will auto-refresh translation resources (e.g. TM hits) upon save
      • Editing the source text after a segment has already been translated will remove the previously-created TM for the segment
  • System Administration Improvements (premise & dedicated cloud installations only)
    • Improved jobs session monitoring and administration
    • Added support for promotional codes
    • New "Open Signup" setting under the Customization page can be enabled to allow anyone to join a community via the community's signup page (i.e. no invite email is needed, just a link to the URL for the community's signup page)
    • Login IDs no longer have to be an email address
    • Users auto-provisioned via a Lingotek Inside integration can now be added to the OID provider table via the "Allow Local Login" button, thus allowing the user to log in directly to Dashboard
  • Platform Improvements
    • Backgrounds jobs and processes now use Quartz Scheduler
    • Updated JQuery to version 1.7.1
    • Updated Struts to version 2.3.1
    • Updated LibreOffice to version 3.5
    • Updated Flot to version 0.7

New APIs

  • getDocumentProgress - single API call which returns the progress percentage for all targets within a document
  • listWorkflows - list all of the workflows for the current active user
  • prefillDocument - pre-fill the segments of a document. Exact matches and matches with a qualifying threshold will be automatically inserted into the target. Vault precedence is determined by the order of the tmVaultId parameter values (highest first)

Modified APIs

  • addContentDocument - added the optional "note" parameter, which allows you to add a document-level note with your upload
  • addDocument - added the optional "note" parameter, which allows you to add a document-level note with your upload
  • addProject - added the optional "workflowId" parameter, which allows you to specify which workflow the project should use
  • getPhase - added the "isMarkedComplete" value to the response
  • getTranslationTarget - added the "isMarkedComplete" value to the response

Bug Fixes

  • Performance improvements when loading the Workbench (New and Classic)
  • Performance improvements when navigating between segments in the New Workbench
  • Performance improvements when loading the details page for projects and documents
  • Notes in the Classic Workbench can be deleted
  • Improved error handling when uploading Microsoft Office files
  • Find and Replace correctly handles documents with zero translatable segments
  • Add Target for TM option can be applied to multiple documents uploaded at project creation
  • List of documents on a project correctly displays document whose targets have mixed due dates
  • Paste Text method for document upload uses the selected document type filter
  • New Project page loads even when the Premise Features for the community have never been set
  • System admin can now clear out background jobs that are not moving through the queue

 


On this page:


Need Support?

Visit the Lingotek Support Center.