Seitenhierarchie
Zum Ende der Metadaten springen
Zum Anfang der Metadaten

The following list shows the important features of translate5 in bullets without detailed explanation.

It aims at giving an overview of the software and is addressed to people, who know how translation software and translation business works.

This list is the most complete list of features, that exists. It will be kept up2date, as translate5 progresses further. More details can be looked up in translate5 JIRA.

The page tree on the left shows some features more in details, yet it is far from complete and not up2date.

General features

  • Tabular views of translate5 (including editor) 
    • are sortable and filterable. Filter can be set as AND filters in several columns.
    • are "endless" grids, which means you can scroll through the table from the first to the last entry with great performance, even with 100000 entries in a tabular view (implemented for editor; soon also for other views).
    • Order of columns can be changed by drag and drop
    • Columns can be hidden and shown
    • default column order and visibility can be configured on system level for task overview and user overview

Editor

  • All general features for tabular views (see above)
  • If language resources are assigned, for all assigned language resources the following features are offered: 
    • Concordance search (can be narrowed to single language resources)
    • Fuzzy matching (queries all assigned language resources, when a segment is opened)
  • In-Context review
    • with visualReview plug-in active (visualReview is a paid plug-in that helps financing the open source development of translate5 - please contact MittagQI)
    • available for practicably all file formats (e.g. MS Office, Indesign, FrameMaker, HTML, etc.)
  • In-Context translation (coming soon)
  • In-line Track Changes
    • with TrackChanges plug-in active  (TrackChanges is a paid plug-in that helps financing the open source development of translate5 - please contact MittagQI)
    • MS Word style track changes
    • different background colors for different users
    • Changes are recorded across all workflow steps, that translate5 offers
  • Cascaded file tree with folder hierarchy for the files of a task and the reference files of a task
    • files sortable, new sort order is respected on export
  • Potential usage as integrated editor in other applications without task overview etc. 
  • Great performance with large tasks - 100000 segments and 1000 files in a tasks work seamlessly
  • Optional pivot language (if reviewer does not understand German, but an English translation is available)
  • Search and replace (with replace all, wildcards, regular expressions and case-sensitive options)
  • Comments on segment level, filterable for content, date and creator of comment
  • Altering of font size
  • Keyboard short cuts for all important user interactions
  • Auto propagate of repeated segments on saving (configurable if it is down automatic, asked each time for which segments it should be done or never done)
  • Max. and min. length restrictions (so far only supported for XLIFF 1.2 import files; also works across all segments of a trans-unit)
  • full tag and short tag mode; hide tags possible with editing disabled
  • Bookmarking of segments
  • Autostatus column shows 
    • workflow step, in which segment was changed the last time
    • if segment was edited or just saved without changes or auto-propagated
    • if segment is untranslated
  • Automated QA checks on segment level:
    • Terminology check against translate5 TermCollections
    • Empty segment check
    • Tag check (potentially configurable, that segment can only be saved without tag errors or only with different white-space tags than the source; for review tasks the comparison is done against the translation at import time)
    • Spell-, Grammar and Style-Check based on languageTool (Open Source)
  • Manual QA based on
    • DQF/MQM on subsegment level (definable on system and task level; error types in cascading levels)
      • including DQF/MQM statistics (visualized and exportable as XML)
    • QA flags on segment level (definable different flags on system level)
    • Manual status (definable different status on system level)
  • Meta data columns (by default visible) for
    • Segment no. in task
    • match rate in TM system
    • match rate type (shows, where a segment content comes from)
  • Meta data columns (by default hidden) for
    • QA flags
    • Manual status
    • last editor
    • edited at workflow step XYZ
    • target language at import time (for review tasks)
  • Optionally add white space tags (non-breaking space, tab, soft return)
  • Display of number of segments in current filtering
  • "Hovering open segment": If segment is opened, it is possible to scroll under it all segments of the table and also filter and sort these segments
  • Segment actions:
    • Save and jump to next in workflow
    • Save and jump to next
    • Save and jump to previous
    • Abort and jump to next in workflow
    • Abort and jump to previous in workflow
    • Abort and jump to next
    • Abort and jump to previous
    • Refocus open segment in context (see "hovering open segment")
    • Reset segment to non-edited state
  • Switch of default modes for font size of and visible columns and panels (normal mode, details mode and simple mode (the latter only available with visualReview plug-in active)
  • Multiple target language columns (if imported like that in CSV format; see "task overview")

Task overview

  • All general features for tabular views (see above)
  • Which task actions are available for a user depends on the rights a user has
  • Add tasks:
    • Importable file formats:
      • All file formats supported by Okapi Framework through Okapi integration
        • An Okapi bconf file can be passed in a zip import package for translate5 to adjust Okapi filter settings (like ignoring Excel cells with certain colors, configuring xml, etc.)
      • SDLXLIFF
      • XLIFF 1.2
      • OpenTM2 XLIFF
      • Across XLIFF
      • Star Transit bilingual file pairs
      • Bilingual CSV (also multiple target language columns possible, that lead to multiple target language columns in translate5 editor, usable for comparing different translations, e.g. with DQF/MQM)
    • decide, if
      • 100% should be editable
      • segments locked in original xliff should be editable
      • source language should be editable (does not export in all export formats, mainly made for CSV so far)
    • Add meta data
      • source language
      • target language
      • client
      • order number
      • delivery date
      • order date
    • Optionally use pivot language (if reviewer does not know the source language, but e.g. an English translation exists)
  • Associate language resources to tasks
    • All language resources are selectable, that are associated with the same source language, target language and client as the task is
      • automatic assignment of language resources, that are assigned as default language resources for the client, to whom the task belongs
    • Decide if match analysis should include internal fuzzy matches (fuzzy matches produced by segments that are only translated while the task is translated)
    • Decide if pre-translation should be done against associated termCollections and Translation Memories and if yes up to which match rate
    • Decide if pre-translations should be done against associated MT resources (only for segments still empty after TM pre-translation)
    • Decide, if terminology should be tagged in segments
  • View analysis; export analysis as excel
  • Assign users to task
    • Multiple users / workflow step
    • Workflow roles 
      • translator
      • reviewer
      • check of review
    • Configure the editor more in detail for users or roles
      •  which source and target columns can be seen
      • if the column name of source and target columns can be read
      • if only DQF/MQM tags can be set, but the content can not be changed
      • if non-editable target columns (target text at import time) can be seen
  • Export
    • task in original format
    • task with diff compared to the import stage (main use case: sdlxliff with change marks, that work with Trados Studio track changes)
    • task as xliff 2.1 with reflection of translate5 in-line track changes (who changed what when) and segment meta data (please note: this track changes is complete workflow based, the "diff to import stage" in the previous bullet does not reflect changes within the translate5 workflow steps)
    • MQM statistics as xml
  • Change a number of task properties, e.g. task name, project manager, delivery date, delivery date (real), order date (only with PM rights, or if special rights are configured in ACL
  • See, filter and sort task meta data
    • source, target and pivot language
    • client
    • task status
    • task name
    • order number
    • number of words in a task
    • associated users with their workflow roles and workflow status (task table can currently not be filtered along this column)
    • associated language resources (task table can currently not be filtered along this column)
    • reference files available
    • terminology associated
    • project manager
    • order date
    • delivery date
    • delivery date (real)
    • number of files in a task
    • 100% matches can be edited
    • task is a translation or review task
    • locked segments in imported xliff are also locked in task
    • source language is editable

Workflow

translate5 has implemented a flexible workflow mechanism:

  • To every task the following job types can be assigned:
    • translation
    • review
    • check of review
  • For every job type multiple users can be assigned, resulting in multiple jobs within one workflow step
  • As soon as all jobs of one job type are finished, the workflow jumps to the next job type / workflow step and the users of this workflow step and the project manager are notified
  • The notifications after finishing of review and "check of review" list all changed segments as HTML mail, including tags, term markup, MQM markup, comments, TrackChanges, etc.
  • On job assignment an email is send (configurable in the system config)
  • Each job has one of the following workflow job status
    • open
    • waiting
    • finished
  • The project manager is completely free:
    • Each workflow step can be assigned to users, even if no other workflow steps are used
    • At any point in time new users can be assigned to any workflow step or removed
    • At any point in time the job status of a user can manually be changed to any value.
      • If this makes sense and works in regarding of the current todos of other users is the responsibility of the project manager. He can change also the job status of other users or notify them manually about consequences, if necessary

New workflow roles or steps can easily be extended on a programming level for a certain installation.

Users

  • All general features for tabular views (see above)

Manage users

  • Create and delete users
  • Associate users to clients
  • Set preferred GUI language of user
  • Set system roles of users (do not mix up with workflow roles)
    • it is configurable via ACL which system roles can be set by a specific system role (example: PMs can only set Editor roles, admin can set all roles)
  • Set password of user
  • Set default languages of this user. When a task is created, users for whom a default language is set are automatically assigned to this task as proofreader.
  • A parent user can be set. This user only has any effect, if in the database-based ACL for a PM role is set, that it is not allowed to see all users. If this is the case, a user can only see those users in the user management, for whom he is the parent user. And in addition all users that are children of his children users and so forth.
  • On database level it can be configured, if it is allowed to edit or delete a certain user

Language Resources

Language resources in translate5 are TermCollections, Translation Memories and Machine Translation resources.

  • All general features for tabular views apply (see above)
  • The following resource types are currently available:
    • TermCollections:
      • Internal translate5 TermCollections
        • Import TBX
    • Translation Memory resources
      • OpenTM2
        • Import / Export TMX and OpenTM2 binary TMs
    • Machine Translation Resources
      • Lucy
      • Google
      • Moses default
      • SDL LanguageCloud
      • Globalese (not as part of Language Resources, but with a special integration)
  • Language resources can be associated to multiple clients
  • Language resources can be associated as default resources for clients. In that case they are automatically associated for tasks of these clients.

Client management

So far the client management is a simple tabular list of customer names and numbers.

All general features for tabular views apply (see above).

TermPortal

At the current stage the TermPortal aims only at serving end-users as a simple interface for looking up the terminology of their companies.

Every user sees the terminology of all TermCollections, that are associated with the a client, that the user itself is also associated with (see user management and language resources).

This can be used by language service providers to serve different clients with the same translate5 instance.

The TermPortal provides:

  • Wildcard search
  • Live search
  • Fast and very easy to use interface
  • Shows the attributes of term entries, language level and term level
  • Also supports none-standard term attributes
  • Term attribute labels can be customized on database level and translated with the translate5 GUI translation mechanism (xliff based) 
  • Term attributes for processStatus and administrativestatus are also shown as icons

InstantTranslate

InstantTranslate is a Google-Translate-like interface for automatic on-the-fly translation of content.

It also can be used as an easy interface to search through translation memories.

The interface sends the entered text to all language resources, that are associated with a client, that itself is associated with the current user.

This means, the text is send to TermCollections, Translation Memories and Machine Translation resources at the same time.

The answers of all resources that have matches are listed.

For Translation Memories also Fuzzy Matches are listed - the difference of what does not match with the requested translation is highlighted and the match rate shown. 

InstantTranslate is only available for translate5 users, that support the community driven open source development with a support and development contract for translate5.

Login, Session, Password Rest

  • Session time out can be set in the configuration
  • Password can be reset in a secure way by the user
  • Password can be set inside the application by a user for himself and by a PM or admin for everyone

Change log plug-in

In the task overview in the right bottom corner a button opens the change log of translate5.

It is filterable by translate5 JIRA issue type.

It only lists the issues that are relevant for the currently logged in translate5 role.

If there are new issues after login (because of a recent translate5 update) the change log window automatically pops up and only shows the new issues.

As all plug-ins it can be deactivated.

Maintenance mode

A maintenance mode allows server admins

  • to notify users time X before a translate5 update 
  • to lock the login time Y before a translate5 update
  • to lock the application for any interaction at point Z in time

After the update is done it automatically requests the browser of the user to refresh itself so that the Javascript is refreshed with the updated code.

Database-based features

  • Detailed configuration of rights and roles possible in translate5 ACL table
  • More than 200 options for configuration on system level for translate5 in Zf_configuration table
  • "Materialized view" of a task for fast performance on reading segments
  • Saving of complete segment history (each editing step of a segment is saved in a history)
  • Saving of the time it takes for a user to edit a segment

Features for developers and good performance

Worker

The worker concept in translate5 allows

  • to run processes in the background
  • to fork processes
  • to chain processes with dependencies
  • to load balance and split processes, especially when using micro services for background tasks

Editor only usage

Integration of only the editor of translate5 into another application. All other features are used via API, if necessary.

Skinning / customizing the layout

It is possible to 

  • completely customize layout and content of the pages that users see, that are NOT logged in
  • customize the head panel of translate5, when logged in
  • to completely customize the whole look and feel of the application by creating a custom ExtJs template.

API

In general: All actions that can be done by a user in the GUI can also be automated via translate5 REST API (by design)

Installer and Updater

Translate5 provides an install and update bash script for linux and a batch script for windows systems.

  • The installer guides the admin through the whole installation process, downloading and installing also all needed third party packages
  • The updater updates the installation with just one call with latest publicly available translate5 version; same for potentially updated third party packages
  • Users of a customized version just have to provide the customized version as zip file to the updater
  • Further features of the install-and-update script:
    • Fast enabling / disabling of the maintenance mode
    • check the current installation usage (if there are running imports for example)

 

 




  • Keine Stichwörter