Phrase TMS Releases 2022

When importing to a TM with .tmx or .xslx files and the Remove tags option is switched on, text resembling the tag is now escaped.

Example: {1>The<1} becomes \{1\>The\<1\}.

Deprecated languages (removed in June 2022) were removed also from submitter portal page.

When a term with only 1 character is imported into a term base, the match type will be set to Exact (instead of Fuzzy).

The subscription term edition has been replaced with plans.

7 new languages/locales added.

If optimizing translation memory matches and the Previous OR next segment context as 101% checkbox was checked in file import settings, the first and the last segment were displayed as 101% match even though the context was different (next segment for first segment, previous segment for last segment). This has been fixed (first and last segments now appear as 100% matches).

In projects created via the submitter portal, the automatically generated quote contained steps to which the selected provider was not assigned. This has been fixed.

Saved custom filters for TM and TB pages were not displayed in the dropdown menu. This has been fixed.

Quality tab was not displayed in ENT Starter organisations. This has been fixed.

Only the first 50 project templates are displayed when setting up APC. This has been fixed.

When creating a job from online repository via API, the file path was not visible in the job tooltip. This has been fixed.

Billing details in the subscription overview page have been updated.

Formality feature for DeepL is now supported in Phrase Translate.

In multilingual CSV files, extra quotation marks were added to rows with text in quotation marks in the completed file. This has been fixed.

Some analyses were not properly named in projects created via Submitter Portal. This has been fixed.

New analytics dashboards are now available with processed word count, on-time vs overdue delivery, and language quality assessment (LQA) charts. Each dashboard provides multiple filtering options and chart results can be downloaded in multiple formats.

A video review of the dashboards.

Custom metadata fields can now be added to project and project templates. Custom fields can be used to filter and sort on project list and jobs pages.

A new fully managed engine is now available for Phrase Translate.

UI buttons are now updated with new shape and colors.

Multilingual .XLSX files with multiple sheets were not being auto-detected. This has been fixed.

In content search for projects, amended segments that no longer match search criteria do not disappear from the search results. This has been fixed.

The alignment feature was generating invalid .XLSX files. This has been fixed.

Time settings in the APC settings were randomly changing. This has been fixed.

When edited, APC timezone was randomly changing. This has been fixed.

Context key for PLIST format now consists of values from all preceding key elements.

Subscription period options are changed and 12 months period is now pre-selected by default.

For some language pairs, MT was failing with unsupported language pair error when glossary is attached to Phrase Translate profile. This has been fixed.

Unsharing projects or jobs by the Buyer was removing and failing to update recently visited items of the user to whom the project or job had been shared. This has been fixed.

COTI's failed import logs and files were not being moved to error folder. This has been fixed.

It was not possible to unassign multiple jobs assigned to a Vendor at once. This has been fixed.

In some cases, APC was not respecting the interval set in the APC settings. This has been fixed.

Ability to log in through single Suite login and access TMS and Strings through a new dashboard. Global profile, organization settings, and subscriptions can be managed from a single place.

Preview of the login procedure.

Submenus for filters and groups on jobs were not working properly. This has been fixed.

In some cases, project numbers on the project list page were split into two rows. This has been fixed.

Job history was not being displayed in Continuous Jobs on the Vendor's side. This has been fixed.

Saving edits to existing APCs was returning an error. This has been fixed.

The editing form for APCs created as Continuous Projects was not loading properly and the suffix automation was being added to their name. This has been fixed.

Upon login, submitters were being redirected to an error page. This has been fixed.

Search across projects enables users to look for any expression in the source and/or target. The user may narrow the results based on advanced filters or regular expressions.

Another set of new languages/locales (94) has been added.

Password rules do not require special characters anymore, but also will no longer accept passwords that contain a user's username or email inbox name.

There is a new UI look and improved navigation with breadcrumbs in the application header.

Upon saving the APC, the number of selected monitored files was being changed. This has been fixed.

Upon creation of the APC without selected files, a new project was not being created. This has been fixed.

In some cases, the Vendor's Linguist of a shared job was receiving Next step ready e-mail notification from the Buyer's owner instead of the Vendor's owner. This has been fixed.

Spellchecker for Bengali (Bangladesh) produced huge amount of alerts, marking almost all the words as error. This has been fixed.

In a project created via APC importing files inside a ZIP folder, the target ZIP subfolder was not being exported back to the repository upon completing the project. This has been fixed.

The option Assign providers from a template was allowing to assign single jobs to the same Vendor of a shared project. This has been fixed.

Project references can be filtered and sorted.

When an organization transferred a project with a job to a Buyer, and subsequently this organization shared the job in this project to a vendor to whom they transferred the project, the job was duplicated. This has been fixed.

Google Sheets are not recognized as a supported file format by Continuous Jobs. This has been fixed.

A new set of English locales (88) is now available as supported languages.

A new option is now available in the Access and Security settings which allows Vendors of shared jobs to upload reference files to projects. The option is disabled by default.

Update target could not be used in workflow steps other than the first one. This has been fixed.

Some Chinese terms were not being displayed in the CAT pane if they were inside a long segment with other TB suggestions. If they were alone in a segment, they were being displayed. This has been fixed.

In projects having Serbian or Serbian locales with Cyrillic as target language, MT suggestions from Amazon Translate were in Latin. This has been fixed.

A new error message is now displayed when a Linguist is trying to accept or open a job that has already been claimed.

The user can now choose whether the changed segments are confirmed or not after selecting Update target.

A new import option is now available for AsciiDoc files.

The formatting applied to entire cells in Multilingual Excel files was lost upon export. This has been fixed.

Projects created from a template had a different TM order than the template they were created from. This has been fixed.

In projects created via the Submitter Portal with the financial setting Create for provider as well enabled, provider's analyses were created by language and not by provider. This has been fixed.

An error report page was shown instead of a message when trying to delete a project template which was assigned to some service. This has been fixed.

It was not possible to remove the Vendor of a shared project. This has been fixed.

In some cases, placeholders instead of languages were shown in the error message of incompatible languages with the project settings. This has been fixed.

It was not possible to import a Google spreadsheet as a Multilingual Excel file using the auto-detection feature via the Submitter Portal. This has been fixed.

All customers in all regions can purchase subscriptions in any of the following currencies: USD, EUR, GBP and CZK.

When entering a VAT number for a country where VAT is not applicable, the subscription checkout did not work as expected and an error message was not displayed to users. This has been fixed.

In shared projects and jobs where Buyer had a custom URL, this URL was mentioned in Vendors' e-mails. This has been fixed.

When a project or a project template has many target languages and the user tries to select a TM with many languages for All target languages, this TM is not shown in the UI for relevant languages and is not found when filtering the name. This has been fixed.

When reassigning a job with LQA enabled from an external Vendor to an internal Linguist, the LQA was reset and previously flagged issues were not calculated in the final score. This has been fixed.

Quantity strings are now imported in Android String XML.

Jobs created via Submitter portal couldn't be downloaded from the Download all button in the widget page if a reference file had been deleted. This has been fixed.

Reference files were sorted randomly. This has been fixed.

Export workflow changes crashes on exotic character in job name. This has been fixed.

Delete all translations feature did not work for more than 50 jobs at the same time. This has been fixed.

When a Linguist account was later changed into a Project Manager account, filters in the Project view dropdown menu stayed the same as for a Linguist account. This has been fixed.

It was possible to enable the option to Import files as Continuous Jobs for HTML files imported via some connectors, even though HTML is not a supported file format for continuous jobs, and jobs couldn't be created. This has been fixed.

When sending LQA Scorecard to a Vendor, it was not possible to search for the provider by Vendor's name. This has been fixed.

In Android String resources file type, actual percentage signs and double percentage signs were converted to tags. This has been fixed.

Primary segmentation rules were not automatically selected when creating jobs via API. This has been fixed.

The drop-down list when filtering clients in the Jobs view contained only the first 50 items. This has been fixed.

Time zone value in the user settings was reset to Europe/London when editing. This has been fixed.

The formatting around comments in some XML files was changed upon export. This has been fixed.

Pre-translate check box was checked or unchecked based on the latest job and not the project settings. This has been fixed.

When reassigning a job with LQA enabled from external Vendor to internal Linguist, the LQA was reset and previously flagged issues didn't count to the final score. This has been fixed.

In-Context Preview was not generated if Process HTML code or Convert to Memsource tags options were applied in MS Excel file import settings. This is fixed.

There is now an email template for sharing LQA scorecards with providers. Email templates can be selected in the Send LQA Scorecard modal.

Usernames option is now displayed in LQA scorecard in Access and security settings to disable showing names in downloaded LQA scorecards. When disabled, all names are replaced with user id and organization name.

TM ID is now added next to TM name when selecting TMs and adding them to a project.

A new job count macro is available.

In ASCIIDOC files, double curly brackets were not converted to tags correctly with regular expressions. This has been fixed.

Some symbols caused incorrect interpreting of literal monospace in ASCIIDOC files. This has been fixed.

Project related macros were not being applied when emailing multiple jobs in the same project to the same provider. This has been fixed.

In-Context Preview was not generated if Process HTML code or Convert to Memsource tags options were applied in MS Excel file import settings. This has been fixed.

A set of an additional 40 languages are supported.

The project name is now added to the filename of downloaded scorecards.

False XLIFF tags QA warning are displayed when ?xm-replace_text? XML element is not closed. This has been fixed.

The Missing term QA check returns false positives for some Japanese terms even though the term was used in the target. This has been fixed.

Norwegian locales are not mapped correctly in translated .DOCX files. This has been fixed.

LRI and PDI marks were added around tags in the bilingual DOCX for LTR languages. This has been fixed.

When an HTML file with a  jQuery script is imported, the underlying structure of the exported completed file is corrupted. This has been fixed.

Projects could not be created using Safari on macOS. This has been fixed.

The last state of the Split file in all workflow steps option is now remembered per user.

The list of UI languages which the user can choose from has been decreased to 11. We now support English, Chinese (Simplified) , Czech, French, German, Italian, Japanese, Korean, Portuguese (Brazil), Russian, and Spanish.

Links to LQA scorecards now expire after 1 year instead of 5 days.

When adding TMs to a project, they can be filtered based on ID, source and target languages, and project metadata (Client, Domain, Subdomain, etc.). Strict locale matching is also supported and when used, only TMs strictly matching project languages and locales are listed.

Another set of supported languages has been added.

Services listed in the Submitter Portal are in a random order. This has been fixed.

An unexpected error occurred when changing Submitter to Linguist or Project Manager. This has been fixed.

Vendor's TMs used in shared jobs were not visible in the project page. This has been fixed.

The email templates page in Settings was not accessible in some editions. This has been fixed.

APC creating two identical projects one minute apart. This has been fixed.

Splitting of CDATA section was causing invalid XML to be generated. This has been fixed.

Provider from a deleted job was not listed when assigning a quote. This has been fixed.

There was no option to select all analyses on the project page. This has been fixed.

Automation Widgets have been renamed to Submitter portals. There is no change to functionality.

Job reports now contain first and last names of assigned linguists.

The Lengoo HALOS MT engine has been added to Memsource Translate as a manually managed engine.

Selected notifications can now be disabled in Submitter portals by selecting empty values in the Email templates section.

Linguists had access to TMs even when the associated jobs were removed from the project. This has been fixed.

Source segment capitalization was used in the machine translation output from DeepL, which was not correct for all languages. This has been fixed.

When more jobs are assigned to the same provider (Linguist/Vendor) within one language pair, only one notification email is now sent.

Creation dates of projects, jobs, TM, TBs etc. now display date and time for entities created in the current year. Entities created in previous years will have a year displayed instead of a time.

Buttons on the linguist Jobs view are now color coded to better distinguish each command.

Create job buttons are now disabled unless a file is selected to prevent users from forgetting to do so and losing their import settings configuration.

The Alexa Translations A.I. MT engine was added to Memsource Translate.

When adding an email address into the CC or BCC copy in the email tool by selecting a suggested email address stored in the browser’s autofill feature, it was propagated into all CC, BCC, and subject fields. This has been fixed.

The timezone set in the organization is not promoted to the user creation form. This has been fixed.

Drop-down menus were not closing when navigating to subsequent fields using the TAB key. This has been fixed.

Updating source imports the original content and unconfirms already confirmed segments. This has been fixed.

Linguists were not able to access analyses and quotes via API. This has been fixed.

It was possible to save templates with same vendor and provider. This has been fixed.

Translation jobs are now in one centralized location on the new Jobs page with advanced filters and customization options. Demo video.

Vendors can now add or change due dates in shared jobs when allowed by their Buyers. Buyers can set the rights in Access and security section in setup, project templates and individual projects.

It's now possible to disable escaping of < and & entities in completed files. The new option is available in HTML file import settings.

In a project or a project template with more than 12 target languages, it was not possible to scroll through a list of available target languages when selecting a TM for only a particular target language. This has been fixed.

It was not possible to place TMs in the desired position when defining priority order. This has been fixed.

When individual languages were selected for TM assignment in a project, only TMs for all project's target languages were displayed. This has been fixed.

Update target now supports multilingual files (XLIFF, XLSX, CSV). Only one language can be updated at a time.

Convert to Memsource tags is now available for .MIF (Adobe FrameMaker) files.

When adding a TM to a project with one WF step, there is now the option to select whether the TM should be added to this WF step (individual WF step) or to all WF steps. If the latter option is selected, the TM will automatically appear in other added WF steps.

When a TM is added to a project and Selected section has no TM in WRITE mode, the added TM didn't have the WRITE checkbox turned on by default. This has been fixed.

There was a problem with loading a quote page when project has thousands of analyses. This has been fixed.

Filters applied to Users were removed after performing an action. This has been fixed.

Vendors with shared jobs were able to see all quotes in a Buyers' projects through API. This has been fixed.

The source locale was not displayed correctly for the selected TM in a project. This has been fixed.

APC: Completed analysis leads to the premature completion of the project. This has been fixed.

When using Update target, segmentation rules for the target segment can now be changed.

User can now select all analyses by using the Select all <number> analyses checkbox (the same behavior as for jobs).

It's now possible to use Convert to Memsource tags in Quark Xpress Tags (XTG) files.

A Broken pipe error occurs when importing some XLSX files into a TM. This has been fixed.

There is an overloading problem when creating/editing a user. This has been fixed.

Trying to open a deleted and subsequently undeleted TB from the term base view triggers an internal server error. This has been fixed.

When Penalize multiple 101% matches by 1% is selected, TM matches with the same target are not penalized (even among multiple TMs). TM matches with different targets are still penalized.

Incorrect Nepali spellchecker warning for ।. This has been fixed.

MTQE score was being generated for third party MT engines. MTQE functionality has been removed from all MT engines, that are not connected to Memsource projects through MTr.

It is possible to create a working PROMT MT engine with invalid credentials. This has been fixed.

Webhooks can now be configured for job exported and job unexported events.

Some punctuation marks cause spellcheck error in Arabic. This has been fixed.

Some tags are not preserved when Google Translate is used under Memsource Translate. This has been fixed.

Top items of long lists of target languages in the scrollable dropdown of the Select TM window were missing as the scrollable list extended beyond the top of the screen. This has been fixed.

It was not possible to download the LQA scorecard nor send it to linguists from the project page when different settings for the same LQA profile were applied to different jobs. This has been fixed.

Links in reference notes were only clickable if located in front of the more… button. This has been fixed.

DeepL translated Traditional Chinese jobs with Simplified Chinese, when it only supports Simplified Chinese as a target language. This has been fixed.

The LQA scorecard issued a FAIL status if the score was equal to the set threshold. This has been fixed.

Spreadsheet documents containing more than 10,000 cells with conditional formatting and are processed with Multilingual Excel, Excel with html (Process HTML code) or Excel with Convert to Memsource tags filters will be rejected as too large/complex.

Ultimate and Enterprise users can now prioritize TMs added to the project and decide which TM is more relevant/important. Based on this TM priority files are pre-translated and suggestions are displayed in the CAT pane.

DeepL MT engine has been added into Memsource Translate as a fully managed MT engine.

Rozetta MT engine has been added into Memsource Translate as a fully managed MT engine.

Webhooks can now be configured for job deleted and project deleted events.

The in-context preview becomes stuck for some .XLSX files. This has been fixed.

A new project settings field has been implemented under Completed file name and export path to specify the target folder for all files imported from an online repository if the jobs are Continuous Jobs or not.

Failed webhook calls can now be replayed on the History page in batches or individually. Only failed webhook calls can be selected.

Exported price list files were empty. This has been fixed.

Updating a source file triggered validation errors with some file types. This has been fixed.

Extra code was being added when importing YAML files with the Markdown subfilter. This has been fixed.

Alignment function was triggering an unexpected error. This has been fixed.