Release Notes


Development including all bug fixes can be tracked on the Github repository for Global Signbank. New user features implemented from 2018 onwards are summarised below.

July 2023

  • For all datasets, 'translation equivalents' have been migrated to a new field 'senses', where multiple senses of a sign can be distinguished in terms of their translation equivalents, and in terms of one or more example sentences per sense. The example sentences can be uploaded as a video, and receive a translation in all of the description languages that are specified for a dataset. Thus, a video in NGT can have a translation in both English and Dutch. Each sentence can be classified in terms of sentence type (declarative, interrogative, ...) and can be marked for negation.
  • Dataset managers can find a function 'Map keywords to senses' in the Analysis menu, to quickly group/revise the old translation equivalents ('keywords') in terms of distinct senses


May 2023

Change in field value: handedness value '2n' (non-symmetrical) has been changed to '2t' (two hands move in tandem in contact with each other) because of continuing confusion.

Changes to functionalities and layout in the interface (there are many more in this period, for details see Github):

  • Queries can be saved by the user; see the Query View in the Analysis menu
  • The edit history for a sign is now saved and can be displayed in one of the tabs when inspecting a sign

September 2020

  • In several edit views, entering information has been facilitated by enabling more keyboard navigation: confirming a menu choice (by clicking 'OK') can happen by pressing the enter key, and this also moves the cursor forward to the next field.
  • The editing of Lemma ID Glosses has been facilitated and can be carried out directly from the Edit view for an annotation gloss.
  • A button has been added in the Edit view to generate an image from the video, in case that this did not happen yet automatically.
  • A menu item 'Show all morphemes' has been added.


March 2020

  • The page where the dataset manager can select which field values to show/hide has been made much more user-friendly
  • Clicking on a dataset label in the header now leads to the info page for that dataset
  • The Public View has been revised so that Notes that are marked as 'published' now show up for anonymous users
  • For users with access to multiple datasets, datasets can now more quickly be selected and unselected


May 2019

  • Global Signbank now has a new url: https://signbank.cls.ru.nl (it's being developed and hosted at the Centre for Language Studies, CLS)
  • Tags are now hidden for anonymous users (which can only see 'published' signs).
  • A "Reference" field has been added to the dataset metadata, so that Dataset Managers can specify how they would like users to refer to their dataset in publications.
  • Dataset Managers can now hide field choices that are not relevant for their language (like specific locations in the phonology panel).
  • A list view for Lemma ID Glosses has been created
  • Lemma ID Glosses can be updated (not created) by export-import routines
  • Under the hood:
    • Many unit tests have been created that check whether new functionalities do not break older ones.
    • Videos and images are now stored in separate folders per dataset.
    • Video and image file names are now updated when an Annotation ID Gloss is changed.
    • Several small bug fixes.


January 2019

  • The relation between the Lemma ID Gloss and the Annotation ID Gloss is now a hierarchical one in the system. Lemma glosses are independently listed, and Annotation glosses refer to items on this list. When creating a new sign (Annotation ID Gloss), reference has to be made to an existing Lemma, or a new one can be created on the spot.
  • Both Lemma ID Glosses and Translations are now multi-lingual: just like the Annotation ID Gloss fields, they come in multiple languages (depending on the translation languages specified for your dataset). The CSV import-export routines and the search forms have been adapted accordingly.
  • In the menu Signs, the CSV import item has been split into two, one for adding new signs, one for modifying existing signs.
  • Added possibility to import Notes (CSV import)
  • Dataset Managers can now exclude specific field choices from phonology and semantics fields (to cut down the list of possible handshapes or movement types, if not all apply to a given sign language)
  • Several speed improvements
  • The whole top part of the interface has been redesigned. Available and selected datasets are listed next to the logo.
  • Images in the list view (search results) are now small videos that start moving when hovering over them with the mouse pointer


September 2018

  • The menu bar has been reorganised, and a new menu with analysis functions has been created.
  • A function to generate frequency reports for phonological fields has been added to the Analysis menu.
  • Frequencies of values in the selected datasets are displayed in the pull-down menus when editing a sign.
  • The ECV that is generated for each dataset now contains the description languages for that dataset (and no longer 'English' and 'Dutch' as a default).
  • Multiple search values can be selected (A or B or C) for all fields with lists of values. For text fields, regular expressions can be used for 'or' searches.
  • Self-registration of new users has been re-activated. Users can request access to one or more datasets. The account will be created as soon as the first dataset manager has allowed access to a dataset, and the new user will be notified by email.


August 2018

  • Selection of multiple search values is now possible in fields with lists of values.
  • A registration system has been set up. For each dataset to which access has been requested, the dataset manager receives an email to grant or deny access.
  • The Feedback functions have been restored.
  • The Lemma ID Gloss is now also multilingual.
  • Small improvements to the selection and use of Datasets.


February, 2018

  • A logo designed by the deaf Chinese designer Sun Qian has been added to the header, and serves as a button to go to the home page.
  • A Handshape list has been added, including both photos and a description in terms of distinctive features. The search form allow for searching for handshapes and for searching for signs based on handshape features.
  • ECVs can be exported for each dataset from the page View available datasets.
  • Selected datasets are listed in the header next to the user name. When no datasets are selected, all of them are listed here (and serve as the scope for searches).
  • Access to Datasets can be configured as 'read only' or 'read and write'. Searches for both signs and morphemes respect dataset permissions.
  • The panel Language & Dialect in the detail view has been removed. The Language field is now part of the Dataset value in the Publication Status panel. Dialect is displayed at the top above the glosses when specified.