(PR#3942) Files in the dataset/project workspace are viewable while updating: Previously when a file updated it was rendered as disabled and no data from it showed in the workspace. Now the data prior to the start of the update is shown until the update has completed and it can be used until the update is finished.
(PR#3971) Save query on CMD+S anywhere in query container: Selecting CMD+S while in the query window of of a project or a dataset (for both SQL and SPARQL queries) will save the active query. If the query hasn't previously been saved or is owned by someone else, the user will be prompted to save the query:
 |
(PR#4005) SPARQL query with CONSTRUCT returns a TTL file: Previously when a user ran a SPARQL query that included a CONSTRUCT statement, the results were presented in a table:
 |
Now the results are presented in a TTL file:
 |
In the case that the data can't be returned as a TTL file, the results will be shown in a table with an error message:
 |
for more information on querying in SPARQL, see our SPARQL tutorial.
(PR#3973) Fixed an error when sorting numeric columns with the UI sortfunction: Previously the sort could return numerics sorted as strings:
 |
Now they are correctly identified as numerics:
 |
(#PR4947): Added the error message "Aggregation used in GROUP BY" when a user tries to include an aggregation in an SQL GROUP BY clause.
(#PR4940): Added the error message "ORDER BY <column index> syntax is not supported by data.world SQL". This functionality has been deprecated in SQL.
(#PR4930): data.world SQL updated to allow WITH subqueries to be named after tables as in:"WITH foo AS SELECT * FROM foo" where foo is a table name in the SELECT statement.
-------------------------------------------------------------------------------------------------------------------------
(PR#4971) Fixed a sorting error onstreams: The API was modified to allow updates of streams endpoints (more information here). In the initial version of the API, the records were sorted newest to oldest. However in order to make the API work with our new update component which allows records to be updated as well as appended to a dataset, the sort order was changed to oldest to newest. More information about streams can be found here.
(PR#$4957) Data syncs through the public API are now limited to 10 per user per minute: If a user attempts more than 10 syncs in a minute, an error message (429 too many requests) will be triggered and the additional requests will fail. The user will also receive notification of how long to wait before attempting to sync again. This fix only affects automatic syncs run through the public API. Syncs run from data.world's autosync are not impacted.