Added

• Clickhouse connector

Security

• Upgraded Electron

Fixed

• Google BigQuery tabs are now closeable and delete scheduled queries

Changed

• Branding and labelling has been changed from Plot.ly/On-Premises to Chart Studio Cloud/Enterprise

Removed

• The "Plot.ly" tab has been removed and the login button has been moved to the "Schedule" tab
• The IBM DB2 connector has been (hopefully temporarily) removed

Security

• Grids created by Falcon are now non-world-readable by default

Added

• More informative Schedule tab with previous run results, tags for organization, next scheduled run time, API calls per day estimates
• Google BigQuery support

Fixed

• Query tab not available after connecting to Oracle (#516)

Added

• OnPrem: Added env variable PLOTY_CONNECTOR_BASE_URL to customise the
  connectorUrl set in a query metadata (#514)

Fixed

• OnPrem: Fixed connectorUrl in query metadata
  (https://github.com/plotly/streambed/issues/11310)

Fixed

• Scheduler: Handle query updates that change the names or the number of columns (#511)
• OnPrem: Fixed login link in modal window (#512)

Fixed

• Scheduler: Upload grid metadata as required by Chart Studio (#504)
• OnPrem: Fixed /external-data-connector page not accessible in private mode (#505)

Added

• Added UI to show scheduled queries (#472)
• Added UI to create, update and delete scheduled queries (#476)
• Support cron rules to schedule queries (#488)
• Added UI to name scheduled queries (#500)

Changed

• The refresh interval of scheduled queries created in previous versions of
  Falcon is now converted into one of the following: every minute, every 5
  minutes, hourly, daily and weekly (#488)

Removed

• Removed the chart editor and the export tab when running onprem mode (#496)

Fixed

• MSSQL: Fixed #299, #409 and #462 (#464)

Security

• Mitigated clickjacking by setting X-Frame-Options: DENY (#489)

Fixed

• Fixed query panel for non-sql connectors (#451)
• Fixed build failure with dtrace-provider (#421)
• Fixed broken oauth redirection in web app (#260)

Security

• Disabled nodeIntegration (#453)
• Ensured new windows are created using shell.openExternal (#453)

Added

• Connector for Oracle databases (#437)
• Backend: Added API for creating a grid to store results from scheduled queries (#444)

Changed

• UI: Added a vertical scrollbar to the database view
• UI: Don't hide the code editor when the chart editor opens (#445)
• Log: Reduced verbosity

Fixed

• Backend: Fixed error message on failed connection

Security

• Upgraded to Electron v2.0.1 (and Node v8)
• Prevent navigation on drag'n'drop (#436)

Many thanks to:

• Shannon Lal (@shannonlal) for contributing the new AWS Athena connector (#352) and joining the team of maintainers
• Kinuthia Ndung'u (@kndungu) for joining the team of maintainers
• Álvaro Abella Bascarán (@alvaroabascar) for his contribution to display query times (#358)

Added

• Connector to AWS Athena service (#352 contributed by Shannon Lal @shannonlal)
• Connector to query CSV files (via URL or drag'n'drop) (#375 and #385)
• New chart editor based on Plotly React Component Suite (#405)
• New code editor, resizeable and with improved performance (#411)
• New table view with row filters (#393)
• Test UI using jest and enzyme

Changed

• ElasticSearch: Use mocked servers for testing
• ElasticSearch: Run tests against elasticsearch v2 and v5
• Query: Show query times (#358 contributed by Álvaro Abella Bascarán, @alvaroabascar)
• Query: Improved error displays (#394)

Fixed

• CodeEditor: Fixed initial query (#371)
• ElasticSearch: Fixed connection to an elasticsearch server behind nginx (#377)
• Query: Prevent the code editor from opening before a table has been selected (#379)
• Test: Improved logging of backend failures

Many thanks to Kinuthia Ndung'u (@kndungu) and Rafael Pereira (@rflprr) for contributing the new data.world connector.

Added

• Connector for https://data.world

Changed

• test: implemented mock connectors
• test: updated DB2, Impala and Spark sample servers
• Enabled asar in the installers for Mac and Linux

Deprecated

• test: DB2 and Spark sample servers will be turned off in the near future.
  Please, instead, use the docker containers:
  • quay.io/plotly/falcon-test-db2
  • quay.io/plotly/falcon-test-spark

Added

• CHANGELOG.md
• Windows installer
• Mac installer
• Linux installers (tar.gz, deb and AppImage)

Changed

• Reduced app size
• Desktop app can run without logging into Plotly
• Do not download fonts

Removed

• OnPrem: Removed Download CSV

Fixed

• CodeEditor: Restored functionality to autocomplete keywords and table names
• CodeEditor: Fixed issue with the button Run becoming unclickable or hidden
• DB2: Fixed error with Falcon being unable to locate the ibm_db driver in Mac
• OnPrem: Fixed multiple OAuth issues
• OnPrem: Fixed export link
• Postgres: Fixed table preview to show tables in all the schemas
• Postgres: Accept schema names and quoted identifiers
• Redshift: Fixed table preview to show tables in all the schemas
• Redshift: Accept schema names and quoted identifiers
• Sqlite: Fixed the file selector in Windows and Linux
• Sqlite: Fixed the schemas overview

Added

• Windows installer
• Mac installer

Changed

• Reduced app size
• Desktop app can run without logging into Plotly
• Do not create SSL certificates if user is not logged into Plotly

Deprecated

Removed

Fixed

• CodeEditor: autocomplete
• CodeEditor: Run becomes unclickable or hidden
• DB2: unable to locate driver in mac
• Postgres: table preview
• Postgres: accept schema names and quoted identifiers
• Redshift: table preview
• Redshift: accept schema names and quoted identifiers
• Sqlite: file selector
• Sqlite: schemas overview

Security

• Documented the test databases

The Falcon SQL Client now supports authenticated requests through OAuth and a fully featured SQL Editor.

New Features:

• Adds Connectors for Hive via SparkSQL, IBM DB2 and Apache Impala.
• Public sample datastores for trying out the app with no fuss.
• All requests are now authenticated by default. This allows this app to be run as a public server while still securely ferrying data between plot.ly and the connected datastores. Authentication is enabled by default, but can be turned Off using AUTH_ENABLED setting.
• Restricted access through OAuth. Access to the app running as a server can be restricted with an ALLOWED_USERS setting.
• Introduces a standalone, interactive SQL Editor. The editor features data preview, typeahead, database schema previews, and inline data visualization.
• Drag-and-drop chart GUI for creating and exporting plotly.js charts.
• Optionally export charts and CSV data to plot.ly. Plot.ly has further editing, export, and online sharing options.
• CSV download of data to desktop.
• PNG download of charts to desktop.

The Plotly Database Connector (PDC) now supports authenticated requests through OAuth and a fully featured SQL Editor.

New Features:

• Adds Connectors for Hive via SparkSQL, IBM DB2 and Apache Impala.
• Public sample datastores for trying out the app with no fuss.
• All requests are now authenticated by default. This allows this app to be run as a public server while still securely ferrying data between plot.ly and the connected datastores. Authentication is enabled by default, but can be turned Off using AUTH_ENABLED setting.
• Restricted access through OAuth. Access to the app running as a server can be restricted with an ALLOWED_USERS setting.
• Introduces a standalone, interactive SQL Editor. The editor features data preview, typeahead, database schema previews, and inline data visualization.
• Drag-and-drop chart GUI for creating and exporting plotly.js charts.
• Optionally export charts and CSV data to plot.ly. Plot.ly has further editing, export, and online sharing options.
• CSV download of data to desktop.
• PNG download of charts to desktop.

• Reorganized app with horizontal tabs to eliminate need for vertical scroll
• Fixed Redshift schema preview
• Added missing font file
• Misc. UI and CSS clean-up
• Organized header links into an upper-right dropdown
• Toggle sample database credentials

See PR https://github.com/plotly/plotly-database-connector/pull/158 for more.

The Database Connector is now 100x easier to set-up by automatically issuing SSL certificates for you.

Some background:

• This application runs a webserver. on localhost. The plotly chart creator makes requests to this local webserver directly from the web browser.
• These requests happen entirely locally, it's just local web browser to local application, so no data is sent over the network. This makes queries wicked fast (no data is sent over the wire) and secure (you don't need to open up your database to any other IP ranges, just the ones on your local computer).
• However, since the chart creator runs on HTTPS, the browser automatically rejects any requests that happen over HTTP. If it didn't, malicious websites could trick you into thinking that all your data is encrypted, when really some requests aren't. This is bad news for localhost web severs because they run on HTTP by default. And even though data isn't sent over the network when connecting to a localhost webserver (therefore HTTPS-level encryption is not needed), the requests still get blocked.
• So previously, we required users to either block this setting in their web browser (an annoying process that you need to do every single time you want to connect) or create self-signed certs (an extremely annoying process to set up).
• But what are we to do? We can't just ask users to buy an expensive SSL certificate for their desktop applications... And besides, these applications are just running on localhost! To create an SSL certificate you need to have a registered domain name and a public facing website. We're just running this app on localhost, there is no domain name (it's just localhost!) and your desktop app is not open up to the world-wide-web (nor should it be!)
• Let's Encrypt to the rescue! Let's Encrypt is service for offering free SSL certificates. What we've done is created a server which generates a unique subdomain and a unique Let's Encrypt SSL certificates for that subdomain. We then re-route that subdomain's DNS to localhost:9495. The original proposal of this architecture was discussed in the Let's Encrypt forums. We've open sourced this certificate issuing server, too
• For example, when you start up your app, your app will make a request to this certificate and domain issuing server. After about six minutes, you'll have a unique subdomain and a unique server. Your application will spin up a server with those certificates and the plotly chart creator web application will be able to make requests to your server, no problem!
• For example, this is what the application UI looks like when the certificate has been generated:
  
  In this example, the domain https://plotly--ab29698f-1b35-420c-881d-6.plotly-connector.com has a unique certificate associated to it. And remember, this domain just re-routes to localhost:9495 on the global DNS network. It's still just loading content from your localhost server - data is not traveling over the network!
• These certificates are saved locally on your machine in the ~/.plotly/connector folder. They expire every 90 days and the application should automatically generate a new one for you.

There are a lot of moving parts here, so please create an issue if you run into any issues.

And finally, big thanks to the team at Let's Encrypt for building the infrastructure that makes this type of architecture possible. 🍻

Adds letsencrypt support to communicate over HTTPS with https://plot.ly.

Include credentials in the browser requests and accept credentials headers server-side.
Update README.