30 Oct Omniscope ROCK build – 2019.3
We’re pleased to announce the latest quarterly “rock” build of Omniscope, version 2019.3 b20709. A rock build marks the end of a development phase (2019.3) with a build that has passed extended manual and automated testing and is suitable for production deployments.
Notable new features:
Summary charts in Tables
You can switch between none, header charts, and overlay charts, using the top-left chart icon. In reports, these are hidden by default. In the workflow block dialog’s “Data” tab (shown above), these are shown by default.
Misc. improvements to the Workflow
Key shortcuts Ctrl+Z, Ctrl+Shift+Z have been added for undo/redo, respectively.
The toolbar is much cleaner and simpler; you can now duplicate and delete selected blocks directly from the toolbar:
The “Batch append folder” and “Append files” now support JSON and XML files, and the “File” block now has automatic configuration for importing data from XML files. The “Elasticsearch” block now supports Public Key Authentication (PKI).
Misc. improvements to Reports and views/visualisations
Report tabs now support a Tab background/overlay image, allowing eye-catching dashboards:
You can also use the Image view to place images, optionally linked, easily within your dashboard.
The Bar view now lets you configure a “dual axis” view with 2 parallel/interleaved measures on different left/right scales:
The Bar, Line, Area and related views now let you choose to hide null values, have better automatic histograms for integer fields, and show a “continuous” axis for number/date fields, supporting cleaner split-axis labels:
Stacked bars have much better default/automatic stack labels; you can also use the “Reverse” option in the Bar view to invert Y axis, useful for cases where you want to show an implicitly negative measure:
The Content view lets you insert “micro-charts”, a set of views optimised for a small size, within the text of the Content view. You can, for example, create your own table layouts combining formulas and “sparklines” or mini pie charts:
The Table, Chart and Details views have a much improved field picker, letting you easily select and deselect fields individually or all at once.
The Scatter view has a new “Trend line” option, for unclustered points.
The Sankey view for relational data has been much improved, though remains experimental until we add support for other data formats/models.
The Pivot view lets you select (sub-)total cells, and exports data in the same column/row/cell structure as displayed, in CSV format.
Filters now allow you to show “Choice” filter types as a drop-down, to save space, and to configure “sort by frequency”; you can also configure a “single choice” option which shows a ‘radio’ single-select list of values.
Parameters are a hugely powerful and versatile addition to support centralised administration and automated publishing. A parameter is a reusable named value that can be defined in a project, or server-wide in Admin:
(Project parameters in workflow)
(Server-wide parameters in Admin)
(Using a parameter)
Your workflow can refer to these parameters in place of literal values for most block options. For example, you might want to parameterise the Database block’s “host” setting, allowing you to centrally reconfigure all database blocks for a replacement database server.
You can also use the Scheduler tasks “Edit project parameters” and “Edit app-wide parameters” to automate changing of parameter values, e.g. to execute a workflow with a different client’s data:
Omniscope Evo now supports translation into arbitrary languages, with near-complete translations bundled with Omniscope for French (thanks to Magali Colin of Avizua) and Italian (shown below), and the ability to add your own custom translations for other languages.
The server default language is configured in Admin; users can override this in their browser using the language ‘flag’ icon in the app. New projects capture the language settings of the user at the time they are created; workflow execution respects the project’s language where blocks produce warnings or errors, for example; reports respect the user’s language if different to the project’s.
The welcome page now has a language option (flag icon) and navigation sidebar:
The Scheduler “Reconfigure block” task can now automate edits to most kinds of workflow blocks e.g. File, Append, etc.
For a more detailed list, see the changelog.
Development of the next version (2019.4) is well under way with the first builds already available. Download the latest daily build here. Some features to look forward to, many of which will appear in 2019.4:
Custom Script block
This new block, which replaces the existing Python and R blocks, with better performance, and allows you to create reusable blocks with your own set of options, which can be shared with your users via import/export, shared repositories, and Github. It will let you work productively with your own IDE as you develop the block, or use Omniscope as a “slave” for data sources and data visualisation, when doing (e.g.) Python-centric data science. The 2019.3 rock, above, includes an early, experimental and incomplete version of this.
A new Map view
Supporting multiple base maps and data-driven layers from different data sources, including Choropleth, area fills, polygons/polylines, point-to-point/spider plots, markers, raster tiles, images, vector tiles with configurable sub-layers, micro charts/pie overlays, geojson, shapefiles, high-performance / larger data volumes…
Undo/redo in Reports
Common mistakes such as deleting a view will be reversible in a click, tap or keystroke.
Data-driven configurations to dynamically create user-specific tailored experiences of a report, with different row/column subsets of the data, different tabs, different branding, …
Better Report configuration user experience (UX)
Using the Report to explore your data or build dashboards will become more fluid and intuitive.
View options will be better presented, and easier to navigate and understand.
Better support for creating, administering and understanding your report’s saved/named queries.
More intelligent data-driven defaults for reports, views and filters.
Better automatic data colour palettes.
Device-specific dashboard previews.
Parallel and push-based execution
A more powerful workflow execution engine, making better use of parallelism and allowing editing workflows while they are executing. Push-based execution, and streaming continuous event-based data.
And much more…
Search & sort files on the welcome page.
Formula measures in Bar, Line etc. views.
More ways to copy/paste various settings within and between projects.
Project and Report commenting and history.
Bookmarks, for saving and reusing blocks and groups of blocks in the workflow.
As usual, please let us know your feedback at firstname.lastname@example.org. Enjoy!