Looker 3.44 Release Notes

release

(Margaret Rosas) #1

Release Rollout Begins: April 7, 2016
Release Final & Download Available: April 21, 2016


Important Changes

  • Loading assets from CDN is now enabled for all Looker-hosted customers.
  • Several Legacy Features are going to be turned off by default, but can be re-enabled. Learn More
    • Field Labels Can Override View Label
    • Leading Period in Field Names
    • Upper Case Labels in Downloads and Chart Configurations
    • Send email from looker@looker.com
    • Default Redshift PDT Distribution Style EVEN

Dashboards

  • Automagic Heatmaps: The “Fit to Data” feature is supported when using heatmaps.
  • Improved Error Messaging on Dashboard Tiles. Incorrect filter syntax errors are now displayed.
  • Fixed an issue where tiles were loading in a compressed way before properly filling the screen.

Visualizations

  • Custom X-Axis Limits. Introduces a setting to show or hide the first or last X rows as specified by the user.
  • Logarithmic Y-Axis is now available for charts
  • Fixed an issue with misaligned stacking pieces in stacked bar charts.

Scheduling

  • New Scheduling Admin Panels
    • Scheduled Plans. Displays all currently scheduled items in your Looker instance with relevant details, including task owners.
    • Scheduled Job History. This section shows recently run items on your instance with details, including the query, runtime, and success/failure state with errors, if applicable.
  • Looks can now have more than one schedule. Bringing Look scheduling into alignment with how Dashboard scheduling works. Currently scheduled Looks will be assigned to the first person who created that schedule and other users will not be able to see the schedule, even if they are recipients of it.
  • New schedules will start at the first scheduled time. They will no longer send an email immediately after scheduling.
  • Schedules that belong to a user who no longer exists in the system (they’ve been deleted) or have been disabled are automatically reassigned to the first enabled user with the “Admin” role.
  • Fixed issue with dashboard display on the iPhone mail app.

LookML & the Developer Environment

  • Custom Field Groups. Adds the ability to group fields together in the field picker using the new group_label: parameter. Learn More.
  • Ability to Rename Projects. Accessible in the “Project Settings” option located under the Sync button.
  • Updated Distance Field
  • New In-App Links to Documentation applies to:
    • LookML validation errors, where applicable.
    • Applies to deprecated features and the Legacy Feature Panel to help transition off of old features.
  • Multiple conditions in HAVING clauses and filtered measures now are always each wrapped in parentheses.

Dialects

  • Spark/Hive Updates:
    • Fixed an issue where EXPLAIN was broken for Hadoop-related dialect issues.
    • Fixed an issue where SQL Runner queries were running slowly on Spark SQL.
    • Tests for PDTs are now run at connection time for Spark SQL.
    • Spark SQL now gets schema information from SQL instead of JDBC calls.

General Tweaks and Bug Fixes

  • Turned on load from CDN for all Looker-hosted customers.
  • Update to the create_table_calculation Permission: Users now do not need the create_table_calculation permission to see table calculations. This permission is still required to create calculations.
  • Fixed an issue where switching into dev mode while looking at a file that only exists in production caused an immediate 404.
  • Made updates to the Explain function in the SQL Runner to apply more dialect-specific functionality and remove forced limits.
  • Fixed a bug where table calculations didn’t work when exploring queries from the SQL Runner or i__looker.
  • Fixed a bug in which one failed SAML validation would cause all subsequent checks to fail, even with a valid file.
  • Fixed an issue where a new version of Chrome was causing dashboards to crash the browser.
  • Fixed a bug where dashboard changes were not being reflected due to caching.
  • Fixed a bug where filtering on nulls as strings did not work.
  • Fixed a bug that made navigating LookML files in developer mode very slow.
  • Override a chat client setting that caused the chat window to open unnecessarily.

Interactive Map Visualizations (3.34+)
(Ken Yeoh) #2

With this feature, can we extend a date axis past what’s available in the data?
For example, if we want to show month-to-date progress for sales, and we’re in the 5th of the month, we would like to extend the X axis to the end of the month without creating derived tables. Currently, the visualization would end at the 5th.


(Margaret Rosas) #3

#4

Can you explain (or link to an explanation), please?


(Morgan Imel) #5

Hi Ken!

The Custom X-Axis Limits feature only lets you limit the rows displayed, thus limiting the range of the X-Axis.

Extending the X-Axis of a visualization can be accomplished by zero filling the results. In your use case, you can zero-fill dates, as described here.

Cheers!


(Morgan Imel) #6

Hi Michael!

We’re in the process of writing some documentation about the distance field. Here’s a quick summary:

The distance field will take two parameters of type location. It will look something like this:

  - dimension: distance_to_santa_cruz
    type: distance
    start_location_field: start_location
    end_location_field: end_location
    units: miles

where start_location and end_location are dimensions of type location.

We hope to have documentation about this very soon!


Looker and Vertica: Geospatial Analytics
(Brett Sauve) #8

@Michael Docs about type: distance dimensions are here now.


#9

Better flexibility for the X-axis limits would be great. Even just being able to truncate rows at the beginning AND end of a chart (rather than being forced to choose one) would be helpful.


(Max Corbin) #10

Hey Ross, I can communicate that feedback to Product for you, seems like a pretty interesting idea. For the time being, a possible workaround would be to use table calculations to limit the rows that are used in the visualization. For example, if you wanted to limit the number of leading and ending rows included in the visualization, you could do something like this:

if(row() > 10 AND row() < (max(row()) - 10))

This would produce a yesno table calculation that would have a value of Yes on most rows, but would return No on the first 10 and last 10 rows. Then, you can click on the table calculation header and Hide on No in the visualization, effectively limiting the first ten and last ten rows.

Let me know if this works for your use case, and in the meantime I’ll take this idea to the Product team!


#11

Oh, good idea. Thanks for the suggestion, Max.


(Nick Morrison) #12

Sorry to dig up an old post, but I would like to request this feature, too. I have a chart that just needs to be trimmed slightly on both the first and last few rows, so being able to hide a few rows using both of the “first” and “last” selections would be useful.