Skip to content

Spring 2020 Virtual Sprint Notes

Jump to bottom
Aaron Crosman edited this page Apr 1, 2020 · 1 revision

First Day Notes

  1. Reviewed Philly Sprint materials, including Issues, the wiki page about ux, and Proof-of-concept code.
  2. Overview of CumulusCI, what it is, and what it does.
  3. Aaron mentioned known outstanding working: Clean up docs (particularly the main ReadMe file), issues from the issue queue, But stressed that this group needed to help provide a sense of where to go next and how.
  4. Paul talked about the new Snowfakery project he has been leading to create a tool that does simple, but slightly different, things. In particular is good for medium to large data sets, but does not contain the enumeration capacity of the Data generation task from Philly.
  5. As a group we talked about some of the use cases people are interested in:
    • Creating data sets for demos that contain all safe-to-share information that looks reasonably close to real data.
    • Creating data sets for QA that covers all edge cases.
    • Creating large data sets for testing automation with a fully populated org.

It was pointed out that while very related, this are a bit different use cases. In the first you don't really care of every possible case is covered, but you want data that looks like a typical a record. In the second you need every permutation of options so you can test that all use cases are represented in the dataset. In the third you mainly care about number of records created.

Other considerations/use cases:

  • Data sets that can be quickly and easily loaded into sandboxes.
  • Data set generation should not be just for developers, it should be something admins can achieve.
  • Being able to use fun, interesting, or informative name sets:
    • Happy Potter Characters
    • Civil Rights Leaders
  • Ensuring that your name set, at least for QA and edge case testing, including things from outside your personal cultural default assumptions.

Options for Next steps:

  1. Take a node from an existing project ERD and see if we can generate all the pieces. 1.
    • Probably NPSP (Account, Opportunity, Contact)
  2. What are the requirements for custom objects?
    • EDA has university sample data, but costly to maintain.
    • What is the context for the data?
    • Creating data sets for demos that contain all safe-to-share information that looks reasonably close to real data.
    • Creating data sets for QA that covers all edge cases.
    • Creating large data sets for testing automation with a fully populated org.
    • Creating data specific to a user story.
    • Important to data for the related objects.
    • Data theme (e.g. Civil Rights leaders)
    • Seeding sandbox
  3. Why do we need to generate new data all the time?
    • New orgs means new custom objects and fields.
    • Consultants have many clients all with different needs.
    • Data structures change over time, and new test data sets are needed.
    • A complete data set allows complete testing.
    • Even if we need to generate the data by hand the first time, being able to convert that to Salesforce objects is the time sink. If we could create something not-totally-unlike a sandbox template and get SnowFakery/CCI to generate data to match it.

Demo:

Paul demonstrated SnowFakery and solicits feedback on modification to demo:

He showed adding to the data schema YML files including adding Affiliations (npe5__Affiliation__c) to connect Contacts and Accounts with only 1 being primary per person.

Question: what’s on your roadmap, Paul?

  • Bringing in data from other system
  • Permutations
  • Frontend ux

Sub-projects for Wednesday:

  • Get setup as OpenSource commons project:
    • Aaron
    • Samantha
  • Declarative definition of the YML files
    • Rebecca
    • Tyler
  • Documentation of process to get SnowFakery to actually work in practice
    • Paul could demo, and then group document.
    • Amanda
    • Erin

Other Project goals:

  • Documenting a process for how to load data without SnowFakery

Shared Links:

Virtual Sprint Team:

Person Company/Organization
Michael Beaty
Rebecca Cole Sullivan
Aaron Crosman Attain
Chris
Paul Prescod Salesforce.org
Erin Schroder Attain
Samantha Shain
Amanda Styles
Cassie Supilowski
Tyler Woebkenberg Salesforce.org
Clone this wiki locally