Fall 2020 Virtual Sprint
- Lawrence Newcombe
- Aaron Crosman
- Jung Mun
- Paul Prescod
- Andrew Beyer
- Marciana Davis
- Debbie Duran
- Michael Beaty
- Andrew Beyer
- Joe Blodgett
- Michael Kolodner
Work during this Sprint was split into two major work streams, one to create documentation and the other to progress on a user experience and interface for generating data. As part of the documentation effort a subteam reviewed the survey results and offered suggestions for how to improve the personas.
A basic frame work was created for the documentation was developed by the documentation team. We explored how to use CCI to move existing data to another org:
- CCI extract dataset
- Turns into SQL file
- SQL can be edited in another platform
- CCI pushes data into destination
Michael K. and Samantha used that outline and existing documentation to created an initial complete draft of a process to move data from one realistic org to another (as opposed to simple clean orgs like Trailheads or Scratch orgs) using CCI. That draft is now being circulated for review and feedback. There is an expectation that this work may uncover issues in CCI that will need to be addressed, but the CCI team needs people doing this kind of work to be able to address those challenges.
The feedback from the survey was reviewed. That team found some places we need to adjust the personas and noted those suggestions in that file. The team will need to look at those again in the coming months to create a final version and then add a copy to this repository. We also now have an inventory document of data related tools that includes already known tools and tools listed in the survey. Work still needs to happen to clean up that document and make it fully public (currently in a shared Google Doc).
There was lots of discussion of the difference between having the UI off platform, on platform, or a mix of both. We used the diagrams below to help guide the discussion. Aaron Demoed a tool he called Snowmakery that's a proof-of-concept level Snowfakery recipe editor. The group also discussed a couple on platform generation tools:
- Data Generation App which has interesting features, and run entirely on platform and is free in the app store. We however found a website where the creator appeared to be trying to charge $99. The app also have a 10,000 row limit. *Forceea an Open Source app that requires significant work by an APEX developer to use, but has tremendous capacity.
A few rough drafts of structures were shared for discussion:
That discussion lead Paul and Lawrence to start proof-of-concept tools of their own.
- Paul demo’ed an iteration of Aaron’s recipe editor based on JSON Schema:
- Lawrence demo’ed a his iteration of importing JSON Schema into Salesforce. Lawrence demo raised a potential scenario of populating data via APEX
Samantha commented that these reminded her of the FormAssembly SFDC connector in terms of user flow.
Lawrence also prepared a series of diagrams that were more polished that the initial set. Of particular interest to the team were concepts four and five.
Four places the UI for the process outside of Salesforce:
.
Five places the UI on platform but still leverages remote tools for on platform generation.
As of this writing Lawrence's full concept drawings are here on Lucid Chart.
Jung shared Apex DML Insert Stress Test Demo (Requires PluralSight login).
Potential Building Blocks for the UI include
- JavaScript
- JSON + YAML
- CumulusCI
- Snowfakery
- CSS + HTML
- Python