Finish Forms System Core

[raywangoctova]

Problem Statement

Smooth and standardized VFS development isn’t being achieved today because of a complex and inflexible forms system for VFS front end engineers and designers which results in complexity and workarounds in the VFS development process and inconsistent UX for users.

Our mission is to develop a new VA Forms System Core (VAFSC) application that is simple and flexible enough to be used either in whole or in part by application teams building forms on VA.gov. Building upon this Core application, we want to eventually provide designers, developers, and authors a way to quickly build forms that work well and are a model that other agencies and civic tech organizations can look to for their own forms.

The end goal of this effort is to deploy the VA Forms System Core application to allow VFS teams to quickly and easily:

• Access documentation and guidance to get started
• Build a standard form
• Customize forms as needed
• Migrate an existing form into the updated experience

User Impact

Impacts all VFS engineers and VFS teams building using the current Forms System. Inflexibility of the current system slows developers down when implementing Veteran-facing forms.

Where was this problem reported?

Found in research with VFS teams conducted by a Forms Spike team that ended in Fall of 2022.

What do we know about the problem space?

The new Forms System Core provides the following:

• Modular feature usage (no more all or nothing)
• Form components are Design System web components (one component to rule them all)
• No need to use a config file to build forms
• No need to use schemas to describe what a UI looks like
• Root URL is no longer restricted to the root URL of the application itself, theoretically making it easier to build a full form inside or alongside a more conventional web application
• App isolation ready: this is a modular system contained in its own NPM package, allowing for safer releases, better testing, and the potential for teams to choose to stay on a specific version of the VAFSC

What do we not know about the problem space?

We don't know if a new team picking up the Forms System Core work will be able to finish the system on the same estimated timeline as the previous team.

What (if any) research or discovery has been done?

https://vfs.atlassian.net/wiki/spaces/FLT/pages/2069495835/Forms+Library+Team+Early+Research+Findings+-+FLT

What is the acceptance criteria?

Develop the Forms System Core so that it has feature parity with the current Forms library. Remaining work is detailed here: https://vfs.atlassian.net/wiki/spaces/FLT/pages/2275180551/VA+Forms+System+Core+VAFSC+Feature+To+Do+List+-+FLT

How should we measure success?

Success would be enabling VFS Teams to build new forms using the Forms System Core product. Teams are currently blocked by missing functionality, notably the lack of authentication and pre-fill support.

[davidconlon]

Blocked due to Waiting on funding for a team.

[humancompanion-usds]

My current thought on this is that the Veteran-facing Forms team will continue to build patterns in the Forms Library. That tool helps us use patterns and will help us build the Form Builder tool. However, teams building more bespoke, white glove approach, forms/services need more flexibility (or would very much like more flexibility). Forms System Core would be a better tool for them to use than the Forms Library.

[annekerr49]

This item has been in the backlog ever since we cut the SPIKE teams and has never resolved what to do about it. Should stay in backlog until we decide how to address issue.

1. Forms library is very prescriptive and is driven by schemas and this kind of keeps developers limited in what they can do. Helpful for us to keep patterns consistent and is going to be helpful for us to build the form builder tool
2. Forms System - for teams that are doing more bespoke white glove treatment forms that are very complicated they need more flexibility and that is what the Forms System Core gives them. It will be open source. So they are two different tools Form System Core had six to nine months of work to get done