Page Comparison

Sprint Architecture

Backlog Item Breakdown

...

Ad-hoc testing

...

Required Backlog Items

Release Process

We will ship once per sprint.

Daily Breakdown

Length: 10 business days

Dev - Backlog item checklist

...

Team Member

...

Task

...

Dev #1

...

• Complete backlog item or bug development

...

Dev # 2

...

• Write test case

• Create automated test

• Manual test of feature, and surrounding features

...

One-off Sprints

• Bug fix only sprint

Sprint Composition

• Can we still use Ewa 4 days of the sprint - ad-hoc testing would really help - this would be very valuable. She could do ad hoc testing, writing test cases, etc. anything that is valuable. Even if she does a timeboxed one day ad hoc testing, that would be valuable - one day timebox on QA and one day timebox in ACC.

• Need to communicate to the user that the cost might not be worth it to do non-valuable things.

• Can use playwright for automated testing - have we solved the issue of authentication?

• Change management - could write the ticket every sprint

• If we follow this method, might not have to do a full regression test. But might want to do a 5 day regression.

• Need - conversation about who will decide what is a bug or not or UI change we are not going to make. Action item: feature we will never fix - e.g. Here is a bug that is not a bug. Team can search to see what decisions were made.

50-70% - New development and testing, including:

• New feature (meets definition of done)

• Bug fixes (old or new)

• Test in test environment and ACC

20% - Might include tech debt, ad hoc testing, documentation, translation testing, release activities, missing unit tests and automated tests, critical bug fixes, release activities

10% - Release activities / release overhead. Release notes / change log - Steve’s notes, documentation, make sure Playwright tests are passing, ad hoc testing - Ewa ACC and test.

Without QA resources

Note: Days are not set in stone

Dev work for each sprint will be pushed into a sprint branch (e.g. sprint 29 branch). Tested in the dev environment, then to QA, ACC and prod.

Items we are working on in sprint 27, 28, 29 will go into a sprint branch. We can create a sprint branch in sprint 29 to support - {need to think about it}.

Release process with shippable feature

We can ship in mid-sprint.

Explore - do we want to start shipping each story? We can just ship the story when it’s ready vs a specific release day. Ashiq will look into it. Meta does this.

Day 1 - Discuss what we can ship in the morning (we can tack it on the scrum), and ship in the afternoon. PO can inform clients of shipment.

Day 2, 3 and 4:

• Dev - Any backlog items that need to be tested from days 9 and 10 of last sprint.

• Dev cannot test backlog items they developed.

• Dev, PO, Ewa, Maggie - Ad-hoc testing of app (30min to 1 hr.)

• Dev - meeting with dev and QA sprint tickets and identify test cases as a group - discuss possible test cases for each ticket, which tickets interact. Test cases stored in DevOps.

• {PO? - UAT prep - Word doc and spreadsheet creation.}

• {PO -

  • {Option 1: Provide list of backlog items to users. They can test if they have time in ACC. If they do not have time, they do not have to test. One hour of testing only max. Perhaps we can send a copy of the weekly agenda, and they can test if they like.}

  • {Option 2: PO hosts regular bi-weekly testing session with users to test backlog items completed last sprint in ACC. Users only attend if they can. }

...

Day 10 - Sprint review, sprint planning. Finish and wrap up what we can.

Note: we shouldn’t worry too much about carry-over at this point.

Day 1, 2 and 3:

• Dev - Any backlog items that need to be tested from days 9 and 10 of last sprint. Dev cannot test backlog items they developed.

• Dev, PO, Maggie - Ad-hoc testing of app (30min to 1 hr.)?

• Dev- Review sprint tickets and identify test cases as a group? (Question: where should test cases be stored?)

• {PO? - UAT prep - Word doc and spreadsheet creation.}

• {PO -

  • {Option 1: Provide list of backlog items to users. They can test if they have time in ACC. If they do not have time, they do not have to test. One hour of testing only max. Perhaps we can send a copy of the weekly agenda, and they can test if they like.}

  • {Option 2: PO hosts regular bi-weekly testing session with users to test backlog items completed last sprint in ACC. Users only attend if they can. }

Day 4, 5, 6,

• Dev work

• Dev testing

• Dev user story writing

Day 8:

• PO walkover (if needed) - 15-30 minutes max.

Day 9:

• Dev: Sprint review planning - add items to present at sprint review to agenda

Day 10:

• Dev: Sprint review dry run - prepare demo and participate in sprint review dry run

• Dev: Sprint review - participate in sprint review

• PO / Users: Sprint review - provide users with {test} URL, and they can test if they are interested.

Definition of Done

• Acceptance criteria is complete

• Unit test is complete

• {Automated testing is complete}

• Dev has tested in QA, ACC and prod (if possible)

• Dev sends screenshot of new UI work, or other items for which feedback is needed, to PO during development process for sign-off. Do not need to send screenshot for bug fixes.

• No other sign-off is necessary as PO has seen screenshots, walk-through and review

• When a story is done, and anything around it is tested so it is shippable

Future consideration:

• Security

• Code coverage

Testing

Ad-hoc testing:

• Dev, QA (Ewa), PO - 30min to 1hr per sprint

Basic level: - Dev’s only

• Writing test cases

• Reviewing each other test cases - manually

• Executing each other test cases - manually

• Still valuable - can start automating in the future

• Careful not to add too many testing activities to the sprint for dev’s

Automated testing:

...

B Unit testing - more like unit testing for the UI - not full UI automated testing

Regression Testing

Release Plan

Release Plan - brainstorming

1 month release schedule - Release Train

Total: 2 sprints

Sprint 1

Tech debt - some tech debt might need to be tested. But it is safe to do tech debt at the end of the sprint. At minimum, as sanity check might need to happen.

Dev:

60% - New development

10% - Bug Fixes

20% - Tech debt

10% - Release activities

QA:

70% - Backlog item testing

20% - User testing - day 10? Either PO or QA?

10% - UAT prep

Sprint 2

Sprint 3

Dev:

QA:

50% - ACC environment testing

50% - UAT testing (with users) - Test environment?

Sprint 4

Dev:

50% - Release activities

?? % - Critical bug fixes?

?? - ??

Dev - Release Architecture Needed

Manual to automated - Jugraj is working with Henry on this

One click deployment

Patch fixes - how do we implement?

Anything else?

Architecture: With QA resource - brainstorming only

Day 1, 2 and 3:

...

QA - Any backlog items that need to be tested from days 9 and 10 of last sprint.

...

QA - Ad-hoc testing of app (30min to 1 hr.).

...

QA - UAT prep - Word doc and spreadsheet creation.

...

Dev - Work on top priority backlog items

{PO -

...

One-off Sprints

• Bug fix only sprint

• Full regression test

Backlog item checklist

Shippable Stories

UAT

...

• Product Owner or QA:

  • Provide list of backlog items to users

  • Users can test in ACC if they have time

  • If users do not have time, they do not have to test .}

  • {Option 2: PO hosts regular bi-weekly testing session with users to test backlog items completed last sprint. Users only attend if they can.}

Day 4, 5, 6, 7:

Testing - if QA resources - brainstorming only

If QA:

• First three days, can catch up on testing

• Writing test cases

• Ad-hoc testing

• Review test cases with one of dev team’s members

• Executing test cases

• Test all tickets within one sprint - if possible

Testing - brainstorming only

Discussion with Grace - April 4

Bugs addressed during the sprint. Leave some space for bug testing during the sprint

10 business days

First 3 days:

• Ad-hoc testing

• UAT prep - documentation regarding the previous sprint

• Users can test what was completed last sprint - Grace or Steve, do not need two team members

• TGD team - QA team tests first, and then PO or user tests, and the user or PO marks it as done. User or PO testing does not need to be complete during the same sprint.

Issues to think about:

• How to track modifications to backlog items. E.g. when testing, if something needs to be added, what should we do?

Discussion with Ashiq

PO and users:

• test in ACC

• They can test on the first day of the next sprint - will will give them a list of what we completed and they can test if they can, if not, that is ok - maybe 1 hr of testing for them

• Dev - send a screenshot of new UI before the functionality is built to get feedback from PO via email. Do not need to send screenshots for things such as bugs, and FR translations. Only things we need feeback on.

• PO walkover - two days before the end of sprint - 15-30 min to take a quick look if screenshots are not enough - flexible. PO may not need it every week.

  • Can show him work still in progress

• Users can test in test environment if they wish after sprint review. We can give them a few bullet points of what we worked on. Either at end of review or first day of sprint.

Test Cases - brainstorming only

Questions:

• What will the test cases be used for? (e.g. backlog testing, automated testing, future use)

• How can we minimize the amount of work required to write a test case? E.g. focus on developing a specific feature per sprint, so test case writing can start on day 1 - ? Perhaps identify which backlog items will need test cases at the beginning of the sprint, so test case writing can start at the beginning?

How would we work with dev’s only to write test cases?

Discussion with Ashiq

QA free from manual testing

• Dev’s test each other’s work / test cases

• Ad-hoc testing - Grace, Maggie, dev’s, Ewa, PO - anyone who can help

• Automated testing - B Unit testing - more like unit testing for the UI - not full UI automated testing

• Concern - not to add to many testing activities to dev team

Basic level: - Dev’s only

• Writing test cases

• Reviewing each other test cases - manually

• Executing each other test cases - manually

• Still valuable - can start automating in the future

If QA:

• First three days, can catch up on testing

• Writing test cases

• Ad-hoc testing

• Review test cases with one of dev team’s members

• Executing test cases

• Test all tickets within one sprint - if possible

Currently we have 3 dev’s and one QA

Need to keep quality or speed

Other work environments - 1 to 1 dev team - work at writing the user stories from the beginning of sprint

Scope

Time

Quality

Automated Testing

• How can we implement?

• How can we keep QA involved?

• What is the outcome of implementation? E.g. less QA manual testing

Unit Testing

{Add short write-up, description}

Release Planning Architecture

2 month release schedule - Release Train

Total: 4 sprints

Sprint 1

Tech debt - some tech debt might need to be tested. But it is safe to do tech debt at the end of the sprint. At minimum, as sanity check might need to happen.

Dev:

60% - New development

10% - Bug Fixes

20% - Tech debt

10% - Release activities

QA:

70% - Backlog item testing

20% - User testing - day 10? Either PO or QA?

10% - UAT prep

Sprint 2

Sprint 3

Dev:

QA:

50% - ACC environment testing

50% - UAT testing (with users) - Test environment?

Sprint 4

Dev:

50% - Release activities

?? % - Critical bug fixes?

?? - ??

Dev - Release Architecture Needed

Manual to automated - Jugraj is working with Henry on this

One click deployment

Patch fixes - how do we implement?

• • (not required)

  • Users should only test one hour max

Sprint Testing

• UAT testing sheet

• Smoke testing

Sprint Branches

Sprint Environments

*When work in QA is ready for next release, they will be pushed to ACC

Deployment

Current: Manual deployment

Future: One-click deployment (Jugraj is working with Henry)

Definition of Done

• Acceptance criteria is complete

• PO (or equivalent) has approved screenshot of UI for new work (as needed)

• Unit test is complete

• Test case is written

• Playwright test is written

• Changes don’t introduce any new security defects, code-quality issues, or lower the code-coverage percentage

• QA manual (main functionality and other components it interacts with) test is complete in correct branch

• Documentation for security groups

Future consideration

Definition of done future consideration

• Security

• Code coverage

Testing