Skip to content

LG-9008: Remove code that supports pre-launch IPP functionality#11440

Merged
eileen-nava merged 4 commits intomainfrom
em/9008-remove-prelaunch-code
Nov 6, 2024
Merged

LG-9008: Remove code that supports pre-launch IPP functionality#11440
eileen-nava merged 4 commits intomainfrom
em/9008-remove-prelaunch-code

Conversation

@eileen-nava
Copy link
Contributor

🎫 Ticket

LG-9008: Remove code that supports pre-launch IPP functionality

🛠 Summary of changes

  • Removed the isPilot field from usps location objects, because it is no longer relevant

📜 Testing Plan

  • Run automated tests
  • Navigate to the PO Search
  • Confirm you are still able to search for a location

@eileen-nava eileen-nava requested review from a team and WilliamBirdsall October 31, 2024 21:27
{t('in_person_proofing.body.location.po_search.results_description', {
address,
count: locations?.length,
})}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we need to phase this for 50/50 state?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This PR doesn't include a database change, a change to anything stored in redis, a change to a job, a change to a route, or a change to a feature flag.

Is there another 50/50 state scenario that I am missing? Or, let me know if you do think it falls under one of the scenarios covered in 50/50 state documentation. Since our team has hit multiple 50/50 state errors recently, I definitely agree with being cautious about it.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's not listed on that handbook article, but another scenario is one where we have JavaScript that requests from an API endpoint and assumes the response will take a particular shape. If a user is running an older version of the JavaScript, it may fall out of sync with the newer instances handling the API responses.

I think in this case it might be okay since it'll treat the absence of isPilot in the response as the same as if it had previously returned false, and the behavior would be the same?

Probably worth documenting this as an additional 50/50 scenario.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Feel free to resolve this. I feel comfortable moving forward. (I didn't want to resolve it if you wanted to keep it as reminder to update docs.)

Copy link
Contributor Author

@eileen-nava eileen-nava Nov 4, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for noting the additional 50/50 state scenario, @aduth. Could you, or another dev who has witnessed this type of 50-50 state bug, please take a look at this documentation PR? (I would be curious to hear what remediation steps you suggest.) Thanks!

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think in this case it might be okay since it'll treat the absence of isPilot in the response as the same as if it had previously returned false, and the behavior would be the same?

I agree with the above.

I created a branch that still uses isPilot in in-person-locations.tsx, but does not have that property present in the response from the backend. This was meant to reproduce the 50/50 state scenario discussed in this thread. I included a video of manually testing that here. I didn't hit any bugs in that manual test.

That, along with what's discussed above, makes me think we'll be okay when this deploys. (Let me know if anyone disagrees or thinks I missed something. If I don't hear any objections, I'm going to move forward.)

zip_code_4: string;
zip_code_5: string;
is_pilot: boolean;
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

One other thought I had- do you think we should version up for address-search (app/javascript/packages/address-search/package.json) because we are changing the interface? If you agree- we should also update app/javascript/packages/address-search/CHANGELOG.md

Copy link
Contributor Author

@eileen-nava eileen-nava Nov 4, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I agree, that's a good idea. 👍🏻 I will get to that tomorrow. Thanks for raising the topic.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@gina-yamada I added commit 4389c9a to version up the address search package. Unfortunately, I don't have access to the 18F npm organization. 😬

I updated the documentation to reflect the new process for requesting access to the 18F npm organization. I also requested access. I think I'm blocked on publishing this new minor version of @18f/identity-address-search until I have npm org access.

Copy link
Contributor

@gina-yamada gina-yamada Nov 5, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If getting access will take a bit of time, add some time on my calendar. In that session, you can merge this in and I can publish it for you. Ideally, we should do it back to back so we know there is no drift. (I want to make sure I have access to publish before you merge it in.)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, Gina! I booked time on your calendar for tomorrow.

Copy link
Contributor

@WilliamBirdsall WilliamBirdsall left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, but needs the changelog message to pass automated checks :)

Copy link
Contributor

@KeithNava KeithNava left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

FYI, I'll wait until this PR is merged before I merge my PR that touches literally the same files #11424

@eileen-nava eileen-nava merged commit 8f41ad7 into main Nov 6, 2024
@eileen-nava eileen-nava deleted the em/9008-remove-prelaunch-code branch November 6, 2024 21:45
@eileen-nava
Copy link
Contributor Author

Gina and I paired to publish the new npm package. 🎉 (cc: @KeithNava for visibility)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants