Wikimedia Apps/Team/iOS/Retrospectives/06-15-2017

Discussion Topics

Bug writing criteria (Try to repro, regress, write repro steps)
Regression tests… how “seriously” can we take them? We need to get 9.x devices!
- Very large doc now. Loads very slowly. Can I delete or archive some of the older results sheets?

Previous Action Items

Team to incorporate reviewing Needs Acceptance during Task Sync
- We have to decide on a case by case basis. If people feel like they weren’t involved enough in reviewing the acceptance criteria, bring it up
Natalia to follow up with Monte re Phab search fix DONE

Good Things That Happened Since the Last Retro

Places

Discussion

Bug writing criteria (Try to repro, regress, write repro steps)

Some bugs get moved to Need Testing Criteria even though they have repro steps in the comments

We have a “bug” link, but probably doesn’t get used (Bug template)

There’s a “Regression” tag that we stopped using (we should start using it again)

Question: What's better - regression tag or putting “Regression” in the title?

- Tag

Some bugs have titles only - that’s not enough, makes it difficult to sing off

Better to file incomplete bug tickets than not to file them at all (from the PM perspective)

Question: What is a certain level of bug that you’d file a ticket for?

- Severity - if it’s severe but you can’t repro it, put it on the board

- Reproducibility - if you can repro but it’s not severe

Severity and reproducibility help you figure out if it should be on the board

Question: How do you deal with bugs?

JW: Normally I file bugs and work on them (if it’s something that should go in the release)

JM: It should depend if the bug is related to the feature that you’re working on. If it’s unrelated, I’d prefer you to file it and take it through Needs Triage etc.

JW: To add to that - if it’s a quick fix, how would you find if there’s room for that (less than 30 minutes)?

JM: Use your judgement (you’re just super fast Joe ⚡️)

MH: For me it’s a 10/20 minute threshold

JW: Anybody interested in starting filing tickets for EVERYTHING?

JM: That’d be my preference.

JM: We should keep track of how many bugs go into the release with incomplete repro steps. We can check how many incomplete bugs were closed during our Process Process meeting.

JW: Other ACTION is to update the template to include regression and reproducibility

Regression tests… how “seriously” can we take them? We need to get 9.x devices!

JM: Sometimes we don’t take device-specific bugs seriously. Having more 9.x devices would give us the option to verify them. Mostly on me to take those failures seriously

Action Items

Natalia to add reproducibility & regression & frequency to the bug template
Natalia to check out the complexity of adding custom form fields
Anthony to increase the priority of getting 9.x devices