Compatibility/Meetings/2021-07-06

• Web Compatibility Meeting Meeting - 2021-07-06
• Minutes: Previous 2021-06-22

Minutes

Scribed by Dennis

Create new project for OKR tasks (SV))

• Oana: Karl can you create a new project for Q3? We'd like to add our tasks there.
• [TODO] Karl to create a new boar before d/project.

Webcompat and Performance Teams (karl)

(for next meeting when Thomas and Dennis are back) Read https://docs.google.com/document/d/1RfV_gV0LgOeADjGM6ZcY6tQeLqwIUHzFfl2MXubrvlY/edit#

• Karl: Patricia Lawless is managing the performance team. I had a meeting with her, and she wants to have better interactions with the WebCompat team. I will send her a list of performance issues we received in the past. She asked us if we can identify the type of performance issues in reports we receive, and she wants to teach us how to better use the profiler tooling, and she suggested to continue open issues on Bugzilla and to use the `[qf]` whiteboard tag.
• Karl: There were also a couple of Joy of Profiling sessions, where you can bring your own profile and people will look at it together. There will be more sessions.
• Dennis: Interested in those sessions, and great to see the perf team is still interested in our reports.
• Tom: Is this about us taking over more issues, or us throwing issues to them?
• Karl: It's more about getting the help of the performance team if we run into an issue, and about filing the right bugs for them to look at it.
• Oana: We'd also be interested in taking part here, it would help us identifying the performance issues.
• Karl: Yes!
• James: Do we know which fraction of WebCompat bugs are performance issues?
• Karl: No, we don't track that. We maybe could search for keywords in the comments.
• Oana: Usually, when a perf issue is filed, we record a profile, so you could search for the links.

Compat interventions testing (jgraham)

(needs thomas) I have a mach command that works and nearly have it running on CI. Need to decide on details.

• tom: we can set up a meeting or two to hammer out details, feel free to book one asap if you'd like.
• James: The current implementation is essentially pytest wired to a mach command that allows you to run tests. Tests have some nice additions, like testing with or without interventions enabled, ... it fills a lot of the requirements we discussed. There are some problems, for example it's currently using the WebDriver harness for WPT, which isn't feature-complete. We could look into adding Selenium, but that might be complicated, but maybe we can could get away with workarounds given those tests will be Tier 3 anyway.
• James: This implementation is there to check if it meets the requirements we actually have, and that it works the way we want to.
• Tom: I'll have a couple of chats with James to figure out details, and will write some additonal tests. Using WebDriver instead of Marionette has some advantages, even if the setup might be a bit more complicated.
• James: Marionette is what we use as the backend for WebDriver, but you're no longer supposed to use it directly.
• Karl: Where's the code located?
• James: mozilla-central, it's pushed to try, https://treeherder.mozilla.org/jobs?repo=try&selectedTaskRun=AoYR3ahtQsKAHascs5THwg.0&tier=1%2C2%2C3&revision=b9185fe0658521dbdbe72f17a8e5910932100e4e
• Karl: Is there documentation on that, or where will it be?
• Tom: I was thinking about adding it to the wiki, including running and authoring tests. Placeholder for Site Interventions Documentation https://wiki.mozilla.org/Compatibility/Projects#Site_Interventions_Automated_Test_Suite https://wiki.mozilla.org/Compatibility/Interventions_Test_Suite

Shower thought: consider having a team-wide public "brag document" (or similar) (tom)

In the interest of keeping everyone's career development and self-esteem in mind, we might consider maintaining a team-wide document where we each list our accomplishments as we achieve them (big and small). If that wouldn't be too uncomfortable for anyone, it would let us present a micro and macro level view of our work as we do it, and keep us motivated to ensure we're all happy with the work we're doing, as well as the trajectory our individual careers are heading in.

• Related: https://jvns.ca/blog/brag-documents/
• James: There's value in having a documentation of what you've done that's all in one place. I've seen other teams use status bots in various forms.
• Tom: The idea is to have a bit like the two minutes, but permenant and with more work and less garden. To keep track of what we do, so it's easier to refer to the list if someone asks and show what we do, and to keep track of things that would otherwise get lost in the noise. Having a permanent work-log is also great for everyone individually, so we're able to look back and see what we achieved.
• James: I'd never keep something updated if it's just me, some social pressure might help. I wonder if we're talking about a "Brag Document", or a "Status Document" here. Brag Documents are more useful for career development.
• Ksenia: I think putting something out is a good idea!
• Karl: I think it also enables the rest of Mozilla to see what the team is doing, which might be helpful.
• James: If you communicate stuff out to the rest of the company, it probably needs a lot more context. Something internal gives you the ability to go back, find what we did, and write something with more context.
• James: What is the process we get into to avoid doing this for two weeks and then stopping?
• Karl: Is there someone who wants to lead this? I think someone should make sure that people actually fill it out.
• Tom: (kinda volunteered) I'll start a document.
• Dennis: I think we should have this private, and then take things out and make it public with more context if we want to.
• James: I wonder if it's worth spending some time investigating if we can add some automation here, like a Slack/Matrix bot to ping people and make collecting that data easier. It's probably better to log work as it happens, not two weeks after the fact because it's easy to forget things.
• [TODO] Tom to Tom to start a team-internal document, Tom to ping people if they haven't added stuff in a while, and someone maybe to look into possible ways to automate it before 2021-07-20

Two Minutes ( 👹 )

The useful, the silly, the things you did since the last meeting. Pro and Perso accepted.

• Dennis: Vacationing! Nothing productive at all, wohoo! Played some Super Mario World instead of Coding. [Some incredibly bad (thunder)storms](https://twitter.com/LucaSzwirblies/status/1409578240801390594).
• Guillaume:
• James: Some WebDriver-BiDi spec work. Apparently if you want an insulating front door the trick is to live in Germany or Scandinavia. Lots of bees.
• Karl: A lot of reviews stuff for the end of semester. Some work on UA freezing. Diagnosis. PTO hunting apartments in Tokyo. Monstera.
• Kate:
• Ksenia: Some diagnosis, biking and hiking during the wellness week
• Oana: Triage and moderation. Finished last UA Override/Interventions verification. We are happy with the ML-bot. Lots of rain, but I have cucumbers, letuce, green onion and little tomatoes.
• Raul: Triage and moderation. Finished last UA Override/Interventions verification.
• Thomas: PTO was gloriously relaxing. a week at a quiet cottage by a lake, and a week doing little or nothing. lots of rain lately, but in that "it's also mordibly hot outside" way which makes it frustrating.
• Joe:

PTO / Travel ( 👹 )

(no more travel until the end of 2021… probably) Coming holidays across the world https://www.timeanddate.com/holidays/ 5th of July (Raul PTO) August 12 (Ksenia, PTO)

FYI: 2021H1 Ideas (karl)

Let's keep this around so we have a reference. https://docs.google.com/document/d/1w3_AU4ZYJJLSkqOW5j_6K49JxmutEsxcpsiwHXz8Tbw/edit?usp=sharing