From MozillaWiki
Jump to: navigation, search


Scribed by James

Fixing some C++/Rust Webcompat bugs (karl)

Let's say we would want to prioritize fixing some of the Webcompat bugs ourselves (webcompat team). What would it take? Resources, time. Would that make the work more appealing?

  • Karl: Diagnosis often leads to core bugs. Sometimes these are not too difficult to fix. What would you think about doing more work to help fix the core bugs?
  • Dennis: There is definitely value in people doing diagnosis learning more about gecko internals, and doing these fixes will help. That in turn will make diagnosis easier. Needs to be a balance between doing diagnosis and doing this. But could be better overall in some cases.
  • Tom: I feel the same. Some mentoring would help to estimate the time required.
  • Ksenia: Layout bugs would be time consuming, but mentoring would make it easier/possible. Could be worth trying.
  • Tom: Some people on the team have knowledge. We can probably make it work. But are interventions a faster approach?
  • Karl: Need to set priorities. Still continue doing the other things. Start very basic e.g. fixing a pref in android. Needed to learn how to modify tests even though it's a small fix. Would still get review from the core team, so it's hard to break things, even though there's a learning curve.
  • Dennis: Interventions and gecko bugfixes aren't mutually exclusive. Gecko fixes are generally better than site interventions. So impact of gecko patches is higher.
  • Tom: If we can do N interventions in the same timeframe, that could still be higher impact. Need to assess on a case-by-case basis. Could be good for team building and development.
  • James: In the small, this seems good for people. New skills, more knowledge, ... But I wonder what the bigger picture is here? If we spend more time with C++, there is less time spent on other things. The question is if we have the time to work on that.
  • Karl: Looking for interest at this stage. How to organise it and balance with other work is still an open question. Could just be one bug per half per person at the start. If it works out we could imagine finding ways to make it sustainable. Performance team is an interesting comparison here. For big issues core engineers look at it, but for small things, the perf team fix it directly. But they are also a bigger team.
  • Dennis: Simple issues often don't get addressed by gecko engineers. Actually fixing those would increase our impact on the product.
  • Tom: Necko team have reached out in the past because I've written simple patches that they didn't manage to prioritize. Helps interactions with other teams if we take on these bugs.
  • Karl: Looks like we should find a way to make this happen.
  • Tom: I did about 1-2 bugs a quarter like this, until ETP work.

bfcache (jgraham)

Plan is to allow documents with beforeUnload to participate in bfcache and look at possibility of also allowing documents with unload handlers. Latter might be more of a compat issue.

  • James: Unclear where the compat issues are coming from (wether that's some UX issues, or site developers expecting the handlers to fire, ...). There might be issues with site's unloading, so keep an eye out.
  • Thomas: Sometimes, Ghost-entries in the history appear where you can't navigate backwards. Is this a concern?
  • James: That's probably unrelated. If we start seeing a lot more of this after the bfcache changes, that might be a problem, but I suspect that's unrelated.

Two Minutes ( 👹 )

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

  • Dennis: Vacationing! 10/10, would recommend. Attempting to fix the UA100 experiment code.
  • James: Saw Mozilla London people for a picnic. Helping with WebDriver presentation for Weekly and TPAC. Initial draft of script execution for WebDriver. Proposed a server-mediated message channel for wpt.
  • Karl: Reorganizing the life/work schedule with the new apartment. Mostly finished setting my office. (I have now a desk and a screen. yoohoo). Trying to land a patch for Android. Meetings and diagnosis. Enjoying long bike rides in Tokyo.
  • Ksenia: Updated flask version on and worked on diagnosis. Travelled to the most southern point of the Canadian mainland, Point Pelee.
  • Oana: Triage & moderation, Enterprise testing. Great weather for biking and hiking in the weekend. Met some SV Mozillians in a get together organized locally, before the new Covid wave stroke again.
  • Raul: Triage & moderation, Enterprise testing. Met with SV Mozillians in a get together organized locally. Managed to learn (without any major issues) how to ride a dirt bike.
  • Thomas: Diagnosis and interventions; v27 of the compat addon, SmartBlock 3.0 release for Fx93, Office365 intervention, and dFPI breakage mitigation on StackBlitz, among other things. Also just escaped from a full burn out cycle, and am figuring things out to prevent it actually happening.
  • Joe:

PTO / Travel ( 👹 )

(no more travel until the end of 2021… probably) Coming holidays across the world

  • Oana: PTO 4-5 October, 2021
  • Dennis: Probably not working Fridays for the rest of the year.