Performance Chat Summary: 5 March 2024

Performance Chat Summary: 5 March 2024

Meeting agenda here and the full chat log is available beginning here on Slack.

Announcements

  • Welcome to our new members of #core-performance
  • WordCamp Asia is this week
  • Updated agenda format to match Core dev chat as suggested in previous chats
  • Reminder on timezone difference for the next 3 weeks, this chat will remain at 16:00 UTC, then it will switch to 15:00 UTC from April 2

Priority Items

Structure:

  • WordPress performance Trac tickets discussion
  • Performance Lab plugin (and other performance plugins) discussion
  • Active priority projects
    • Plugin checker
    • Improve template loading
    • INP opportunities research

WordPress Performance Trac Tickets

  • For WordPress 6.5:
    • @joemcgill The only open ticket we still have in the milestone is #60127, which can happen at any time. @thelovekesh was planning on taking a pass at this soon.
      • The only other thing that I think we’ve been aware of is that this release is showing a slightly higher TTFB than 6.4, but we can discuss that in more detail later.
    • @thelovekesh tried to replicate the issue on 6.4 but facing some issues with local setup. Will try to set it up again.
  • For Future Releases:
    • We already have 15 items in the 6.6 milestone, but it’s a good time for folks to start thinking about what they are wanting to focus on for the next release.

Performance Lab Plugin (and other Performance Plugins)

  • @flixos90 There has been a lot of progress on the unbundling of Performance Lab (see overview issue)
  • https://github.com/WordPress/performance/pull/1011 and https://github.com/WordPress/performance/pull/1019 are close to being merged into trunk
  • While that would in principle set us up to ship the 3.0 release, there have been a few conversations on the UX of the new screen that deserve more attention. There is a lot to unpack, but it falls into 2 categories at a high level:
    • The migration from modules to plugins is a confusing user experience.
    • The PL screen itself, with the plugins and modules to activate, is a confusing user experience.
  • The first point affects a one-time functionality, as it was recently decided that the concept of modules will be permanently removed. In other words, this migration will never need to happen again. As it’s already released and users are already migrating, it is not worth investing time to improve it
  • For the screen itself though, there is things we can do.
    • Most importantly, it would be more intuitive to make the experience about the features, rather than the plugins. As in, each card should describe a feature, instead of focusing on the technicality that it’s a plugin. We shouldn’t pull down .org information, but instead provide more solution driven feature descriptions. Things like the star ratings and other things related to the underlying plugin don’t need to be shown.
      • Under the hood, enabling a feature will still install and activate the corresponding plugin. But that shouldn’t matter for the user, their focus should be on the actual features.
    • We need to think about a better onboarding experience, e.g. a step which allows enabling all recommended features at once. There used to be modules which are activated by default, which is no longer possible now that the features are in the form of standalone plugins. So we need a solution to make that intuitive, which goes hand in hand with a better onboarding experience.
  • @thelovekesh suggested Maybe showing some on-boarding wizard? Where user can select what they want to improve in terms of performance on their site?
    • @flixos90 Potentially. I usually think of multiple steps when I think “onboarding wizard”, so not sure that’s needed here. But at least it would need to be a dedicated screen, or a customized version of the regular
  • @joemcgill I’m definitely in favor of the first point above, and see the second as a “nice to have”. Do we already have a GH issue to track this discussion?
  • The most immediate course of action because of those additional points regarding the Performance Lab plugin unbundling: Because of the additional discussion and engineering required, I think we should postpone the 3.0 release. Hopefully only to April. We need to make sure the UX is good enough before we ship it, but at the same time we want to minimize the time we remain in this limbo where the 2.x branch and trunk are so diverged, as that complicates shipping potential fixes in the existing code.

Active Priority Projects

Plugin Checker

  • No updates this week

Improve template loading

INP research opportunities

  • No updates this week

Open Floor

Our next chat will be held on Tuesday, March 12, 2024 at 16:00 UTC in the #core-performance channel in Slack.

#core-performance, #performance, #performance-chat, #summary

Leave a Reply

Your email address will not be published. Required fields are marked *