1713382 - Clean up pre-proton about:welcome components, images, strings, prefs

Status: RESOLVED FIXED

(Firefox :: Messaging System, task, P1)

Description

[Ed Lee :Mardak]

Based on results of bug 1709718 holdback experiment in 89 release showing new proton welcome and pre-proton design, we might get rid of the old stuff. At earliest results would be during 91 nightly cycle.

Bug 1711942 reintroduced browser.aboutwelcome.protonDesign pref to allow switching to the old design.

Bug 1712802 updated the 90 proton experience without directly touching pre-proton, so mac users see "Dock" with proton but pre-proton still shows "primary browser" which only has strings for "taskbar" as well as images that are only packaged on windows from bug 1694578.

Comment 1

[:Gijs (he/him)]

Ed, is this a meta or do you expect the cleanup work to happen here? In the latter case, can you assign a point value roughly corresponding to how many engineering days you think it'd take to get this done? Thanks!

Comment 2

[Ed Lee :Mardak]

Trickiest part will probably be the components cleanup of css. There's already a split between MultiStageScreen and MultiStageProtonScreen:
https://searchfox.org/mozilla-central/rev/bf8d5de8528036c09590009720bc172882845b80/browser/components/newtab/content-src/aboutwelcome/components/MultiStageAboutWelcome.jsx#314-336

but there's the styles for the former component were mostly reused then overridden with .proton specificity:
https://searchfox.org/mozilla-central/rev/bf8d5de8528036c09590009720bc172882845b80/browser/components/newtab/content-src/aboutwelcome/aboutwelcome.scss#504-505

There'll be various logic simplifications that can happen without proton checks including welcome defaults, nimbus feature, importable top sites that aren't used by proton design.

As for strings and images, the more involved part is probably working with product to decide if these are desired for remote experimentation. Having translated strings does allow for expanding experiments to non-en locales.

Comment 3

[Ed Lee :Mardak]

Bug 1709718 holdback experiment only just launched, so it'll enroll for a week before we even get 1 day results and probably want a couple weeks of results before deciding to remove or not, so maybe June 25th?

Comment 4

[Ed Lee :Mardak]

Early results seem to show no harm from the new proton designs. We'll remove the code at the beginning of the nightly 92 cycle.

Comment 5

[Punam Dahiya [:pdahiya]]

With dependencies resolved its safe to close this meta pre-proton clean up tracking bug. I will update Bug 1720763 to be fixed with or as dependent of bug 1704128, thanks