The intent here is to understand the risks around multi-profile experiments. With this spike, we hope to gain an understanding around: * which multi-profile use cases are being prioritized for the initial work * how set-to-default as a special action should work in the supported multi-profile cases * how targeting, triggers, and frequency-capping should work in the supported multi-profile cases * which things make sense to happen across all profiles, and which should be restricted to a single profile * what sort of infrastructure might make sense for the shared prefs, data, notifications * better understanding of the assumptions and trade-offs being made by both Nimbus and Multi-Profile Team * documentation about what infrastructure work is likely to be needed now and in the future * which risks are apparent and which need further investigation One approach would be to take a past multi-profile experiment, clone it, and start playing with it with multi-profiles enabled, investigating the above points, and keeping notes. Perhaps even just copying the above points to a Google Doc and filling it out as things become clearer.
Bug 1927614 Comment 0 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
The intent here is to understand the risks around multi-profile experiments. With this spike, we hope to gain an understanding around: * which multi-profile use cases are being prioritized for the initial work * how set-to-default as a special action should work in the supported multi-profile cases * how targeting, triggers, and frequency-capping should work in the supported multi-profile cases * which things make sense to happen across all profiles, and which should be restricted to a single profile * what sort of infrastructure might make sense for the shared prefs, data, notifications * better understanding of the assumptions and trade-offs being made by both Nimbus and Multi-Profile Team * documentation about what infrastructure work is likely to be needed now and in the future * which risks are apparent and which need further investigation One approach would be to take a past set-to-default experiment, clone it, and start playing with it with multi-profiles enabled, investigating the above points, and keeping notes. Perhaps even just copying the above points to a Google Doc and filling it out as things become clearer.
The intent here is to understand the risks around multi-profile experiments. With this spike, we hope to gain an understanding around: * which multi-profile use cases are being prioritized for the initial work * how set-to-default as a special action should work in the supported multi-profile cases * how targeting, triggers, and frequency-capping should work in the supported multi-profile cases * which things make sense to happen across all profiles, and which should be restricted to a single profile * what sort of infrastructure might make sense for the shared prefs, data, notifications * better understanding of the assumptions and trade-offs being made by both Nimbus and Multi-Profile Team * documentation about what infrastructure work is likely to be needed now and in the future * which risks are apparent and which need further investigation One approach would be to take a past set-to-default experiment, clone it, and start playing with it with multi-profiles enabled, investigating the above points, and keeping notes. Deliverable: Google doc with the above points (and any other info that seems worth capturing) filled out, along with a TL;DR section at the top.
The intent here is to understand the risks around multi-profile experiments. With this spike, we hope to gain an understanding a variety of things; see https://docs.google.com/document/d/1Z5pU_VcuRaWBfb26JTWqTOXG-55nxtQrovKN6x1YITQ/edit?tab=t.0 for details. Proposed approach: clone [FOX_DOODLE_SET_DEFAULT](https://searchfox.org/mozilla-central/source/browser/components/asrouter/modules/OnboardingMessageProvider.sys.mjs#1087), and start playing with it locally with multi-profiles enabled, and go through the above document, adding info/commentary/questions as you go. We _could_ consider deploying this to nightly once multi-profiles are turned on and looking at the telemetry to learn more about behaviors in the wild. Deliverable: * The Google doc has been updated with ** any info that seems worth capturing ** a short summary about this particular investigation and how it is similar / different from the other spikes discussed. Include a link to this ticket.
The intent here is to understand the risks around multi-profile experiments. With this spike, we hope to gain an understanding a variety of things; see https://docs.google.com/document/d/1Z5pU_VcuRaWBfb26JTWqTOXG-55nxtQrovKN6x1YITQ/edit?tab=t.0 for details. Proposed approach: clone [FOX_DOODLE_SET_DEFAULT](https://searchfox.org/mozilla-central/source/browser/components/asrouter/modules/OnboardingMessageProvider.sys.mjs#1087), and start playing with it locally with multi-profiles enabled, and go through the above document, adding info/commentary/questions as you go. We _could_ consider deploying this to nightly once multi-profiles are turned on and looking at the telemetry to learn more about behaviors in the wild. Deliverable: The Google doc has been updated with 1. any info that seems worth capturing 2. a short summary about this particular investigation and how it is similar / different from the other spikes discussed. Include a link to this ticket.
The intent here is to understand the risks around multi-profile messages outside of experiments and rollouts. With this spike, we hope to gain an understanding a variety of things; see https://docs.google.com/document/d/1Z5pU_VcuRaWBfb26JTWqTOXG-55nxtQrovKN6x1YITQ/edit?tab=t.0 for details. Proposed approach: clone [FOX_DOODLE_SET_DEFAULT](https://searchfox.org/mozilla-central/source/browser/components/asrouter/modules/OnboardingMessageProvider.sys.mjs#1087), and start playing with it locally with multi-profiles enabled, and go through the above document, adding info/commentary/questions as you go. We _could_ consider deploying this to nightly once multi-profiles are turned on and looking at the telemetry to learn more about behaviors in the wild. Deliverable: The Google doc has been updated with 1. any info that seems worth capturing 2. a short summary about this particular investigation and how it is similar / different from the other spikes discussed. Include a link to this ticket.
The intent here is to understand the risks around multi-profile messages outside of experiments and rollouts. With this spike, we hope to gain an understanding a variety of things; see https://docs.google.com/document/d/1Z5pU_VcuRaWBfb26JTWqTOXG-55nxtQrovKN6x1YITQ/edit?tab=t.0 for details. Proposed approach: clone [FOX_DOODLE_SET_DEFAULT](https://searchfox.org/mozilla-central/source/browser/components/asrouter/modules/OnboardingMessageProvider.sys.mjs#1087), and start playing with it locally with multi-profiles enabled, and go through the above document, adding info/commentary/questions as you go. We _could_ consider deploying this to nightly once multi-profiles are turned on and looking at the telemetry to learn more about behaviors in the wild. The intent is to try and an initial investigation done by end of day, Monday Dec 2nd, so that drafting of a plan to address what we have found can start on Tuesday, December 3rd. Deliverable: The Google doc has been updated with 1. any info that seems worth capturing 2. a short summary about this particular investigation and how it is similar / different from the other spikes discussed. Include a link to this ticket.
The intent here is to understand the risks around multi-profile messages outside of experiments and rollouts. With this spike, we hope to gain an understanding a variety of things; see https://docs.google.com/document/d/1Z5pU_VcuRaWBfb26JTWqTOXG-55nxtQrovKN6x1YITQ/edit?tab=t.0 for details. Proposed approach: clone [FOX_DOODLE_SET_DEFAULT](https://searchfox.org/mozilla-central/source/browser/components/asrouter/modules/OnboardingMessageProvider.sys.mjs#1087), and start playing with it locally with multi-profiles enabled, and go through the above document, adding info/commentary/questions as you go. We _could_ consider deploying this to nightly once multi-profiles are turned on and looking at the telemetry to learn more about behaviors in the wild. The intent is to try to complete an initial investigation by the end of the day on Monday, December 2nd, so that drafting a plan to address what we have found can start on Tuesday, December 3rd. Deliverable: The Google doc has been updated with 1. any info that seems worth capturing 2. a short summary about this particular investigation and how it is similar / different from the other spikes discussed. Include a link to this ticket.