Gijs pointed out today that preferences don't generally survive profile refreshes. He also pointed out that it is possible to something kinda gross to make a specific (or three, in this case) survive refreshes, with https://searchfox.org/mozilla-central/rev/3c194fa1d6f339036d2ec9516bd310c6ad612859/browser/components/migration/FirefoxProfileMigrator.sys.mjs#261-284 being an example. He did, however, point out that we should use IOUtils instead of OS.File and see if we can avoid additional file reads in our hook, if we choose to go this direction. Together, Gijs and Leif concluded that it may be possible to avoid doing this if we're willing to accept some amount of error. Leif agreed to do a bit of research into how much error this would be, and how this would (at a high level) change what targeting we needed to use in experiments so that we can come to a decision about which way to move forward. Leif, if you have any particular recommendation once you've got this info, I'd interested in hearing about it too... Giving Leif a needinfo to put this data along with any thoughts in this bug... Thank you!
Bug 1794617 Comment 8 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
Gijs pointed out today that preferences don't generally survive profile refreshes. He also pointed out that it is possible to something kinda gross to make a specific (or three, in this case) survive refreshes, with https://searchfox.org/mozilla-central/rev/3c194fa1d6f339036d2ec9516bd310c6ad612859/browser/components/migration/FirefoxProfileMigrator.sys.mjs#261-284 being an example. He did, however, point out that we should use IOUtils instead of OS.File and see if we can avoid additional file reads in our hook, if we choose to go this direction. Together, Gijs and Leif concluded that it may be possible to avoid doing this if we're willing to accept some amount of error. Leif agreed to do a bit of research into how much error this would be, and how this would (at a high level) change what targeting we needed to use in experiments so that we can come to a decision about which way to move forward. Leif, if you have any particular recommendation once you've got this info, I'd interested in hearing about it too... Giving Leif a needinfo to put this data here along with any thoughts in this bug... Thank you!
Gijs pointed out today that preferences don't generally survive profile refreshes. He also pointed out that it is possible to something kinda gross to make a specific pref (or three, in this case) survive refreshes, with https://searchfox.org/mozilla-central/rev/3c194fa1d6f339036d2ec9516bd310c6ad612859/browser/components/migration/FirefoxProfileMigrator.sys.mjs#261-284 being an example. He did, however, point out that we should use IOUtils instead of OS.File and see if we can avoid additional file reads in our hook, if we choose to go this direction. Together, Gijs and Leif concluded that it may be possible to avoid doing this if we're willing to accept some amount of error. Leif agreed to do a bit of research into how much error this would be, and how this would (at a high level) change what targeting we needed to use in experiments so that we can come to a decision about which way to move forward. Leif, if you have any particular recommendation once you've got this info, I'd interested in hearing about it too... Giving Leif a needinfo to put this data here along with any thoughts in this bug... Thank you!