Pausing work on Mentat

Nicholas Alexander nalexander at mozilla.com
Thu Sep 6 23:12:48 UTC 2018


Hi folks,

The Browser Architecture team is pausing work on Mentat
<https://github.com/mozilla/mentat>, the storage layer that we’ve been
building.  We don’t feel we can bring the technology to a meaningful market
fast enough to have the impact we want, so it’s better to focus on other
projects.

The most technically challenging user data our flagship browser products
store is the bookmarks and history data in Places in Firefox for Desktop.
Although Mentat was designed to ultimately serve as the backend for Places,
our initial target has been storage for new mobile products. To that end,
we implemented a logins store and Firefox Sync functionality targeting
Lockbox for Android (logins being technically simpler than bookmarks and
history).  We learned a great deal modeling logins in Mentat, however the
Application Services team has some very tight deadlines in shipping new
mobile products so they have elected to deliver Firefox Sync for new mobile
products based on SQLite, rather than Mentat
<https://mail.mozilla.org/pipermail/sync-dev/2018-August/001676.html>, and
we support that decision.  Looking at the larger schedule, we conclude that
Mentat isn’t on the intercept path we thought we were on and we don’t think
we can adjust Mentat’s trajectory to ship with new mobile products.

Mentat has taught us much, and many of the ideas that informed Mentat will
live on.  The Application Services team’s tactical approach to
commoditizing Firefox Sync is very similar to Mentat’s approach: a single
sync and storage implementation
<https://github.com/mozilla/application-services>, with bindings to Java,
Kotlin, and Swift, expressed in Rust, shared across platforms.  That
approach, as it reduces the number of Firefox Sync implementations, will
reduce the cost of maintaining Firefox Sync and allow the team to evolve
towards a better syncing solution.  And some of the fundamental ideas
<https://medium.com/@rnewman/thinking-about-syncing-part-4-keeping-track-4485207a611d>
about syncing user data look likely to find their way into improvements to
Firefox Sync in 2019.

I’d like to thank all those who have supported Mentat and especially the
Application Services team, who worked closely with us to provide a vehicle
to bring Mentat to market.

Yours,

Nick Alexander, Emily Toop, and Grisha Kruglov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/sync-dev/attachments/20180906/73c27bbf/attachment.html>


More information about the Sync-dev mailing list