Quantum Flow <3 Photon

Mike Conley mconley at mozilla.com
Wed Mar 29 17:01:23 UTC 2017


Hey folks,

Just following up here - some earlier meetings went long, and so lunch just
started. The meeting is being pushed back a half hour, and will start at
1:30PM (so, 30 minutes from now). I've managed to get us into a room with
Vidyo, and we'll be in the Firefox vidyo room.

We'll be covering (roughly) the following areas:
* Photon metrics and measurement
* Tab opening and closing optimization
* Window opening and closing optimization
* Anything else Photon + Performance related that happens to surface

Feel free to idle, or participate.

-Mike

On 28 March 2017 at 18:53, Mike Conley <mconley at mozilla.com> wrote:

> Hey folks,
>
> Sorry I couldn’t make the meeting today. I’m at a Quantum Flow work week,
> and have a few follow-up things to add:
>
> Regarding specs for machines that we should be testing on, not sure if it
> was mentioned, but we should probably make sure that the Performance group
> (along with the Animations group) is roped in on that.
>
> Regarding Quantum Flow - here's a quick description on what that is, and
> how it related to Photon. If you already know this, feel free to skip
> ahead.
>
> There are a lot of cool Quantum projects underway. These are discrete
> projects that are more or less independent of one another (Quantum DOM,
> Quantum Style, Quantum Render, Quantum Compositor). The teams are highly
> focused, and made of domain experts for that particular region of Gecko.
>
> Quantum Flow is a project that recognizes that swapping out our
> compositor, style engine, etc, does not guarantee us a fast and smooth
> browser. QF is concerned with finding what we do *now* that is slow or
> sub-opitimal, and fixing those ASAP, to give the browser the highest
> probability of being kick-ass for 57 (and leading up to 57).
>
> Up until now, ehsan has been taking point on doing profiles and filing
> bugs. That’s been pretty native-level-centric (with some exceptions), and
> I’ve agreed to be intake (profiling and bugs) with more of a front-end
> focus.
>
> QF is still kinda taking shape, but the process for getting bugs filed
> and triaged is starting to emerge.
> <https://docs.google.com/document/d/1Ka8eNAISQodT1mS_OXapFG-_kk94GoXyo4eKH1j7EV4/edit#heading=h.jm1nihfir2by>
>
> Now how does that relate to Photon? Well, QF wants to make sure that
> Photon is as fast and smooth as possible. Work has already been done with
> this kind of cross-team collaboration and support - the most recent example
> being the film-strip technique that jaws and sfoster are working with for
> the animated icons from UX (which was an idea out of Graphics after they
> saw the UX specs).
>
> I’ll be working with florian and Enn in conjunction with the work already
> ongoing elsewhere with Quantum Flow to make this happen. As per usual, I
> seem to be straddling two worlds here, but I think it’s going to result in
> the smoothest browser ever.
>
> One thing to note: The Quantum Flow schedule
> <https://docs.google.com/spreadsheets/d/1w6n5Jq7GRoEqeDTll5nbIPE-WTdr6yWpg_szi4F3vtE/edit?ts=589224a5#gid=1156329444> has
> Photon stuff being talked about tomorrow (March 29th) from roughly 1PM -
> 5PM ET. I’m going to do everything in my power to get those talks and
> meetings to happen in rooms with Vidyo capabilities. Assuming that’s the
> case, I’ll connect to the Firefox vidyo room in case you all want to
> participate[1].
>
> If you can’t make it, I’ll be taking notes. A few others (Jean,
> shorlander, phlsa, Enn) will probably be in those meetings as well, so we
> have some meeting-note / representation redundancy there.
>
> Feel free to let me know if you have questions! Stoked to ship a smooth
> browser,
>
> -Mike
>
> [1]: I assume the Firefox room is the way to go. Do message me if for some
> odd reason I should choose a different room.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/photon-dev/attachments/20170329/033244f0/attachment.html>


More information about the Photon-dev mailing list