Weekly Engineering Newsletter 2011-08-18
Asa Dotzler
do my best to give you all the week's developer news in 500 words or
less. I'm a day late but I'm also still here so that's sort of a win.
What a week it's been.
*What You Need To Know Right Now*
These trains run on time!
*Planning Update*
Development of Firefox 9 on mozilla-central is already well underway
with about 100 change sets landing in the first three days of the new cycle.
On Tuesday [2011-08-16] morning, the ~3,000 change sets that you all
have landed into mozilla-central over the last 6 weeks were uplifted to
Aurora. This was the largest of the 4 rapid release migrations from m-c
that we've had. Firefox 8 is gonna rock!
Migration day was also kind to Beta. We uplifted more than 2,400 change
sets from Aurora to Beta. Among these Firefox 7 changes are some amazing
memory improvements. When this makes it to users in just under 6 weeks,
Mozilla will re-take the memory efficiency crown and give our users the
highly responsive Firefox they deserve. Thanks to everyone in the
mem-shrink effort for making this happen.
And, it seems worth noting also that we shipped Firefox 6. Cool.
*Must Read Blogs & News*
Armen Zambrano Gasparnian � Go Faster: improved download times for test
jobs and merged few talos suites together
http://armenzg.blogspot.com/2011/08/go-faster-improved-download-times-for.html
Laura Thomson � Rapid releases: one webdev's perspective
http://www.laurathomson.com/2011/08/rapid-releases-one-webdevs-perspective/
Nicholas Nethercote � MemShrink progress, week 9
http://blog.mozilla.com/nnethercote/2011/08/17/memshrink-progress-week-9/
Gavin Sharp � How you can help make Firefox multi-process friendly
http://www.gavinsharp.com/blog/2011/08/16/how-you-can-help-make-firefox-multi-process-friendly/
Benjamin Smedberg � Help improve Firefox power usage!
http://benjamin.smedbergs.us/blog/2011-08-16/help-improve-firefox-power-usage/
Katherine Noyes, PCWorld � Firefox Numbering Debate Is a Tempest in a Teapot
http://www.pcworld.com/businesscenter/article/238412/firefox_numbering_debate_is_a_tempest_in_a_teapot.html
*Cool Features & Fixes*
Two-finger swipe on Mac OS X Lion is much improved with the changes
landed at https://bugzilla.mozilla.org/show_bug.cgi?id=678891 and
https://bugzilla.mozilla.org/show_bug.cgi?id=678607
The Prefs and Find bar visual regression is fixed at
https://bugzilla.mozilla.org/show_bug.cgi?id=677091
The Orion editor has been turned on for Scratchpad on m-c at
https://bugzilla.mozilla.org/show_bug.cgi?id=679475
*Meetings I Went To So You Didn't Have To*
Weekly Firefox/Gecko Development Meeting:
https://wiki.mozilla.org/Platform/2011-08-16
Weekly Product Planning Meeting:
https://wiki.mozilla.org/Firefox/Planning/2011-08-17
Thrice-Weekly Aurora Meeting:
https://wiki.mozilla.org/Firefox/Aurora/2011-08-16
https://wiki.mozilla.org/Firefox/Aurora/2011-08-18
Weekly Beta Meeting:
https://wiki.mozilla.org/Firefox/Beta/2011-08-15
Nicholas Nethercote
>
> Among these Firefox 7 changes are some amazing memory
> improvements. When this makes it to users in just under 6 weeks, Mozilla
> will re-take the memory efficiency crown
It's definitely better, but please don't say things like that -- it's
not true that we're the most memory-efficient, and my interactions
with users make it really clear that they hate it when false
performance claims are made.
That's not to say we shouldn't talk about the memory improvements -- we should.
But when we do, I suggest using the opening sentence of
http://blog.mozilla.com/nnethercote/2011/08/09/firefox-7-is-lean-and-fast-2/
as a a starting point: "Firefox 7 uses less memory than Firefox 6
(and 5 and 4): often 20% to 30% less, and sometimes as much as 50%
less." I worded that very carefully because we have numerous numbers
to back it up, but it still gives us wiggle room to account for the
fact the exact improvements are highly dependent on workload and so
not everyone will see the same numbers.
(Of course, I've already seen numerous headlines "Firefox 7 will use
50% less memory"... sigh.)
Nick
Asa Dotzler
>> improvements. When this makes it to users in just under 6 weeks, Mozilla
>> will re-take the memory efficiency crown
>
> not true that we're the most memory-efficient, and my interactions
> with users make it really clear that they hate it when false
> performance claims are made.
Why do you not think we're the most memory efficient. I get better
results loading any dozen tabs in Firefox than I do in Chrome.
Considerably better. This is a false claim?
> That's not to say we shouldn't talk about the memory improvements -- we should.
> But when we do, I suggest using the opening sentence of
> http://blog.mozilla.com/nnethercote/2011/08/09/firefox-7-is-lean-and-fast-2/
> as a a starting point: "Firefox 7 uses less memory than Firefox 6
> (and 5 and 4): often 20% to 30% less, and sometimes as much as 50%
> less." I worded that very carefully because we have numerous numbers
> to back it up, but it still gives us wiggle room to account for the
> fact the exact improvements are highly dependent on workload and so
> not everyone will see the same numbers.
>
> (Of course, I've already seen numerous headlines "Firefox 7 will use
> 50% less memory"... sigh.)
>
> Nick
I like your opening sentence for posts for the general population. I
don't think this is a press-facing or a user facing newsletter, though.
This is the Weekly Engineering Newsletter posted to our planning forum
for the benefit of our engineering organization.
- A
Nicholas Nethercote
>
> Why do you not think we're the most memory efficient. I get better results
> loading any dozen tabs in Firefox than I do in Chrome. Considerably better.
> This is a false claim?
What came to mind immediately was
http://www.neowin.net/forum/topic/989780-meet-firefox-next/page__st__720__p__594231040#entry594231040.
With FF7 we're much better, but we're still the worst, almost 2x
Opera, which is the best.
Nick
Michael Lefevre
> Nicholas Nethercote wrote:
>> On Thu, Aug 18, 2011 at 11:02 PM, Asa Dotzler<a...@mozilla.com> wrote:
>>> Among these Firefox 7 changes are some amazing memory
>>> improvements. When this makes it to users in just under 6 weeks, Mozilla
>>> will re-take the memory efficiency crown
>>
>> It's definitely better, but please don't say things like that -- it's
>> not true that we're the most memory-efficient, and my interactions
>> with users make it really clear that they hate it when false
>> performance claims are made.
>
> Why do you not think we're the most memory efficient. I get better
> results loading any dozen tabs in Firefox than I do in Chrome.
> Considerably better. This is a false claim?
The last comparison I saw (which wasn't with Firefox 7), it was Opera
and Safari that were doing best on memory usage. If you want to make
grand claims, there should be evidence to back them up.
>> That's not to say we shouldn't talk about the memory improvements --
>> we should.
>> But when we do, I suggest using the opening sentence of
>> http://blog.mozilla.com/nnethercote/2011/08/09/firefox-7-is-lean-and-fast-2/
>>
>> as a a starting point: "Firefox 7 uses less memory than Firefox 6
>> (and 5 and 4): often 20% to 30% less, and sometimes as much as 50%
>> less." I worded that very carefully because we have numerous numbers
>> to back it up, but it still gives us wiggle room to account for the
>> fact the exact improvements are highly dependent on workload and so
>> not everyone will see the same numbers.
>>
>> (Of course, I've already seen numerous headlines "Firefox 7 will use
>> 50% less memory"... sigh.)
>
> I like your opening sentence for posts for the general population. I
> don't think this is a press-facing or a user facing newsletter, though.
> This is the Weekly Engineering Newsletter posted to our planning forum
> for the benefit of our engineering organization.
If you're posting in a public newsgroup, then while you can obviously
still have an intended audience in mind, you should expect that whatever
you say will be picked up by users, press or anyone else.
Michael
L. David Baron
> On Thu, Aug 18, 2011 at 11:02 PM, Asa Dotzler <a...@mozilla.com> wrote:
> >
> > improvements. When this makes it to users in just under 6 weeks, Mozilla
> > will re-take the memory efficiency crown
>
> not true that we're the most memory-efficient, and my interactions
> with users make it really clear that they hate it when false
And we certainly shouldn't go telling people that specific problems
they were seeing before are now fixed unless we've actually tested
that. I don't think we know if the synthetic benchmarks that we've
improved on correspond to (any, some or all of) the cases where
users were observing increased memory usage after using the browser
for a long time.
-David
--
𝄞 L. David Baron http://dbaron.org/ 𝄂
𝄢 Mozilla Corporation http://www.mozilla.com/ 𝄂
Boris Zbarsky
> I like your opening sentence for posts for the general population. I
> don't think this is a press-facing or a user facing newsletter
Asa, everything you say in public is press-facing. Just learn to deal
with it.
-Boris
Andrew McCreight
----- Original Message -----
> On Thursday 2011-08-18 23:30 -0700, Nicholas Nethercote wrote:
> > It's definitely better, but please don't say things like that --
> > it's
> > not true that we're the most memory-efficient, and my interactions
> > with users make it really clear that they hate it when false
> > performance claims are made.
>
> And we certainly shouldn't go telling people that specific problems
> they were seeing before are now fixed unless we've actually tested
> that. I don't think we know if the synthetic benchmarks that we've
> improved on correspond to (any, some or all of) the cases where
> users were observing increased memory usage after using the browser
> for a long time.
There were actually seven or more different bugs that were confirmed to be fixed, bugs that were just regular users noticing that memory usage went up over time during regular browsing (for example, "JS heap grows continually when www.gamespot.com is left idle due to insufficiently frequent GC"). So we do have evidence that at least some of it was fixed in real world examples.
For a more detailed breakdown, see njn's comment here: https://bugzilla.mozilla.org/show_bug.cgi?id=656120#c44
Andrew
>
> -David
>
> --
> L. David Baron http://dbaron.org/
> Mozilla Corporation http://www.mozilla.com/
> _______________________________________________
> dev-planning mailing list
> dev-pl...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-planning
Robert Kaiser
> Why do you not think we're the most memory efficient. I get better
> results loading any dozen tabs in Firefox than I do in Chrome.
> Considerably better. This is a false claim?
Just because you compare us with the least memory-efficient competitor
does not mean we're the most efficient one out there.
And actually, most of Chrome's waste is AFAIK because of the overhead
process separation, which we are looking forward to getting as well, so
let's be careful there.
If we compare being lean, efficient, and fast, we always need to compare
against Opera, who are doing a marvelous job on many fronts there, and
only if we are better then them as well as Chrome, IE and possibly
Safari in *multiple* vastly different scenarios common to different
kinds of users, _then_ we can claim a crown, not before.
And I agree with bz, everything you say in your current position in some
kind of public is press-facing and given how they as well as
significantly visible parts of our own community (and remember that
Mozilla is the community in the end) reacted on some of your recent
statements, you need to be doubly as careful how you are saying things
than already usually in this position.
Robert Kaiser
--
Note that any statements of mine - no matter how passionate - are never
meant to be offensive but very often as food for thought or possible
arguments that we as a community should think about. And most of the
time, I even appreciate irony and fun! :)