Extended Support and Dropping Platforms

[Kai Engert]

Dropping support for a specifc platform has a frustration effect for
owners of devices that cannot be upgraded.

For example, while the Maemo OS platform is not developed any longer,
the N900 device is still usable. However, despite having been expensive
and being just two years old, it's no longer possible to find a quality
open source browser that still receives security updates.

The same is true with the proposal to drop OSX 10.5 support, which will
the end of support for PowerPC Mac hardware, if I understand correctly.

Because of the above, here is a proposal (which repeats some of the
arguments also made in the 10.5 support thread):
"Only ever drop support for a platform after the release of an Extended
Support Release."

In other words, if Firefox 10 is a long term support release, then it's
fine to drop a platform in Firefox 11. If currently at Firefox 12, and
the next Extended Support Release is planned for Firefox 20, then the
earliest time we'd drop a platform would be Firefox 21.

Using this approach, we'd extend Firefox browser support for all users
who cannot afford to upgrade to a new device (or bought a second hand
device), and thereby reaching a bigger amount of users.

Also, if the last supported release for a device was an Extended Support
Release, this is probably a better basis for projects that try to keep
the software alive (cf. Wamcom/Classilla/Tenfourfox).

In my opinion it's Mozilla's mission to empower as many people as
possible, instead of just focusing on recent hardware.

In my opinion enabling a larger amount of people (think poorly developed
countries that use second hand hardware) to be able to use the web
(using a secure maintained browser) should have a higher priority than
adding new features.

When we had a similar discussion in the past, people used the argument
"but you don't any more OS upgrades anyway". I think this shouldn't be
an argument. The browser is the primary gateway to the Web, and it's
used to visit potentially harming sites. In my opinion it's more
important to have a current browser.

Someone will continue to use that old hardware that still runs. If we
keep our browser compatible with the OS, we help those users.

Regards,
Kai

[Boris Zbarsky]

On 12/2/11 7:50 AM, Kai Engert wrote:
> The same is true with the proposal to drop OSX 10.5 support, which will
> the end of support for PowerPC Mac hardware, if I understand correctly.

I believe we have already dropped PowerPC support. That happened with
Firefox 4. PowerPC users are encouraged to use TenFourFox.

This discussion is about dropping 10.5 support on Intel, where upgrades
to 10.6 are in fact possible at least in theory.

-Boris

[Henri Sivonen]

On Fri, Dec 2, 2011 at 3:15 PM, Boris Zbarsky <bzba...@mit.edu> wrote:
> On 12/2/11 7:50 AM, Kai Engert wrote:
>>
>> The same is true with the proposal to drop OSX 10.5 support, which will
>> the end of support for PowerPC Mac hardware, if I understand correctly.
>
> I believe we have already dropped PowerPC support.  That happened with
> Firefox 4.

For the time being, 3.6 still supports Mac PPC and 3.6 has become a
kind of grandfathered ESR.

--
Henri Sivonen
hsiv...@iki.fi
http://hsivonen.iki.fi/

[Mike Hommey]

Note, though, that they're not free. That's a possible reason not to
upgrade for some people.

Mike

[Asa Dotzler]

ESR is not intended as a consumer release. It is for managed deployment
situations.

- A

[Cameron Kaiser]

On Dec 2, 8:23 am, Asa Dotzler <a...@mozilla.com> wrote:
> ESR is not intended as a consumer release. It is for managed deployment
> situations.

Even so, there is still value in only decommissioning platforms after
an ESR. Even if Mozilla doesn't want to issue builds for that platform
anymore, if someone else wants to, they can (and it will be from a
branch that still gets updates). This would have made a long-lived
WaMCom possible back in the day, for example (tip of the hat to Kai).

I fully intend to leverage the upcoming ESR for TenFourFox because I'm
worried that there may be features that the core utilizes in 10.6 that
have no easy, arbitrary or possible replacement in 10.4. For a period
of time we would issue official releases based on the ESR, while still
trying to keep core patches working against mozilla-beta for the
testing audience. If this worked, we jump from ESR to ESR, and if we
fail, we can still fall back on the last ESR and do our own security
patching after that. This is only possible because we know the ESR
will still support what it did support.

There are also a few amateur ports of Firefox to PPC 10.5 (I suspect
they will have some problems with Fx9 and Fx10, but they do exist)
which are essentially unmodified. Guaranteeing the ESR still runs on
10.5 still supports them, and lets amateur Intel builders do the same
when 10.5 Intel support goes away. I continue to get mail from Intel
10.4 users on a regular basis. I'm quite sure I'll get even more from
Intel 10.5 when that support terminates. This lets them help
themselves.

The ESR could also be a safe place to *start* ports from as well --
I'd like to see the AmigaOS guys pick up where they left off with
Shiretoko and Timberwolf, knowing that the widget and XUL code will be
relatively stable.

Cameron Kaiser

[Justin Dolske]

On 12/2/11 4:50 AM, Kai Engert wrote:

> For example, while the Maemo OS platform is not developed any longer,
> the N900 device is still usable. However, despite having been expensive
> and being just two years old, it's no longer possible to find a quality
> open source browser that still receives security updates.

It's very strange that you would use this as an example. Nokia's Maemo
devices are a dead-end platform (both in HW and SW), and have negligible
market share. Why should Mozilla be expected to expend effort to support it?

Browsers on the desktop are vibrant and alive, and the mobile world is
evolving rapidly. We should focus efforts on where they have the
greatest impact for the Mozilla Mission.

P.S. Would anyone like to buy my N800 and N810? They've been gathering
dust for quite some time.

Justin

[Kai Engert]

On 02.12.2011 20:49, Justin Dolske wrote:
> On 12/2/11 4:50 AM, Kai Engert wrote:
>
>> For example, while the Maemo OS platform is not developed any longer,
>> the N900 device is still usable. However, despite having been expensive
>> and being just two years old, it's no longer possible to find a quality
>> open source browser that still receives security updates.
>
> It's very strange that you would use this as an example. Nokia's Maemo
> devices are a dead-end platform (both in HW and SW), and have
> negligible market share. Why should Mozilla be expected to expend
> effort to support it?
>
> Browsers on the desktop are vibrant and alive, and the mobile world is
> evolving rapidly. We should focus efforts on where they have the
> greatest impact for the Mozilla Mission.

I don't quite understand your argument.

The N900 is a mobile device. It was expensive. It's less than two years
old, and it appears to me that the hardware will probably continue to
function for another 2 years. Mozilla has raised hopes that Firefox runs
on it. That was actually the reason for me to buy the device. Now
Mozilla has stopped supporting it. It's not a big problem for me,
because I can afford a new phone. But other people cannot afford it.
Those people are now disconnected from security updates. If those people
continue to use the N900, because they have no other phone, they are at
risk e.g. when using the phone.

I haven't asked that Mozilla tries to keep the platform alive. All I'm
asking is, use smart timing when dropping support for an OS, because it
allows people to be able to use their investment for a longer amount of
time.

What's the Mozilla mission? Is it to support the development of the
Internet? No, in my opinion it's about the people. We want to empower
people to actually be able to use the Internet with our software (and
taking part in the Internet development is just an important part of
that mission). I hope I haven't misunderstood our mission.

If we focus on the group of people who can afford to regularly keep up
with hardware purchases, have we really achieved our mission? In my
understanding, our mission is to do good (I read that on the billboard).
This is what distinguishes us from our browser competitors.

Yes, I understand it makes sense to focus on the platforms that have a
significant market share, because Mozilla doesn't have the manpower to
support any given platform. But if we have *already* done the work, if a
platform *is* already supported, why not try to keep it alive a bit
longer, and make many people happy? Especially if the original equipment
manufacturer has already stopped producing security fixes for the
builtin browser, this is a great way of showing the world why we are
different and how we can help them.

The N900 might, right now, sound like a small problem. I don't know
about the number of devices that have been sold. So, instead, let's
hypothesize about Android 7.

Let's say a new generation of mobile processor is developed, new cool
architectural changes in the OS, new innovative human-machine
interfaces. By that time there might be one billion Android phones in
use that are stuck with Android 6, and you could argue, Android 6 is a
dead-end platform, because it doesn't support all the exciting new stuff.

But how many Android 6 devices will continue to be used anyway, because
they will be second-hand devices for someone who cannot afford an
Android 7 device? I would expect that half of the devices would continue
to be used for at least another 2-3 years, by students, by unemployed,
by people with a small income.

In such a scenario, we shouldn't simply drop delivering security updates
for Android 6 devices, just one year after Android 7 was released. In my
opinion, we should have some extended support release just before
dropping the Android 6 platform. That's a good deed by empowering these
devices to be useful for another while.

I don't want to start a long debate. My understanding of our mission is
to enable many people to use the Internet, and we should avoid cutting
off devices from Security fixes unnecessarily. Smart timing is an easy
way to keep the devices online and safe. If you share my view, then
please consider to use smart timing in the future.

Thanks and Regards
Kai

[Gavin Sharp]

Trying to establish a blanket policy for when we drop support for
arbitrary platforms isn't going to be very successful - each case is
different, and we need to evaluate them on their own merits.

The Mozilla project made a decision to drop support for Maemo because
in our estimation, the cost of maintaining that platform for us far
outweighed the benefits to the relatively small number of users we had
there. We didn't do this blindly - everyone was well aware that this
would mean "abandoning" some users. But that abandonment needs to be
traded off against the abandonment of a much larger set of users who
wouldn't be able to benefit from our awesome new Android browser were
we to spend too much time working on Maemo. (And the costs aren't just
a matter of "time" - split focus has costs too.)

Different people are going to have different evaluations of that
tradeoff, obviously. Some people (users) use the product but don't
need to do any work to support it, while others (developers) need to
do work to support it but don't use the product. The debates usually
tend to revolve around which group isn't empathizing enough with the
other. But these debates are subjective and opinions depend a lot on
perspective, so I don't think we're going to be able to avoid them by
trying to come up with policies in the abstract.

Gavin

On Sat, Dec 3, 2011 at 8:17 AM, Kai Engert <ka...@kuix.de> wrote:
> On 02.12.2011 20:49, Justin Dolske wrote:
>>

>> On 12/2/11 4:50 AM, Kai Engert wrote:
>>
>>> For example, while the Maemo OS platform is not developed any longer,
>>> the N900 device is still usable. However, despite having been expensive
>>> and being just two years old, it's no longer possible to find a quality
>>> open source browser that still receives security updates.
>>
>>
>> It's very strange that you would use this as an example. Nokia's Maemo
>> devices are a dead-end platform (both in HW and SW), and have negligible
>> market share. Why should Mozilla be expected to expend effort to support it?
>>
>> Browsers on the desktop are vibrant and alive, and the mobile world is
>> evolving rapidly. We should focus efforts on where they have the greatest
>> impact for the Mozilla Mission.
>
>

> _______________________________________________
> dev-planning mailing list
> dev-pl...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-planning

[Robert Kaiser]

Kai Engert schrieb:

> The N900 is a mobile device. It was expensive. It's less than two years
> old, and it appears to me that the hardware will probably continue to
> function for another 2 years. Mozilla has raised hopes that Firefox runs
> on it.

And it did and does. It's not "unsupported" completely, it was moved to
tier-3, which means we are welcome with patches by community members to
keep things building and we are happy to even take those probably
(following the usual process), and we're happy about community members
doing community builds (though probably not under the "Firefox"
trademark) and making them available to users.
We'll just not do this with the paid staff of Mozilla or the official
Mozilla build and test systems. We are giving away the N900s er had for
testing probably (a bug is open for that) and if someone wants to
continue community support, I guess (s)he'll even be a preferred
recipient of (some of) those.

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! :)

[Jean-Marc Desperrier]

Henri Sivonen a écrit :

> 3.6 has become a kind of grandfathered ESR.

The trouble is there never was a pushed upgrade.
Which means that many people find themselves on an ESR, when what they
really need is the normal release and it's improvements.