Aside from the technical solution, I think this is quite inherent in the
market itself.
At the Mozilla Festival last year a group of Mozillians who were keen to
keep old devices alive for various reasons and myself discussed what it
might take. A wide variety of reasons for example it has a nice hardware
feature or they don't want to create landfill. They wanted to keep
devices alive for 2+ years and way longer than what current consumers
would consider the lifespan. I was the only staff member and person that
worked in the industry, the others were all contributors/consumers.
We discussed both the pace of technical development in the form of
pressure from competitors and consumer criticism which leads the
software developers to push onto the latest driver versions and hardware
and leave the old ones behind. This has led even us to have hundreds of
Otoro, Unagi and Leo paperweights.
We also discussed the financial model of selling mobile phones and that
the financial income is not tied to the life of the phone. It is only
tied to the life of the user's SIM card by the contract. There's no
financial incentive to keep pushing updates any more than is needed to
not piss people off. Even then it's probably only a minority of
customers actively pursuing updates.
If the financial model was tied to the life of the phone then there
would be incentive for the carriers to push updates out more frequently,
to keep that phone alive for longer. It might even put pressure on all
parties to streamline and cheapen the process of certifying and pushing
updates.
I think we'll just come up against this problem with every carrier that
sells a device. Thus we either do the updates ourselves (which could get
costly for every device) or try and change the industry's model.
Zac
On 15/04/14 06:58, Fabrice Desré wrote:
Hi all,
The situation with updates on b2g is far from being ideal. On one side,
phone manufacturers have little incentive to ship updates for a long
time. On the other side, users legitimately expect to be able to get the
latest version of the OS for as long as their phone works. I think a
reasonable target is to support devices for 2 years.
Since there's no hope that we will get partners to provide updates for
so long, I've been thinking about what we could do to help users.
One piece is to provide an *easy* way to change the update channel.
Fortunately that's actually simple: we can just use a web activity that
will be provided by the settings or system app, and then do the usual
setting -> gecko pref mapping. That would let any 3rd party or mozilla
provide gecko+gaia updates without any legal issues.
Where things get a bit more complex is when oems use their own update
mechanism, removing gecko's standard one. Currently in this case we
can't even switch users to another update channel. To fix that, we'll
need to provide a proper webapi for updates management instead of our
current mess, with pluggable implementations and the ability to switch
back to the default one.
If we get these 2 parts done, I believe that will move us to a much
better position, where we would be able to confidently recommend devices
to users as upgradable. And I'm sure communities will provide
cyanogenmod-like updates too.
Another annoyance is that we still can't update gaia apps without doing
a full update. That's pretty ridiculous to not be able to upgrade
productivity and media apps for instance. That is blocked on some APIs
being only available to certified apps. If we look at the calendar app,
only the "settings-read" permission is blocking it from being privileged
though. I may be wrong, but I'm less confident in a short term solution
here.
I really want to get the update channel and web api parts happen for b2g
2.0. Volunteers welcome!
Fabrice
_______________________________________________
dev-b2g mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-b2g
