Fedora 20 (yes, F20).

Bill Nottingham notting at redhat.com
Mon Sep 13 20:59:07 UTC 2010


Dunno if I'll make the meeting, so...

Máirín Duffy (duffy at fedoraproject.org) said: 
> Think for a minute or two. What do *you* think Fedora should focus on
> for F15 and F16? How about F17? F18? F19...or... F20? What should F20
> look like, how should it be, who uses it.... at least, for us to have
> succeeded towards achieving our mission as a project?

Given our normal release schedules, that's three years down the road.

A lot can change in computing in three years. Three years ago, there
were no netbooks; there was no Android, and the iPhone had only been
out for six months. Even Facebook had only been open to the public
for a year. For those in the cloud space, KVM was less than six months
old.

So, asking where we want Fedora to be there requires us all to put
on our IDC analyst hats and predict where the market is going. (Or, it
requires us to state we don't care what other people use, we're going
to do our own thing. Which is a valid choice, although I'm not sure
it's an interesting one of we're attempting to spread free software,
content, culture, etc.)

In the past three years, the biggest trends in the basic computing space
are miniturization, mobilization, and the disappearance of the local
app.  More and more the basic user activities are done on a mobile
device (whether laptop, pad, or phone), via wireless networking (whether
wifi or 3G), to some sort of online service. Furthermore, these mobile
devices are taking the place of more special-purpose devices in
the low-end consumer space; the standalone media player, the standalone
navigation unit, even the standalone camera are all being merged
into a single device.

This leaves Fedora in an interesting bind... if we're targeting this
sort of user, we're targeting a shrinking market. The traditional desktop
with local user apps appears to be dying pretty quickly. The devices
that are getting used more for these tasks have OSes pretty well tied
to the hardware by the manufacturer... I suppose targetting people who
want to replace the OS on their tablet PC is akin to where the 'desktop'
in Linux was targeted 10 years ago, but is that a market that grows? Will
it matter? To properly attack this market, we may need to reorient
around something like the proposed GNOME OS. Howver, with Android, and iOS,
and Meego, this is a tightly competitive space.

There are still usage cases and markets where the mobile computing
devices don't have much of a strangehold. No one's developing for Android
*on* Android, as far as I can tell. High scale digital content editing
isn't done too much on mobile, I believe. So, Fedora could be targetting
the developer (corporate or otherwise), the content creator, and so on.
The advantage here is that these people are likely already technical.
However, you're also making the bet that these usage cases aren't going
to move more to the mobile/cloud space, which may be a losing bet. (I
suspect someone's working on the equivalent of Google Apps for coding
as we speak - edit online, get your own SCM automatically, one-click
deployment to a cloud server, all without any infrastructure of your own.)

On the server side, it's all about virtualization; in many respects, you
have the same server that you had three years ago, it's just more about how
they're deployed, and the services you run may be a bit different. While
Fedora could concentrate in this space, it's obviously where RHEL has
the largest foothold, and where RHEL will find the largest deployments.
It's hard for Fedora to push a cultural shift here as it will likely be
seen (right or wrong) as a beta for RHEL in this space.

So, yeah, I'm not sure there's a silver bullet of what I'd want to see
here. Fedora could concentrate on being a server base, which has a use,
but not one that has a lot of growth, and will constantly but up against
its big brother. It could concentrate on the mobile/tablet/etc. space,
but that's a big shift. Or it could concentrate on the developer/content
creator space, which is a small market that I suspect the mobile market
is going to eat up eventually.

There aren't any easy answers, unfortunately.

Bill


More information about the advisory-board mailing list