Draft Product Description for Fedora Workstation

[Stephen Gallagher]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 11/04/2013 05:32 PM, Kevin Kofler wrote:
> Stephen Gallagher wrote:
>> That's really not true any more in a world built on top of Ruby, 
>> Django, Node.js...
>> 
>> In most cases, the people using these technologies don't use the 
>> distribution packages at all. They instead use
>> 'rubgyem-install', 'easy_install/pip' and 'npm install'. The only
>> thing they care about from the distro (sometimes!) is having
>> those tools present.
> 
> If for you "end user" means "web developer"…

tl;dr: Heavily-qualified "yes".


Well, that's circling back a bit to what we've stated we want from the
Fedora Workstation. (Note: I'm not on that WG). We called it
"Workstation" and not "Desktop" for a reason. When I first proposed
the three products, I wanted us to focus on the things that Fedora has
an opportunity to be great at.

We have some excellent desktop environments (and if you're curious,
I've spent time running most of them; I currently run a
heavily-extended GNOME, but I was using KDE 4 for about a year prior
to that), but while we're having flame-wars over which one is better
for which reason, we have a tendency to lose sight of why they exist
at all. The desktop environment is (to me) an extension of the concept
of the operating system. It should facilitate the work you want to get
done (whether that's actual work or putzing around on Youtube) and not
get in the way of that. For most users, all they really want out of
their desktop is the ability to run their desktop and web applications.

Now, the question we have to face is this (in my opinion): What set of
users are we as the Fedora Workstation going to go after? What group
can we define that is likely to use Fedora and ultimately be
interested in contributing to it, either directly (traditional
contributor) or by building applications and tools atop it.

I'd argue that the "general home user" ship has sailed. Most home
users are happy with their Windows, OSX, iOS or Android devices. That
last has even scooped up a sizeable chunk of the tinkerers and
open-source advocate folks. Targeting this group *right now* is not an
effective use of our time and effort. The desktop environment that
people get pre-installed on their Best Buy laptop is "good enough" for
that group of people.

So let's take a look at who are the users that might be in our reach.
In my mind, there are two groups (with differing needs, of course)
that we have the best chance of reaching. The first are the
traditional Fedora contributors; the people who are working on
delivering Fedora itself. I think it stands to reason that our
Workstation offering should directly benefit their ability to continue
developing Fedora (all of its products, base design and the
ecosystem). This is a place where we do a pretty good job now, at
least in part because it's a bit of a self-fulfilling prophecy.

The second group that we have an opportunity to capture are the people
who are doing the really "edgy" development today. These are the folks
that are using the toolchains we don't love because they don't fit
into our current model of how the world works. These people are
working primarily in web application development because it's the
easiest way for them to target the largest number of users. They don't
have to care if their users are running Windows, iOS or BeOS as long
as the user has a reasonably-compliant browser. This is a group of
developers that right now are developing in one of two ways: They're
developing on Ubuntu to deploy on Ubuntu or they're developing on OSX
to deploy on Ubuntu.

I'm trying in the Server WG to drive an answer to that deployment
problem (as is Matthew in the Cloud WG). What I'd like to see from the
Workstation group is an answer for how to draw people to *prefer* to
use Fedora for this purpose. I have mentioned this on other threads,
but it's not going to be good enough to *match* Ubuntu and OSX in this
space. We need to exceed their ability here. We need to be finding
ways to make Fedora the absolute easiest and *best* place to be trying
all of these new crazy ideas that are taking off.

I absolutely believe that there are risks here in the long run. Some
(many) of these applications are eventually going to succeed and these
developers are going to [re-]learn the difficulties of doing a stable
deployment with an unstable software stack. However, I'm really not
convinced that it's Fedora's place to try to teach them this lesson. I
think it should be our job to enable them to deliver their best new
ideas *right now* and let them figure out when it's time to re-examine
the stable deployment problem.

So this reply has gone on for quite some length, but I hope that I
have represented my position accurately.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.15 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlJ4SEwACgkQeiVVYja6o6OshACfYaIuT1ZR6iB7yK0eo8pWXyMJ
6XgAnjvkSNoSB9ZB43SFdPv+OWzoGwd7
=FG7p
-----END PGP SIGNATURE-----