Re: focus! (was Re: Focusing on innovation re: mono, python et al)

[Havoc Pennington <hp redhat com>]

Iain * wrote:
> Not sure if this is one of them rhetorical questions or if this was
> even what you meant but its late and I'm bored, split the way I
> understand best; generationally:
>
>  - Younger teenagers who want to stay in touch with their circle of
> friends, share the latest funny video they've found and play some cool
> flash game.
>
> - Older teenagers who have to do coursework and school reports, look
> up information online, develop hobby abilities around the computer,
> express themselves.
>
>  - Students who want to type essays, and keep in touch with friends
> back home and at other unis, letting them know what they've been up
> to.
>
> - People with a role in a club or society (say youth group, church
> activities, sports club, etc) who want to write newsletters, or send
> out flyers.
>
> - Small business/medium owners who need to do basic bookkeeping, stock
> checking and making silly signs that say "No, we have no bananas" and
> "I assure you we're open" when required.
>
> Now I'm getting silly, maybe other people have other ideas if this was
> indeed what you wanted people to do.

I did not mean it rhetorically, no - it's a serious exercise.

I did mean something more in addition to what you listed though - you 
have here audiences and benefits, but the second part is how GNOME 
provides those benefits vs. what they have now.

For example, if I'm a younger teenager, I likely already stay in touch 
with my circle of friends on MySpace/Xanga/Facebook and AIM, share 
videos on YouTube, and play Flash games in Internet Explorer.

So the question is what GNOME (the project and community, not the panel 
and window manager) could offer that an audience _doesn't_ have. Also, I 
was suggesting giving serious consideration to whether the desktop was 
really an intrinsic part of the each individual benefit.

IOW, first think purely about the nail (or screw, or bracket). Then look 
in our toolbox for the right tool. Thinking "how do we hit this thing 
with a hammer?" or "what things can we hit with a hammer?" right off 
just doesn't work. ("hammer" = "a desktop" if it isn't clear)

In the list of existing GNOME successes for example:

a) historical UNIX workstation users want something similar to UNIX

  - this intrinsically involves an alternate desktop, though
    Windows XP and OS X both do have some "UNIX friendly" features,
    GNOME/Linux uniquely offer almost complete compatibility
  - this is a benefit not available to people already, since
    most of the historical proprietary workstations are discontinued
    or not cost effective

b) tech fans want a set of apps (and desktop components) they can mess 
with and customize, and they want just plain old _lots_ of apps

  - this intrinsically involves the mostly open source OS and the
    resulting zillions of free apps an "apt-get" or "yum" away,
    and also the arcane config options and ability to tweak
    than an open source platform offers
  - this is a benefit that was uniquely introduced by the idea of an
    open source OS, so Linux/GNOME are the original market leaders here

c) thin client / computer lab deployments who want something with good 
manageability/security and low cost

  - this intrinsically involves a complete OS + apps solution with all
    of it working in a thin client or lab environment - it couldn't be
    just add-ons to an existing OS very easily
  - people _do_ have serious options with Citrix, Windows, Wyse, etc.
    for this... so GNOME has been strongest when low cost is a factor,
    but there are also some limitations to the competition that Linux
    might be able to improve on
  - the overlap of this audience with the "tech fans" audience probably
    helps GNOME

d) server administrators packing in the terminals, using some web-based 
admin tools, and goofing around online from time to time

  - this most likely only weakly requires a complete OS + apps, and
    indeed lots of sysadmins do run Windows. But Linux/UNIX admins might
    be in categories a) or b) also in many cases, and the Windows
    terminal emulators are pretty crappy. Still, the right app suite for
    Windows might make a complete desktop pretty irrelevant here.
  - the "what do they have already" answer is probably covered by
    the a) and b) overlap above, i.e. "they were already historically
    using Linux or UNIX" - but given that, it's in no way clear
    why a Windows-using server admin would switch, since a) and b)
    hinge on an admin that's already Linux-oriented


Anyhow... take a starting point like:

  - People with a role in a club or society (say youth group, church
activities, sports club, etc) who want to write newsletters, or send
out flyers.

First you have to ask, how do they do newspapers and flyers now. And 
then question 2 is, from their point of view, what's the offering 
(staying open to making the offering an app, a web site, a desktop, or 
anything else) that will really make a meaningful difference in how they 
do newsletters and flyers.

Or step back even further; what's the purpose of the newsletter or flyer 
- e.g. is it to keep club members updated? Would a blog, mailing list, 
wiki, something else be an even better way to do that, thus obsoleting 
the whole newsletter thing? Can something be invented that's even more 
ideal?

An easy trap with a desktop release packed with apps is this line of 
thought:
 - amateur photographers need/want to organize their photos
 - it's possible to organize photos using an app that comes with GNOME
 - therefore GNOME meets the needs of amateur photographers

This is wrong because switching to Linux/GNOME really has nothing to do 
with organizing photos. The tie to the desktop is artificial and 
unrelated to what the amateur photographer is trying to do. If someone 
asks "can you recommend a way to organize my photos?" and you give them 
a whole new computer, they will rightly consider you insane.

A thought experiment here is whether making iTunes Mac-only would have 
gotten a lot of people to switch, or just kept Apple from rebounding and 
becoming relevant again (as it has, due to the iPod).

Another angle on that thought experiment is whether iTunes/iPod could 
have been invented and made successful if thought of as "a way to get 
people to use OS X" instead of as a valuable product in its own right.

Havoc