'Commercial Partners'

[Nicolas Mailhot]

Le vendredi 31 mars 2006 à 14:57 -0500, Eric S. Raymond a écrit :
> Avi Alkalay <avi at unix.sh>:
> > To help Eric fine tune his accuracy, I have to say that I work in a
> > global well known IT company, on Linux solutions sales. We fight
> > everyday for Linux not to die in the commercial world, because of this
> > annoying "missing packages". On the commercial desktop space this is
> > specially annoying.
> > 
> > This is so ridiculous that you still can find Linux sales people
> > making Linux presentations on Linux events using MS PowerPoint. Just
> > because to use Linux isn't that straight forward as Andy may think.
> > 
> > I believe that too much religion in our OSS world doesn't help much.
> > Exactly as real religions, it makes its followers a bit blind to see
> > whats really happening out there.
> > 
> > So I think Eric is pretty accurate in its statement.
> 
> If this list weren't 99% populated by techies, you'd see hundreds of
> mails like this.  

In case you didn't know it a lot of techies work for dilbertian Actual
Users which spend their time asking vehemently for specific features
then changing their mind once the features are delivered.

Filling the feature matrix users ask is nice. It'll get you some initial
sales. And lots of former customers once they discover :
1. they don't really need all that stuff
2. you've spread so thin trying to do everything it all sort of works
but none of it works well (also most of the features are not in your
core competency perimeter so you couldn't have implemented them well
even by focusing your resources)
3. the existing apps on which they based their initial feature request
have been superseded by newer versions (or lost their market share to a
radically different competitors), the latest IT infomercial reviews
they've ingested promotes something else, and you must have
misunderstood their requests the first time you asked.
4. if they had known it would be so bad (believed what you told them in
the first place), they'd rather have it done your way (why don't you
re-do the implementation for free now they've conceded your point?)

User opinion no matter how strongly expressed has to be taken with a big
grain of salt. Usually focusing your efforts on what you're good at is a
better long-term strategy. (As Red Hat proved). Plan for your future,
not your competitor present/past.

FOSS may be bad at realplayer or flash or quicken or whatever, but it's
good at localisation, interoperability, tabbed browsing, thin client
stuff etc. If efforts had been poured in wine or trying to replicate
exactly the windows 95 desktop experience we'd still be nowhere today.
Microsoft is very careful to morph the GUI of its software every few
years, so feature/cloning parity is a loser race. Every other bit of the
windows desktop is irrelevant, as MS kills apps faster than we can
duplicate them (sourceforge is littered with unfinished FOSS clones of
extinct windows apps)

And even if Fedora managed to be as bad as the competition :
1. it's not preinstalled on systems
2. no orchestrated media campaign will happen to brainwash users into
thinking the most stupid misfeature is a brilliant idea (Tim Waugh
declared just does not have the same media weight as Bill Gates thinks
clippy is a brillant feature)
3. software users inertia is such parity means no switching and no new
market share. You want switches you have to be better, and to be better
you can't expend your resources were your competitor excels and you
suck.

If you can bring all you've asked for to Fedora without diverting its
resources or weakening its message fine. Net win, no one will protest.
But anything more costly is like building the steam car customer polls
asks for when your competitors are working on an explosion engines.

Regards,

-- 
Nicolas Mailhot
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 199 bytes
Desc: Ceci est une partie de message num?riquement sign?e
URL: <http://listman.redhat.com/archives/fedora-devel-list/attachments/20060331/fec7cd46/attachment.sig>