Re: Linux on the desktop

On Mon, 8 Dec 2003, Daniel C. von Asmuth wrote:

> Why is Linux so slow to conquer the desktop?

I would dispute that if you're talking about corporate desktops - I think
real movement is happening in that space these days.

As far as the _consumer_ desktop goes, the users of such machines are not
really ready to accept the changes that FOSS brings. They're used to buying
individual shrinkwrapped software from supermarkets and high street stores
and expecting it to work on their machine (rather than searching the
Internet for equivalents, or even having it included on the original
installation media!) They're used to buying any old pieces of hardware and
have them work out-of-the-box (to a lesser or greater extent!) rather than
searching for and compiling a driver set. Lindows' Click-N-Run service might
suit this kind of user nicely, if they can build up a critical mass of
software and drivers and get the subscription charges right.

> The major obstacle seems to be the lack of applications.

Which in turn has a significant cause that is the paucity of preloaded Linux
machines for sale. Which in turn has a significant cause of "lack of
consumer demand", no doubt.

> Why are there so few applications for Linux? 
> 
> Well, that's no true. If you search freshmeat and sourceforge you will
> find an ever growing number of software projects. The question should
> read "why are there so few commercial applications?".
> 
> The size of the market is smaller than for the Competition, but this is
> not the whole answer.
> 
> When you consider the failure of Linux desktop software like WordPerfect,
> ApplixWare and Kylix, an important reason would be the Gift Culture of the
> Linux crowd. They will go for anything that comes free of charge if they
> can. There is no Photoshop for Linux, because Gimp is good enough. There
> is an Oracle, but MySql is good enough for most users. TextMaker is a nice
> word processor, but the majority favours OpenOffice. (giving Sun a hard
> time to sell StarOffice). 

Commoditisation. It happens to everything eventually. For 80-90% of users,
the non-commercial alternatives you highlight will suffice. Only a small
proportion of users need (and, more pertinently, /are willing to pay for/)
the advanced features provided by "best of breed" products. When
conveniently presented with a viable alternative, they'll generally go for
it if it's cheaper or free.

> A viable commercial Linux application will need some unique feature. There
> are examples, like Mathematica, SPSS, and BitKeeper, but these are outside
> Joe Sixpack's budget.

Joe Sixpack probably couldn't find a need for those kind of applications
anyway. Only a small proportion of Linux users (who in turn presently
represent only a small proportion of the total software market) have a need
for that sort of thing.

I foresee a future where the masses are mostly provided for with FOSS,
leaving proprietary software vendors to cater for the high-end and
specialist users who have wallets to match their requirements. Such vendors
may even choose to extensively use FOSS (whenever possible) in creating
their products so as to reduce the cost of development and the
time-to-market.

> Happy Holidays,
> Daniel

Best Regards,
Alex.
-- 
Alex Butcher      Brainbench MVP for Internet Security: www.brainbench.com
Bristol, UK                      Need reliable and secure network systems?
PGP/GnuPG ID:0x271fd950                         <http://www.assursys.com/>