Desktop Relevance

[gaf-mNDKBlG2WHs at public.gmane.org (Jerry Feldman)]

On 03/26/2009 11:17 AM, Jarod Wilson wrote:
> On Thursday 26 March 2009 10:47:32 Dan Ritter wrote:
>  =20
>> On Thu, Mar 26, 2009 at 10:14:10AM -0400, Jerry Feldman wrote:
>>    =20
>>> I recall back in the 90s when Unix vendors decided to halt workstatio=
n=20
>>> production.  Basically, my point is that while desktop systems don't =

>>> make money, it tends to be the desktop systems that people tend to le=
arn=20
>>> with. I would also agree that in the corporate world, even if a data =

>>> center is all RHEL (or SLES), the desktops are primarily Windows.=20
>>> Actually, this is a good thing for IT people because they spend a lot=
 of=20
>>> time fixing broken Windows systems :-)
>>>      =20
>> It's odd that he would say this just at the time when it becomes
>> clear that ordinary people can enjoy using Linux desktops...
>>    =20
>
> I don't see anything in the part you quoted that says anything to the
> contrary.
>
> He's simply arguing that you can't make money on a desktop OS (unless
> you are ginormous and own 90%+ of the market, aka MS, and get resellers=

> to shoulder the majority of the support burden). But at the same time,
> the desktop is where people learn, and thus to get more Linux server
> users, you might want more desktop Linux users. Then he's just
> conceding that even big Linux shops are still mostly running Windows
> on the desktop. Then he goes for a bit of humour, saying busted Windows=

> boxes are good for IT people's job security.
>  =20
Jarod, that was my comment. Everything after the IMHO was me. I happen=20
to know a few IT guys who are always complaining about having to fix=20
windows boxes where the users have screwed things up. One guy who is the =

only IT guy at his company gets calls nights and weekends.
> Bottom line is that a publicly traded company has to make money, and th=
e
> easiest way to make money on Linux is in the data center. Making money
> on the desktop is HARD. Red Hat very much likes to see more Linux on th=
e
> desktop, but it simply doesn't make financial sense to try to sell and
> support Linux on the desktop. You'd have to build up market share slowl=
y
> over time, and until you reach critical mass, which may well be never,
> you aren't going to actually make any money. Investors don't take kindl=
y
> to things like that, its jut cold hard business facts.
>
>  =20
This is certainly correct, but we probably need to look at the effect of =

how Linux desktop users influence the choice of Linux in the data=20
center. Or the reverse, if a major Linux vendor decides not to offer=20
enterprise desktop services. I recall a study at a major savings and=20
loan in Atlanta where they were going to eliminate the small passbook=20
savings accounts. If I recall, their study showed that a large number of =

them were related to many very large accounts. The bottom line was that=20
they decided to retain the small accounts because they had an indication =

that by eliminating them, they would also affect the large accounts.

Additionally, Jim is talking mainly from the perspective of the=20
enterprise desktop, not the personal desktop. With an enterprise desktop =

there is an expectation of support, where in the personal area (eg.=20
community supported such as Fedora, OpenSuse) support is unpaid=20
community support.

--=20
Jerry Feldman <gaf-mNDKBlG2WHs at public.gmane.org>
Boston Linux and Unix
PGP key id: 537C5846
PGP Key fingerprint: 3D1B 8377 A3C0 A5F2 ECBB  CA3B 4607 4319 537C 5846