Viability of Linux companies (Was Re: A Linux Today story has been mailed to you!)

[Raju Mathur <raju@xxxxxxxxxxxxxxx>]

>>>>> "Arun" == Arun Sharma <arun@xxxxxxxxxxxxxxx> writes:

    Arun> On Mon, Apr 16, 2001 at 08:55:23PM +0530, Raju Mathur wrote:
    >> You're falling into the same trap again: confusing product and
    >> service.  Linux distributions are products, and will probably
    >> not sustain any distribution maker.

    Arun> I won't say something as radical as - how much money has RMS
    Arun> made by selling Emacs manuals ? Have you tried comparing the
    Arun> amount of money product companies make and software service
    Arun> comapanies make in the market today ?

Have you had a look at most software and hardware companies' support
and service strategies today?  They're all shifting focus from product
to service, including MS.

    >> What will sustain Redhat, for example, is as you said, manuals,
    >> support, direct service, franchisees, etc.  In fact it's not
    >> too difficult to envisage a world where the RH distribution
    >> remains a loss for RH while their real revenues come from
    >> licensing RHCE exams.

    Arun> As I said again, this idea has been around since early 80s
    Arun> and not one serious company has shown any ability to sustain
    Arun> itself selling manuals, exams or whatever else. However,
    Arun> after reading RH's quarterly reports,

    Arun> http://www.edgar-online.com/bin/edgardoc/finSys_main.asp?dcn=0001021408-01-000325

    Arun> I do recognize that their revenues from RHCE is increasing,
    Arun> but I believe that without their flagship product, RHCE
    Arun> would be worthless.  Note that subscription fees continues
    Arun> to be the king.

It was an example.  Subscription == support if I'm not mistaken IAC.

    Biju> Without them, it will be a lot harder for a lot of hackers
    Biju> to sustain their hobbies. Thats what I meant when I said
    Biju> that $free_software is a luxury. Unless _somebody_ can make
    Biju> large anounts of money off of it, it will always stay in a
    Biju> niche.
    >>  Not too many hackers employed by the distribution makers,
    >> actually.

    Arun> Are you kidding ?

    >> And continuing with AC's example, he can probably make much
    >> more money working out of his home than working for RH.  I
    >> doubt if kernel development will slow down: it didn't when
    >> Linus joind Transmeta.

    Arun> (a) Linus seems to do mobile Linux most of the time at TMTA
    Arun> (b) Linus' percentage contribution to Linux is shrinking.

*Shrug* so someone else will step in.

    >> And if Alan decides to spend more time making money than
    >> working on the kernel, there'll be a 100 more people willing to
    >> take on that mantle.  Finally things will find their own level
    >> anyway.

    Arun> And all these people would be:

    Arun> (a) Paid by the government (b) Be in a research lab or
    Arun> university and do linux under the garb of research

Maybe they'd be paid by SGI.  Or IBM.  Or Compaq.  Or HP.  All of whom
invest significantly into Linux.

    Arun> Most of them don't have a viable economic model to sustain
    Arun> themselves.  They're usually exceptions, rather than the
    Arun> rule and being rewarded by their employers for being
    Arun> exceptional. That model just doesn't scale.

    Arun> Again, compare it to the *BSD world - where most people have
    Arun> a viable economic model to sustain themselves.

I still don't see IBM putting in $1.3B into *BSD.  And Linux outsells
all the BSD's put together anyway, so let's not talk about a losing
product.

    >>  BTW, 58% of the web server market, 70% of the mail server
    >> market, 99% of the DNS server market and 25% of the server OS
    >> market (FWIW) can hardly be described as ``niche''.

    Arun> 100% of the web server market is also useless and
    Arun> unsustainable, if it doesn't pay you any money.

Umm, go back and read the first few lines again.  Try to distinguish
between product and service.  I earned more money last year than 99%
of India's population, without writing a single line of proprietary
code and without touching a single piece of proprietary software.
Looks sustainable enough to me.

Just because a particular individual can't do it doesn't mean it can't
be done.

Regards,

-- Raju
-- 
Raju Mathur          raju@xxxxxxxxxxxxx           http://kandalaya.org/