[ILUG] openoffice.org

[Matthew French]

(warning: long)

Bryan Hunt argued:
> 1) Its an industry monopoly based on little else other than
incompetent/lazy
> sys admins/support staff who are too dumb to handle anything other than
> point and click.

Hmmm. I am not sure this is entirely fair. Often the people making the
decisions (the competant ones, anyway) are aware of the alternatives. The
problem is, mandating an enterprise wide operating system is more than just
the license fees. You have to find people who can support the product, who
understand how it works, who can deploy it. You need to have applications
that can be supported because people are familiar with them. You need to
understand all the risks involved in changing over.

You then have to sell that to people who who have the latest gadgets but
don't know how to turn them on (the board of directors). You also have to
convince those users who are *gatvol*[1] of having to call technical support
when something breaks and see the roll out of a new operating system as a
new way of imposing on their productivity. And don't forget the MSCE's who
are petrified that you are trying to force them out of work and will
unconciously conspire to undo your efforts. (Politics in a big organisation
can be a difficult thing to manage.) And any cost savings from changing over
will most probably take at least two years to realise - assuming that your
costing includes increased support requirements and user training. And
remember that Linux is not perfect either, although IMHO a better
alternative to Microsoft.

Given that you have dozens of other issues to deal with, from disaster
recovery to enterprise planning, would you really want to consider such a
risky option?

Well, actually, I would. But this is coming from someone who has a well
earned reputation for reformatting any system before taking control of it,
so I am hardly the best person to ask about less risky alternatives.

> > "its proven in industry",
> 2) It's proven to be slow, buggy, a breeding groud for macro viruses

:)

It might be better to say that Linux is "unproven in industry" - as a
desktop OS, especially. This always reminds me of the space shuttle, which
until relatively recently used ferrite core memory - heavy, archaic and not
particularily will suited to the task. So why use it? Because when you are a
thousand miles up in the air and your survival depends on that memory you
would rather use something that has been proven to work even if you know
there is probably something better.

One of the big selling points to investors in the dot-bomb era was that by
being first to market, it was possible to dominate that market[2]. This is
the other aspect of "industry proven". Most people, especially those who do
not care much about their computer so long as it works, would have no
problem installing PalmOS on their desktop if everyone else did it.

Obviously the way around this problem is to prove Linux is ready for the end
user, and the best way to do this is by convincing people to install it one
by one. When (If?) these people see the benefits, they will tell their
friends and so the change gains momentum. The same applies to companies: as
some companies install Linux and hold it up as an example of what can be
done, others will follow.

This is what "industry proven" means.

> > "its well supported"
> 3) Has anyone ever gotten tech support from Microsoft?  Even IBM and
people
> like that never had a clue what was wrong with MS systems that they
shipped
> with their machines.

Well, my experience has been that MS technical support is completely
shocking. I have never, ever, received a meaningful response from them. The
closest I ever came was getting Siemens Nixdorf to explain that the blue
screen problem I was having by reading the MS documentation on the error
code. But today I can get this answer from my friend Google.

The MS knowledge base was useful too, before it followed the rest of
Microsoft and was dumbed down so that the one useful article is buried under
five hundred obvious and useless but similar articles.

In fact, I would argue that the Linux knowledge base, my friend Google
again, is considerably better than the Microsoft alternative. And if I
really get stuck with a problem, I can usually read the source code in
Linux. There have been many occasions when I wanted to do this with
Microsoft products but couldn't.

The problem, though, is that the majority of support comes from people who
happen to know the product. Until there are enough people that know Linux
well enough, and who aren't experienced enough to demand big salaries
because they know it all, then the support issue will go away. I know many
people who do tech support because they enjoy helping people and because
they know enough about computers to help them. They are not the type of
person to play with all the alternatives, and understand them.

As Linux gains popularity, so more people will be available to support it.
As the support grows, so more companies adopt it. And you can be sure that
training companies will be touting Linux as the "hottest career path" as
soon as the market becomes big enough.

To summarise: it is very easy to write off these arguments as stupid. And I
will agree that some people use these arguments because they are a cliche
and an easy way out of argument they do not wish to have. But they are also
valid.

We have forgotten to mention the other old argument: Linux will never catch
on because there are not enough applications for it. Why? Because this
argument is no longer true and it is the reason why I expect Linux will
succeed where OS/2 and  BeOS have failed (I exclude Solaris and BSD because
they are niche operating systems for specific uses. Not better or worse,
just different.).

Finally, I believe all these arguments are valid for a CTO to use, and may
even be a good reason for a small company to hold off on adopting Linux. But
they are no reason for software developers to use MS technologies. C/C++ is
available on Linux, and is better in many respects than Microsoft's
offering. Why use C# when Java is a much better alternative? Visual Basic -
give me Python or Java any day. And on the Microsoft platform you don't have
shell scripting or perl. You have SQL server, but PostgresSQL is not a bad
alternative if you can live without the GUI. Of course, I would fire any
developer that suggested using IIS as a web server, (although I would demote
one that suggested using Netscape Enterprise Server which is almost as bad).

IMHO, the only reason why developers still write code for the Microsoft
platform is because of its market share. I cannot think of any other good
reason.

- Matthew

[1] gatvol: South African expression meaning (roughly) "sick and tired
of/had more than enough of" but in stronger terms.

[2] The flaw with this argument was that often (a) there was no market to
dominate, and (b) there were hundreds of other companies also trying to
"dominate" the same or similar markets. But it provided a ready incentive
for investors to part with cash before doing any kind of due diligence.



_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com