BBS水木清华站∶精华区

发信人: alinew (冷无晴), 信区: Linux
标  题: Where does Linux want to go today?
发信站: BBS 水木清华站 (Thu Mar 25 09:46:24 1999)

Where Does Linux Want to go Today?

By Eric CaldWell

We have a good understanding of where Linux is going today and maybe tomorrow
but, the question remains, how do we get there, and will Linux succeed?  Now,
before you start lighting the pilots on your flame throwers, hear me out.
Linux has become an
overnight sensation due to factors like, repeated delays of the Windows 2000
release, stability and scalability issues with NT4, and the skyrocketing
implementation and support costs associated with Microsoft solutions and
products (not to mention the
ever escalating Client Access Licenses).

So, how can Linux exploit these gaps? Simple, start where Microsoft started
and work your way up. Microsoft is so focused on tackling the enterprise that
they've forgotten where their roots are, small companies, departments within
big organizations,
and the obvious desktop market. We're going to go into a little detail as to
how the Linux community can accomplish capturing these markets.


The small company
Small companies (say 50-200 people) don't have the financial resources or
staff to keep up with big companies. Linux now has a once in a lifetime
golden opportunity to help these companies effectively compete with their
larger competitors by delivering
a mission critical platform, essential for running their business. This
market segment is nothing to sneeze at either since it represents about $42
billion in annual IT spending, and when you look at it from this perspective,
it represents an enormous
opportunity.

Since most small companies already have file and print servers in place
(probably Netware or NT), they are looking at ways to propel revenues by
leveraging technology without spending all their cash reserves in the
process. Linux has a key opportunity
to displace these monolithic systems with an open system that incorporates
key features and has the software necessary to provide the robust mission
critical systems needed to compete in a global market.

Lets look at the major systems that  propel most big companies: email,
database servers, file and print servers, intranets, etc. Linux and many open
source applications can deliver the same functionality at a fraction of the
cost. Believe me when I
tell you that there's plenty of open source software providing commercial
application quality that can deliver a lot of value. These dynamics make
costs palatable for even the smallest of companies.

Small companies are a key area in which Linux needs to succeed because, when
these companies need to grow, they have a platform that can grow with them
and they won't have to displace their current systems or retool their IT
staff to get there.


Departments within big organizations
The next obvious leap is how do you effectively penetrate the corporate
market where NT thrives? Get the big guys on board with Linux (IBM, Oracle,
SAP, etc.) and you have the ammo you need to penetrate this market.

Departments within big and large companies are in a peculiar situation. For
many years they've been in control of their own IT spending which allowed
them to build systems cheaper and faster than their own internal IT
department could. Microsoft saw
this opportunity and seized it by providing cheap robust application servers
which allowed departments to implement systems at a fraction of their
internal costs. Now the tide is turning, before departments would tolerate
crashes and lost data since
these systems were not viewed as mission critical. But now they are mission
critical. Not only have these systems become mission critical, they are
getting expensive to develop for, implement, and maintain.

NT used to be a cost effective alternative to the low/high end UNIX systems
and Mainframes. This model provided departments with a lot of autonomy
because support and implementation costs were lower with NT. Microsoft is now
starting to lose this grip
because of the fees associated with their software and NT application
development costs are also creeping up to match their Unix equivalents. This
does not bode well for NT since it does not provide scalability and mission
critical up times that
departments and applications are demanding.

Enter Linux, stage left. With purse strings tightening at the corporate level
and departments needing to implement their next generation applications for
mission critical systems, Linux now has the opportunity to do to Microsoft,
what Microsoft did to
IBM. Displace them at the departmental level by delivering a cheap robust,
mission critical, reliable platform.


The desktop market
This will be one of the most difficult and lucrative markets for Linux to
displace. Microsoft has a tight grip on this market and will not give it up
without a fight. The way to win this war is to wage the battle at the
corporate desktop, not the home
market, and here's why. When you go to work, you sit down, you then start in
on your day to day tasks like, email, typing that word document or building
that presentation. Because of this routine, the Windows interface and office
applications have now
become second nature to you.

The first thing the Linux industry needs to do is produce a stable robust
window manager that emulates the Windows 9x/NT interface. There are currently
two great window managers that fit the bill and accomplish this task well,
GNOME and KDE (more on
this in a later article). This is very strategic due to the costs associated
with retraining users on a new environment (look at what IE4 did to
productivity). The second item on the agenda is to get a productivity suite
that emulates the Microsoft
Office look and feel so that your users don't go into withdrawl or rebellion
because they are overwhelmed with a new interface. StarOffice 5.0 does this
very well. Once you have the user interface and applications welded together,
then-and only
then-are you ready to do battle.

One other advantage Linux brings to the desktop is stability and control. The
Linux environment is very robust and secure.  This allows IT to keep users
out of key system areas that can get them in trouble (I know this from my own
personal
experiences). Windows is very lax in this area which is why support costs are
so high with Windows in general. These costs are usually associated with
employees downloading or purchasing that new "Gee, this is neat" gizmo which
winds up totally
wreaking havoc on the system. A lot of these gizmos exist for Linux as well
but, Linux does a much better job at protecting itself from poorly written
software and non-technical users.


The enterprise
The final frontier. This could be one of the toughest areas for Linux
successes, unless big companies like IBM, HP or SUN completely toss out their
proprietary UNIX's, and start developing for the open source Linux.

Now, hold on before you let the flame throwers rip. Linux is lacking in some
very key areas that the big companies have poured millions of dollars into
development on. These areas are:

Systems management
how do you partition a Linux to have a test and production region or shared
memory spaces
Directory services
we need a robust and scalable LDAP server
Clustering
the type of clustering that makes 4 UNIX boxes look like a single OS image
and the list goes on. These are critical areas that will need to be addressed
before we can consider Linux an alternative to commercial UNIX's. The good
news here is, NT doesn't have this ability either and will not in the
foreseeable future. For the
small company, Linux can deliver high availability and scalability that
applications like databases and ERP systems require.
As you can see, the Linux community has a lot of work to do to close the gap
on Windows. Don't get discouraged, the gap is not as big as it seems. There
have been many key announcements of recent from major corporations that
deliver many of the key
components that are necessary to run your business. This can give Linux the
chance to change history just as Microsoft did with Windows. If we set our
priorities and attack these critical areas, Linux will be a huge success.

There is one last topic that needs to be mentioned about Linux, the industry
and Linux developers can not afford to have Linux become fragmented with
different versions of kernels and libraries. This will be the death knell for
Linux as it almost was
for the commercial UNIX's. Linus, Alan Cox and other key contributors need to
maintain an active role in kernel development and other key areas of Linux to
achieve what Linus and the rest of us gees have always envisioned, "World
domination with open
source software at the expense of Microsoft".

Make it count! Add your comment to our reader forum


--
[1A一江春水向东流
一江春水向东流
一江春水向东流
一江春水向东流
一江春水向东流

※ 来源:·BBS 水木清华站 bbs.net.tsinghua.edu.cn·[FROM: 166.111.54.156]