OT - Re: [Evolution] Advice - Why emulate MS?

[Tom Cooper <tom_cooper bigfoot com>]

Sorry that this turned out to be so long, but I think that the answers
to Dan's questions matter.

I'm sorry that it's 
a) off topic, and 
b) so dang long


"Dan B. Mann" wrote:
> How do you see Evolution being able to replace Outlook in your
> organization?  I see some fundamental flaws:
Replace?  I don't think it will replace Outlook.  Completely removing an
entrenched technology is expensive and difficult.

> No equiv of MS Exchange
Who says that we need to dump Exchange?  
Even if we do, there are alternatives.....It depends what functions are
really important to our organization.  At this point, I think my
enterprise is committed for quite some time to the MS platform.

> No real "NT Domain" type of user cohesion/administration
Who said we'd rid ourselves of NT?  Long term all organizations will be
moving to user admin based on a directory model rather than an OS model.

> Linux is too hard for some people to administer in it's current state,
That's certainly true, but having worked with MS products professionally
since '88, I can say that they're no barrel of laughs, either.  Sure,
the setup wizards make it easy to get started, but just let the smallest
thing get out of place, and the system is really hard to deal with.  As
a matter of practice, most of our 9x boxes get the OS wiped and
reinstalled during their operating life because of the cruft that comes
as that crummy design is actually implemented.

MS ain't cheap to acquire or to operate!

> and most people don't want to take the time to learn when they already
> know something that puts bread on the table
You're right, and no one should implement a technology for philosophical
reasons.  The reason we're supposed to implement technology is for the
purpose of helping our companies to make money.  A part of that is
finding a comparatively inexpensive mechanisms.  

Linux in some segments already provides that - from ISPs to movie
production (Titanic's producers used Linux so they wouldn't need to buy
DEC/OSF/Compaq unix - and they saved a bundle!)  In other
segments....it's coming.


> I can't see each computer needing to have an
> administrator go to each machine individually when a new user will be
> added.  It just doesn't make any business sense, and nobody in the
> Linux movement seems to notice this gigantic "black hole".  
Just a minute.  Why is this a Linux issue?  We have the same problem
dealing with any OS implementing a moderate amount of security.  NT has
this VERY same issue, and it's a cause of much heartburn to us.  

Sure, NT2K provides group policies and the like, but there are real
issues with modeling the level of permissions needed by specific users
and granting them only what is needed - as systems are part of a domain
(MS domain) model they are subject to domain permissions - which local
groups and global groups are related, which take precedence?  How much
can be effectively locked down, etc....

How do you handle the issue of local users?  Who gets the
administrator/root password for the box?  How is it maintained/modified?

These issues are cross-platform, and don't go away with Linux or with
NT.

> think about samba 
(rant removed to save space)
Why?  It's a great solution for people who must support mixed
environments, and it's really effective.


> Why can't we create our OWN ...system of file and print sharing 
Must we fight city hall?  Like it or not (I choose not) MS owns the
desktop.  The way to make an effective change is to make the cost of
change as painless as possible.  Novell did this by making an excellent
product, and then making it easy for others to use it to integrate
disparate systems (Mac, Windows, DOS...to a certain extent SNA and Unix)

If the technology we prefer is to succeed, it must be cheaper to buy, or
demonstrably better to own, and easy to migrate to.  This last point is
one of the ways that MS managed to get large organizations to implement
their products - they made the products work well together (and broke
their opponents) and gave companies project plans and free consulting to
help smooth the process.

Besides, there are a couple of file sharing technologies already
available - Samba, NFS, and to a certain extent ftp and web servers
provide some of this as well.

> It seems totally ass backwards to me to go through all of the reverse
> engineering when 
> we could set the standards, design the system with
That would be nice, but in order for that to happen, we'd HAVE to get MS
to play along - the desktop is their sandbox, not ours, and they're the
big bully who tells us all what to do.

The only way to really 'beat' them is to develop a new paradigm to which
MS cannot adapt.  This is what MS did to IBM, and they've become very
successful in the process.

MS will miss a technology shift, it's just a matter of time.  It almost
happened with the Internet, and again with Palml computing.  

They're talking seriously about the centralized computing paradigm, but
I'd bet that there are real opportunities available....even if MS gets
it, they can't really effectively change the computing model that
they've spent years developing - just like IBM couldn't get used to the
idea of distributed computing.


> I Cannot see Linux (READ: EVO) taking over the desktop of a
> corporation until some really basic sound business features are added
> to Linux.
I'm not sure that I agree with that.  

There's more maturity needed on the Linux desktop before it's ready to
replace what we're using today, BUT we spend buckets of money on
proprietary licenses and maintenance, upgrades, and the like.  A
business case can be made for replacing a certain percentage of desktops
with open source alternatives.  I believe an argument can be made for
incurring the additional cost of supporting a different platform, if the
acquisition and maintenance costs can be demonstrated to be sufficiently
low.

I'm not even all that concerned about retraining people who already get
the current model - Linux can be configured to work-alike Windows, and
also can be used for point platforms to solve problems we already have
with our existing systems.


In the mean time, I'm using Linux at home (and unofficially at work,)
providing feedback to developers, answering questions on mailing lists,
and helping encourage people who are open to new/different
technologies.  I'm watching carefully, and I like what I'm seeing.  I
think that we can see some changes in the desktop landscape, and it will
be a fun ride along the way.

When it gets just a _little_bit_ more mature, I'll be ready to propose
this for us as a business decision, not just a cool technology.  Until
then, I'm hanging on tight!

Regards,
Tom Cooper
-- 
Standard disclaimer applies:
This message represents the opinions of the
author, and not necessarily those of any 
organization to which he may be related.