[CLUE-Talk] "red hat - the new redmond?" comment from mainstream online media

[Jed S. Baer]

On Sat, 07 Sep 2002 09:46:02 -0600
David Anselmi <anselmi at americanisp.net> wrote:

> So in the sense that Red Hat is *the* enterprise Linux, no others will 
> be considered, you could say they are another Redmond.  Certainly this 
> is a stretch--there are plenty of differences, too.  It is too early to 
> tell how Red Hat will balance their commitment to share-holder value 
> with their commitment to Open Source and their commitment to their 
> customers.

I think there has to be a shareholder knowledge piece to that commitment
though. When people, or institutions, buy stock in a company, it should be
with full knowledge of the business plan, and culture of the company. That
is, after all, what you're investing in. If it doesn't work, but the
company has done exactly what they said they were going to do when you
bought the stock, do you have anything to complain about? I don't think
so.

(Note, I'm talking about your standard investments here, not leveraged
buyouts, or "strategic" purchases such as Micro$quash's large purchase of
Corel.)

> I wonder if large companies really gain so much from standardized, 
> centralized management, or if they'd be better off with more distributed
> management.  On the one hand, the claim is that standardization reduces 
> costs, improves efficiency, and enhances security.  On the other, it 
> forces me to pay $5,000 for a new web server when I could have got it 
> out of the junk room for nothing.

As always, two ... four, maybe more sides to that. If I'm standardized,
then I minimize the knowledge base that I have to maintain and train for.
I need to know how to administer only one mail server program, so it takes
me less time to manage multiple mail servers. I can manage multiple
databases more easily if they are all Oracle, or PostgreSQL, etc.

> One of the things that is interesting (a parallel to the programming 
> world) is that standardization happens at the product level, not the 
> protocol level (that's implementation vs. interface for you 
> programmers).  "We use Word", not "we use 8.5x11 white paper" (paper 
> documents).  "We use Word", not "we use html" (electronic documents). 
> "We use GroupWise", not "we use LDAP" (address book).

Well, standardization happens at both levels, but it is driven more by the
implementations or products.

> In the process, we force everyone to use standard tools rather than 
> allowing them to use tools that suit them (and that interoperate with 
> others' tools).  The system becomes brittle (tightly coupled).  It costs
> a lot to take a bunch of veteran Word Perfect users and move them to 
> Word.  It will cost a lot when we move from GroupWise to Exchange. 
> Well, maybe that's just the nature of the beast.

I hadn't thought of the term "brittle" to describe this, but I think it
mostly fits. The migrations costs are variable. Yes, there are exceptions,
but most office workers don't get deep enough into the fancy aspects of
office programs to be bothered a lot by these things. If they can change
typefaces, bold and underline text, and do outlines, that covers a lot of
ground, and doesn't represent (with modern word processors) a big learning
curve. And yes, there are some people for whem is is a major problem.
Migrating existing documents can be a large task, but one which can also
be done on an as-needed basis, and so, essentially, amortized. As office
suites become more complex, and business begin to use more of the
features, such as calendaring and room scheduling, the burden of
conversion rises, but even this can be mitigated by limiting how much
history is retained. Vendor-supplied conversions scripts/tools can ease
the difficulty.

Of course, as a consultant, I've never become heavily invested in any
particular office suite.

Businesses have to weight the cost/benefit of all these issues, including
the ability to hire people who are already conversant in "technology X".
OK, I'm contradicting my previous argument. ;-) But that isn't always the
case. So it still comes down to, not TCO, but "total cost of
productivity". Can a business save money by switching to Linux? In the
very short term, especially with larger businesses, probably not. In the
long term I suspect the lower licensing and server costs would outweigh
other factors.

jed
-- 
We're frogs who are getting boiled in a pot full of single-character
morphemes, and we don't notice. - Larry Wall; Perl6, Apocalypse 5