[KLUG Advocacy] Interesting . . .

[Robert G. Brown]

>>http://www.pbs.org/cringely/
>>"Invincible: How Microsoft and Wal-Mart Can Lose, but 
>>Can't Be Beaten"
>
>Wow,  this guy is on the PBS site?  The article seems rife with
>conservative/republican axioms.  Not that they are neccesarily wrong,
>they are just tossed out in such a, well..., axiomatic way.  Below the
>journalistic standards I expect from PBS.  Reads a bit like an InfoWorld
>article.

Yup, on the PBS site, but I've never seen him on the network. He is a 
commentator who comes out of the computer biz, more or less on the 
technical/corporate side.

I've started something of a dialog with him; with the same book reference
posted on this list. This morning my third reply went out. Cringely is 
taking another look at the book (he read it when it came out); we'll see
what he says after absorbing what I've written, and the message the book
carries, with an eye toward Linux and Microsoft.

>"But at their hearts, each business is successful because it has managed
>to reduce its cost per transaction until it is lower than all its
>competitors"

>Is this the heart of Microsoft's success?  Do they really have a more
>efficient cost-per-unit operation than other huge software houses? 
I tend to think the answer is yes. If you consider this is, in effect,
another way of saying that Microsoft has been very successful in pushing
the installation of product as far up the supply chain as possible. As I
pointed out to Cringely, if MS could figure out how to injection-mold 
Windows and Office onto hard drives, they would.

This lowers the cost of [end user] transactions a great deal. It's costly
to get a copy on Windows out on shelves, with fancy packaging, shelf
space, and installation support. It's CHEAP to get Dell to pre-install
a few jillion copies; only one contract to sign, and the counter-party
does all the work!

>They've grown and scratched more immensely expensive projects over the
>years (Bob, Cairo, etc...).  In my mind their success stemmed from there
>ability to provide a usable platform on cheap ubiquitous hardware, 
>while all their "competition" stumbled about like so many dazed and
>confused trolls.
This is about right. Let's suppose that, except for Windows and Office
being pre-installed, Microsoft has just about exactly the same batting
average as any large software house. They're still way, WAY ahead over-
all, billions to the good, based on their pre-install base alone. They
can actually support a LOWER batting average than many other software
producers.

>"IBM might compete with Microsoft, for example, and its recent
>flirtation with Linux is aimed at exactly that by lowering transaction
>costs."
This is correct, in an indirect way. IBM has recognized for a long time
that writing the platform (OS, compilers, libraries, etc.) is almost
pure overhead. Linux provides IBM with a huge opportunity, something
they are only starting to grasp. IBM is also transforming itself from
[primarily] a hardware supplier to a full-service, diversified provider
of IT solutions. 

>Erm.. no.   I am an IBM customer.  ...
Now, now, Adam, you don't have to disagree! Cringely has a point here. He
simply doesn't have (or write) the WHOLE point.

>Sure IBM would be happy to shelve AIX.
Yes, and MVS, oops, OS/390, or whatever they've called it this week, as
they dismantle an OS brand name that's been in place for 20 years.
A "full-service, diversified provider" means a lot of changes, this being
only one (scrapping in-house OS'es).

>But the "flirtation" (~4.x billion dollars at this point)
How much?
Lemmetellya, it's chump change to IBM, and for every dollar they've spent
on Linux, they've spent 2 or 3 building consulting services, industry-
specific applications and expertise, and capital spending, like facilities
to handle outsourcing. The billion IBM has spent on Linux development plays
very well, and it has resulted in some real gains on some projects, but it
is not what I would consider a heavyweight commitment. Actually, much of
the enthusiasm that IBM has for Linux is that it is a LIGHTWEIGHT consumer
of corporate resources, and this thus rather lower risk as a result.

>...customers said - "Wow!  Apache, Samba, Cyrus, etc... runs great
>on that dual x86 box, but not so great on my {AS/400 | S/390 | RS/6000,
>zSeries ??? | iSeries }  Hmmm...  What kind of server will I spend my
>money on?"  The minute IBM heard this enough times they spun their Linux
>attitude around so fast you could hear the corporate tires squeel.
So you are saying that 


>You can have immeasurable low cost-per-unit,  but piss off your
>customers and watch out.  IBM bends over back wards NOT to do just
>that,  the M$ guy practically moons you on the way out.
But no one sees "the MS guy"! They're all too busy selling to the Dells
and Gateways, or getting into bed with mid-level developers and "consulting"
firms.

Customers can be really angry, but if they have no alternatives, there's 
no place for them to go. This is why the book I cited was so interesting;
it provides a reasonably consistent model for HOW leading firms in a market
lose their dominance, over time, and across a number of industries. I'm
looking for some reason why the same forces don't apply to the software 
business.

>"If IBM really intended to compete with Microsoft it would COMPETE WITH
>MICROSOFT, which would require exiting the hardware business
>completely."
>Why?
Because the hardware business is low-margin stuff (compared to software),
and because it keeps IBM involved in the sort of sales that don't allow
many per-transaction costs to be eliminated. This seems to be the Cringely
thesis, anyway.

>So they only "pretend compete" now? Hey, we are only talking about those
>"pretend" dollars you might spend on either an NT cluster or a main
>frame.
The point here is that IBM and Microsoft are not quite in the same business,
and in some ways they had more overlap in the past. In fact, IBM is currently
a fairly large installer of Microsoft-based systems, and they make a lot of
products that rely on Windows. The difference is that IBM wants to be in every 
aspect of the product life-cycle now... they'll not only sell you the hardware
and service it, but they'll come in as consultants and study your business,
recommend what you need to meet requirements, and provide the applications,
upgrades, and custom programming talent.

Microsoft doesn't want any part of many of these activities. They will help
with some of the specification process, sell you the software, and try to
show you which MS off-the-shelf apps just might fit your requirements, but
I believe that's about the extent of it. Outsourcing? Support (other than
what they provide any paying customer)? Outsourcing? Custom consulting and
programming? Forget it! Microsoft isn't in those businesses, and they don't
want to be, not for you, and not for the largest corporates. Oh, if you are
a multimillion dollar a year customer of MS, and your enterprise-level 
machines need some high-level optimization, they'll send out a team (at, I
understand, protean cost), but that's about the extent of it.

IBM sees many of the business activities that MS won't touch as being some
of their most lucrative, while understanding that most of the basic stuff
(platforms and basic server stuff) are becoming commodities.

>This statement is a bit over-arching.
Yes, it's a partial statement of a broader picture. I get that feeling
from reading some commentators; Cringley is no exception.

>"Both Wal-Mart and Microsoft will eventually founder, probably from
>inbreeding and corporate crankiness decades down the road. Until then,
>it is their game to lose, and the best way to compete is probably by not
>competing at all. Just be what they aren't and where they aren't. And,
>like mammals in the age of the dinosaurs, wait for that comet to
>strike."
Doesn't seem like he's learned some lessons very well, mostly about how
emerging technologies have their own proponents, and how premium companies
become welded to their pricing and strategies.

It's easy (and a good percentage play) to forecast the dominance of the
currently dominant players for an indefinite period into the future, but
such conventional wisdom often does not pan out, and that's what the
book is actually about.

>I don't know jack about Wal-mart.  But decades to develop corporate
>crankiness?   From someone who has tried negotiating reasonable
>licensing Microsoft has already hatched that egg.
You have a different definition of this, I suspect. I tend to think both
rather miss he point.

>And "By dominating shelf space, concentrating on market share, and
>making product feints into segments it doesn't really care about,
>Microsoft keeps many potential competitors literally off the field" is
>illegal, after a certain point.  It certainly is unethical.
Yes, to both of your comments. MS'es tactics only prolong their position,
it is not preserved or guaranteed.

>The idea that the internet, pda's, etc... came into being like a comet
>smashing the earth is baloney....
I did point out that things don't often happen like this, the analogy is
somewhat too confrontational for my taste. More often, dominance in a
market fades, replaced by a new technology that has strengths that exploits
the weaknesses of the old. I fell that OSS meets these requirements.

>Smart people saw these things coming, and Microsoft probably sees things 
>coming I/we don't even know about...
Yes, but are they going to be important. MS has missed a number of things
in the last few years which have turned out to be fairly important, and it
is a predictable pattern.

>(after all, according to the article, they have all the smartest people).
Which, even if true, is no guarantee of ongoing success. Moreover, this 
is exactly the sort of thing that breeds over-confidence, a source of failure.

>The only way to compete is to compete;...
Sure, but the competition is often not, perhaps cannot be... face-to-face,
direct. It starts with scattered smaller applications, and goes up in 
scale and value as the technology matures. I can't say how far along this
is with regard to Linux, but I do think it fits the profile (Read the book!),
and that I see that continuing.

>...to establish some elvish enclave and only play in arenas where you 
>KNOW you can win while you wait for the world to change is a one way 
>ticket to the ash heap of history.
Right, the world doesn't change by itself.  It does change when something 
comes along that, even if not mature, offers a better value proposition.

Oh, did I mention that you ought to read the book? :)

							Regards,
							---> RGB <---