Subject: Re: Wal-mart drives software industry
From: Tom Lord <lord@regexps.com>
Date: Wed, 27 Feb 2002 01:39:05 -0800 (PST)



   From: Zimran Ahmed <zimran_ahmed@yahoo.com>

   > lord:
   > I think that "everyone" has goals that are best
   > advanced by freeing
   > nearly all software, but that that isn't universally
   > appreciated yet,
   > in part because FSBs, taken as a whole, aren't yet
   > doing their whole job.

   It's *exactly* comments like this one that lead me to
   my conclusions. "Everyone's" goals are "best" served
   by freeing nearly all software? Clearly proprietary
   software companies don't fall into this category.

In the case of enterprise IT software, the customers have an interest
(whether they recognize it or not) in GPL'ed software.  Therefore,
their suppliers ultimately share that interest.

Why do customers have such an interest?  lma said it pretty well:

    IT is generally a risk, not a competitive advantage.  There are a
    few exceptions.  Some companies have chosen to make IT a
    competitive differentiator (Wal-Mart comes to mind).  But that is
    not the general case.

IT is a bit like the power supply.  The customers have a common
interest in its ubiquitous availability, functionality, and
interoperability and the engineers have an interest in its
transparent, robust, and tractable architecture and implementation.
In an increasingly complex economy, the quality of your neighbors' IT
matters as much to your business as the quality of your own IT -- for
most companies, this consideration outweighs even the worth of a
differential in the quality of your IT vs. your competitors' IT.
Wal-Mart is a forcing function - it forces a lot of retail to either
get with the picture or go away forever.
 


   Buyers don't want to be locked in, and buyers want
   competition for their dollars, and free software may
   or may not have an important role in this, but this is
   not the same thing as buyers wanting software to be
   freed. 

Buyers of software are buying into a process and that process is best
implemented on top of Free Software for reasons well and often
rehearsed on this list.  In addition to being an issue about
competition, it's an issue about engineering.


   It is the job of free software businesses to serve their customers
   by hopefully selling them stuff for more than it cost them to put
   it together.


Accurately measuring that cost, though, is an interesting problem.
Linux distributors inherited a large amount of software that developed
very slowly, through volunteer efforts paid for by other organizations
and by individuals.  While they have contributed much beyond that in
lines of source code, they have so far failed to contribute much by
way of lines of source code having comperable impact.

If they are indeed "solution providers", then the solution must
include the continuation of the process that gave them birth, but at
business speed.  What does that cost and what's it worth?  How can it
be implemented?  I've put forward my proposals -- where are yours?


   You contribute great volume to this list, and it's
   good to see your dedication (I am dedicated too) but
   missionary zeal will not help free software in the
   enterprise. Mission critical apps on GNU/Linux will,
   and free mission critical apps are thin on the ground.

"thin on the ground" for structural reasons which I've described at
length.


-t