Re: bias

On Fri, 7 Jun 2002, Anthony Taylor wrote:

> Deven T. Corzine wrote:
> 
> >On Fri, 7 Jun 2002, Anthony Taylor wrote:
> >
> >>The idea that the GPL is going to destroy all computer jobs is a red 
> >>herring.
> >
> >It may be.  Then again, most of the arguments against it assume proprietary 
> >software remains dominant.  Do they still apply if GPL software dominates?
> 
> I'm not sure what you mean; is the antecedent for "it," my red-herring 
> statement?  If so, then the arguments to which you refer are the pro-GPL 
> arguments that concern the availability of work for computer professionals.

"It may be a red herring."  I was saying you might be right.  I don't know 
that you're wrong here.

> In this case, I think there are several assumptions.  First: proprietary 
> software is superior to libre software only to those who own the 
> copyright, and are able to profit from it.  So in this case, professions 
> not based on commercial software programming will benefit more from GPLd 
> software than from proprietary software.  This includes proprietary 
> in-house extensions of GPLd software, and the use of unencumbered 
> standard software, like OS and office suites.

GPL software is also proprietary -- only those willing to submit to the GPL 
are able to benefit from it.

It's not clear to what extent you can make proprietary in-house extensions 
of GPL software, actually.  If you modify it and distribute it to another 
site, even within the same company, is that "distribution" that triggers 
the associated rights in the GPL?  Or is that internal use that can legally 
be contained by the employer?

Obviously companies can benefit from getting software for zero cost, but 
it's not clear they can benefit much from enhancing that software if giving 
it to their competitors is unavoidable, and creates a competitive advantage 
in the competitor's favor...

> Second assumption: that programming will continue on worthy projects, 
> even after it becomes economically non-viable to produce proprietary 
> commercial software.  People work on libre code right now; I don't see 
> that changing in the years to come.  If a need arises which lacks 
> software and a hobbyist/professional doesn't assume that task, someone 
> will fund the development of that software, just as Sun, AOL, IBM, 
> Compaq, and HP have all contributed to the development of Linux or 
> Apache or Mozilla or Openoffice.  Their reasons for doing this are 
> varied, I'm sure: perhaps they need this software to make their business 
> more viable, or perhaps they consider it R&D, or maybe they wish to 
> hedge their bets against future Microsoft betrayal or failure.  Doesn't 
> matter; it gets done on both large and small scales, every day, right now.

Apache is a special case; it's basically an underground "skunkworks" 
project carried out by webmasters while their superiors remained ignorant 
of what they were doing.  Apache only came to industry attention as a fait 
accompli.  It was already there and better than anything else out there, so 
it garnered official support after the fact, much like VM did with IBM.  
Virtual machines were implemented on IBM mainframes as a skunkworks project 
also; that fait accompli was gained official approval after the fact...

As for Linux, most of the volunteers are there trying to make it better, 
and it did indeed grow from a hobbyist background quite impressively.  The 
companies who support Linux, however, tend to be either (1) seeking to 
profit from all the volunteer labor that went into the software's creation 
(e.g. Red Hat) or (2) seeking to replace proprietary Unix systems to save 
money and/or (3) seeking to disrupt Microsoft's stranglehold on the market 
by switching people from Windows to Linux.

Mozilla and OpenOffice are special cases.  Yes, they appear to be poster 
children for open source at first glance, but there's a catch.  The vast 
majority of the effort spent on these projects came from proprietary 
software companies, AOL/Netscape and Sun/Star Division, in particular.

Like version (3) of the Linux supporters, AOL/Netscape and Sun were willing 
to do this NOT because of their undying belief in free software, but as a 
calculated business risk, intending to disrupt Microsoft's stranglehold on 
the market.  They bore enormous costs to bring these projects to fruition, 
and expect their rewards to come in the form of a weakened Microsoft who 
will no longer be in a position to fling the rest of the computer industry 
around like a rag doll.  The potential benefits of worldwide collaboration 
are entirely secondary.  That's a nice bonus if they can get it, but the 
real goal here is to take down Microsoft.

If Microsoft wasn't a player in either the browser market or the office 
software market, do you REALLY think either of these products would have 
been released, regardless of any arguments Eric Raymond may have made?

If Microsoft is eventually reduced to impotence, the free software efforts 
driven by competitors trying to take Microsoft down may lose all support 
from commercial entities and be left for volunteers to take up the slack.  
At least they'll be starting from a solid base of code, but it's still not 
a foregone conclusion that companies will automatically "step up" to write 
new free software in a GPL-dominated world.

> Third assumption: new industries will spring up supporting GPLd 
> software.  This is also happening, as with Red Hat and IBM both taking 
> support and development roles.  IBM is using Linux to push big iron; Red 
> Hat is proving to be a viable company based on development and support 
> of libre software.

These may yet prove to be niche markets.  It's a bit early to tell.  I'm 
not throwing all my eggs into this basket.  Many business models for open 
source software sound good in theory and don't work so well in practice; 
Transgaming appears to have been burned this way.  This sounds risky.

> For me, there is a fourth item as well, but this one is based more on 
> observation and faith in the market: as long as computers are hard to 
> use, and hard to maintain, there will always be work.  And once 
> computers become easy to use and easy to maintain, there will still be 
> work maintaining and designing systems, networks of systems, data flows, 
> data storage, interfaces, extensions, and enhancements.  There is a lot 
> of life in this chosen field of ours, and miles to go before we sleep. 
>  GPLd software has already made my life easier, and made possible things 
> that would otherwise have been impossible (for me).  I imagine I am not 
> alone.

Well, this is the one assumption where I'm still hopeful, but I'm still 
concerned that we may still be needed to do this hard work, yet end up 
effectively as indentured servants to the dictates of the GPL, should it 
become too successful...

> Anyway, three reasonable assumptions, and a fourth debatable assumption. 
>  These are my assumptions; yours may be different.

I've addressed your assumptions.  Now I'm sure you'll tell me I'm wrong. :-)

Deven