Subject: Re: Software quality: free vs. proprietary
From: kragen@pobox.com (Kragen Sitaker)
Date: Thu, 28 Oct 1999 12:31:45 -0400 (EDT)

Stephen writes:
> >>>>> "Kragen" == Kragen Sitaker <kragen@pobox.com> writes:
>     Kragen> It looks like it [free software] will do a better job of
>     Kragen> producing good software than the proprietary software
>     Kragen> system.
> 
> On what evidence?

Just observation, so far.  I don't claim theoretical grounds like ESR
does.

> Theoretically, there's no reason to believe that an arbitrarily high
> fraction of the "more eyeballs" effect can't be captured by adopting
> some open source practices while keeping more or less proprietary
> rights.  (This is sort of cheating, since "free" is pretty absolute,
> while proprietary is everything else; but then, when the Aladdin Free
> Public License is defined to be "not free," that's your definition,
> and not my problem.  :-)

Whether it is your problem or not is irrelevant.  Whether it captures
eyeballs is the question.  I believe LPD said on this list a year or
two ago that he didn't get much of that for GhostScript, so
anecdotally, the answer is no.

I don't think eyeballs is the whole story, though.  The potential for
forking and the lack of the churn motive (ship something NOW to get
money, then fix the bugs in three releases that each give us more
money) seem likely to be important.

GNOME and KDE have the churn problem, and despite their high eyeball
levels, they are of notably lower quality than most free software.

> Pragmatically, remember that current levels of quality are based on
> taillight-chasing (and therefore design bugs are less of a problem)
> and cherry-picking 

4.2BSD, Emacs, gcc, and EROS are neither taillight-chasing nor
cherry-picking.  The GIMP was, but isn't now.  KDE, GNOME, AbiWord,
Linux, and glibc are clearly doing a lot of both.

But even if you were right about t-c&c-p, Linux seems to have done a
much better job of t-c&c-p than, say, AIX.

> (it's reasonable to suppose that free software has
> entered those parts of the industry where it has a comparative
> advantage, or anyway least disadvantage, vis-a-vis the proprietary
> model---thus it is going to be hard to maintain current levels of
> efficiency and quality as the domain of free software spreads).

It is reasonable.  It is also reasonable to suppose that this is not
the case.  Do you have some evidence one way or the other?

> If I'm misunderstanding you, and you are contemplating a mixed system,
> then it's not clear to me whether the free and proprietary sectors are
> comparable on quality in an interesting way.  "Hello, world," as
> presented in K&R, is provably bug-free (on the GNU/Linux system,
> anyway ;-).  Surely it is higher quality than Windows 98 in that
> sense, but who cares?

Well, how about the 'fuzz' study?  That compared several
implementations of the Unix toolset on quality in a measurable way:
what percentage of them crashed or hung when fed random data.  The GNU
and Linux versions were head and shoulders above the proprietary
implementations.

> You would have to compare head-to-head on comparable projects, but it
> seems likely to me that where free software succeeds, proprietary
> software will abandon the field.

Well, I hope so.  :)

> Where proprietary software succeeds,
> free software may continue development,

By "where proprietary software succeeds", do you mean "where
proprietary software does not abandon the field because free software
has not yet succeeded"?

> but evidently in the eyes of
> consumers free plus bug-free (by hypothesis) does not outweight
> whatever it is that the proprietary software is offering, presumably
> at a higher monetary price, besides.  So what's the relevant quality
> measure there?

I am not clear on what you are saying.  Are you saying that if free
software is less buggy, but people still use proprietary software, then
it is because the proprietary software offers them something they want
that the free software does not?

I do not agree with that.  Here are some other possibilities:
- people don't switch because they don't know about the free
  alternative.  It has been apparent for a long time that free software
  currently has some major difficulty getting publicity, with many
  worthwhile and useful projects remaining almost completely unknown --
  even inside the free-software community.
- people don't switch because they know less about the free software
  than the proprietary software.
- people don't switch because they don't like change.  (Many old bugs
  may be better than a few new bugs.)
- people don't switch because they know a lot about the free software,
  but it's wrong.
- people don't switch because other people don't want them to, such as
  upper management or the IS department.  Many large companies have made
  special deals with Microsoft to use Windows exclusively, for example.

And if it *is* because the proprietary software offers them something
they want, what they want may not be quality:
- people don't switch because spending lots of money gets them more
  power in their organization.
- people don't switch because frequently fixing things when they break
  gives them more visibility and prestige.
- people don't switch because they want to fail.
- people don't switch because they want to complain.
- people don't switch because they don't think they're technical enough.

All speculation, of course.

>     Kragen> Perhaps we can come up with a
>     Kragen> military-industrial-university system for software as we
>     Kragen> have for science.
> 
> Heaven forbid, on efficiency grounds (not ethical ones, I put that hat 
> away).  (Wisecrack only, I do not plan to defend it.)

Heh.  :)

-- 
<kragen@pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Thu Oct 28 1999
12 days until the Internet stock bubble bursts on Monday, 1999-11-08.
<URL:http://www.pobox.com/~kragen/bubble.html>