[LINK] [Fwd: [Aus Innovate] Misconceptions in Support for Op en Source Software]

[Chirgwin, Richard]

Oh, what the hell, it's lunchtime. Tony H, if you read this, keep in mind
that I am a firm believer that all debate, on both sides, have to be founded
on clear, accurate fact.

> In fact, there is already extensive software development by
> Australian companies and individuals. What's more, those companies and
> people would be harmed, not assisted, by open source.

The first statement is true and can be demonstrated by research; the second,
however, is not backed by hard data; it is merely an assertion of opinion.

> Open source software is based on software being free, which 
> means developers
> receive no revenue.

That is not so, Tony. It is based on software designs being made public;
they may be sold, but not obscured. 

However, the conclusions you draw are partly in error:
> because other people produce rival
> programs using that source code, or modify it and pretend it 
> was their own
> work. In this sense, software is different from all other 
> copyrighted works.

Copyright can still subsist in open source software: in fact, its main role
in this context (as I understand it) is to protect authors from "passing
off". My work must be acknowledged as my work, even if it's being
redistributed for free.

This is a model borrowed from other disciplines like architecture and civil
engineering, where copyright is used as a persistent assertion of
authorship. "Passing off" someone else's work as your own, or passing off an
inferior work as the output of someone else, are both actionable even under
a license such as the GPL. Nor is Australian copyright law rendered
ineffective as an assertion of ownership, merely because of the license
under which an author publishes the software.

> because revenue is
> essential for all serious practitioners in the economy, 
> including software
> developers.

No argument. However, this statement does not support the theorem that
"there is only one possible business model for software production".

> Open source explicitly relies on the free labour of 
> programmers. 

Yes. But this is a willing donation. The conclusion thyat it is therefore
exploitative is not one which can be automatically drawn.

Number (3) is non-sequitur to open source pro/con. I note that local
companies have developed operating systems, but why bother? - it's
impossible to market against Microsoft. 

> 4. Open Source Would Favour Commodity Firms
Tony, I don't quite know where to start on this one, because there are too
many unsupported assumptions.

> Government adoption of open source would favour less capable 
> firms that use
> free junk from the web instead of developing first class 
> software. 

Component models are strongly pro-quality, as you know. And outside the
world of marketing budgets, the best components propagate most easily; for
example, look at the world of PC-based music recording and production, where
there exists a great many nearly-identical components, but any given
component class "thins" very, very rapidly.

> As well, firms and organisations that use a lot of open 
> source software
> typically have no respect for the amount of work and 
> expertise required to
> develop advanced applications, 

Such a statement can only stand if it is supported by hard, researched
evidence.

> Those competitors include foreign
> outsourcers, who will gain hefty consulting and support fees 
> from any switch
> to Linux.

This appears to contradict the notion that there's no money in Open Source.

> For Australia, these are serious dangers, because our 
> technology industries
> depend on exporting software; we do not have the 
> manufacturing and design
> base to design and export computers.

There's a serious lack of hard data either for or against the first part of
this statement. It's well-nigh impossible to accurately measure the export
dollars attributable to Australian software development.

Regarding the second: certainly, we have no consumer PC manufacturing.
However, open source is definitely a plus for the hardware industries which
exist here; industrial control, niche telecommunications systems, and
defence electronics are all favoured by an open source model.

It's true to say that "hardware manufacturers support initiatives which help
them sell more hardware". But that's irrespective of whether we're talking
about the open source model or the 'proprietary' model.

> 5. "Many Eyes" Means Nothing If They're No Good
I am not sufficiently technical, nor sufficiently familiar with the
literature, to examine this section. I also suggest that simply flaming this
section would do no good.

I would however observe that:
> More importantly, developing a modern desktop operating 
> system is a huge
> task, like designing a 747 aircraft, and is thus not a 
> feasible prospect for
> Australian industry. 
(From section (3))
and
> Linux, as a development 
> task, being an
> operating system, is relatively straightforward. Operating 
> systems are well
> defined and understood.
 
...seem to lack internal consistency.

Section (7) relies on different views of democracy, and I don't propose to
get into a flamewar about government openness.

> 8. Protection of Source Code is the Only Feasible Way of Protecting
> Copyright in Software

> Copied movies, books and articles can't be provided in public 
> without their
> origin being obvious, thus preventing widespread pirating. 

Not so. Plagarism is frequently practised and infrequently detected. 

> The source code used to build a product is not visible 
> in the final
> product, so freeloaders can claim it to be their own work.

As has been asserted of proprietary software - that it has, on occasion,
"borrowed" from the open source community functions which are then obscured
by the lack of source code? So I'm forced to conclude that this section is
non-sequitur - it can be considered of equal weight on either side of the
argument.

> the original product, and would not otherwise have been 
> created, it is still
> not certain that the copying will be provable. 

There is nothing to stop an open source developer, if he or she so desires,
from lodging an escrow copy; this would be prudent to inhibit "passing off".

Richard Chirgwin

> -----Original Message-----
> From: Brenda Aynsley [mailto:bpa at iss.net.au]
> Sent: Tuesday, 8 July 2003 13:07
> To: link
> Subject: [LINK] [Fwd: [Aus Innovate] Misconceptions in 
> Support for Open
> Source Software]
> 
> 
> Linkers
> 
> This argument has been propounded by Tony Healy, a person 
> known to have strong 
> views on the issue.  Is anyone/s prepared to provide a rebuttal?
> 
> cheers
> brenda
> 
> -------- Original Message --------
> Subject: [Aus Innovate] Misconceptions in Support for Open 
> Source Software
> Date: Tue, 08 Jul 2003 13:00:16 +1000
> From: Tony Healy <thealy at siliconchic.com>
> Reply-To: thealy at siliconchic.com
> To: Aus Innovate <aus-innovate at siliconchic.com>
> 
> 
> 		Misconceptions in Support for Open Source Software
> 			A position paper
> 		Tony Healy, Aus Innovate, 08 July 2003
> 
> 
> ALP and Democrat parliamentarians are currently preparing 
> policies that aim
> to mandate open source software for government. The drafters of those
> policies represent them as boosting the Australian software industry.
> 
> In this paper, I argue that the alleged benefits of open 
> source software are
> greatly exaggerated by proponents, that software development 
> as a discipline
> and industry is not understood by parliamentarians, and that 
> open source
> embodies concepts that conflict with core ideals of Democrats 
> and the ALP.
> 
> 
> 1. Open Source Won't Further Australian Software
> ------------------------------------------------
> 
> Much advocacy seems to presume that open source is the only 
> alternative to
> Microsoft. In fact, there is already extensive software development by
> Australian companies and individuals. What's more, those companies and
> people would be harmed, not assisted, by open source.
> 
> Open source software is based on software being free, which 
> means developers
> receive no revenue. Open source advocates dispute this, but 
> the fact is
> that, once the source code is publicly available for a product, it is
> difficult to charge for software, because other people produce rival
> programs using that source code, or modify it and pretend it 
> was their own
> work. In this sense, software is different from all other 
> copyrighted works.
> I discuss this further in Section 8.
> 
> Any concerted move to open source would kill innovation, 
> because revenue is
> essential for all serious practitioners in the economy, 
> including software
> developers. It's worth noting that Linux creator Linus 
> Torvalds created a
> product worth billions of dollars, yet still has to scrabble 
> for a job. This
> is not a model we want for Australian software.
> 
> 
> 2. It's Inconsistent With Concern for Workers
> ---------------------------------------------
> 
> Open source explicitly relies on the free labour of 
> programmers. In other
> arenas where employers, government or business attempts to 
> exploit workers,
> especially naive young people, socially committed people such 
> as the ALP and
> Democrats generally condemn the employers. Open source 
> actually is the same
> concept.
> 
> 
> 3. The Desktop is a Platform, Not an Opportunity
> ------------------------------------------------
> 
> There seems to be a belief that displacing Windows would create
> opportunities for Australian developers. In fact, it wouldn't. The
> alternatives intended by open source advocates already exist and are
> nominally free. There would be no or minimal development 
> opportunities for
> Australian developers.
> 
> More importantly, developing a modern desktop operating 
> system is a huge
> task, like designing a 747 aircraft, and is thus not a 
> feasible prospect for
> Australian industry. It would cost a fortune and provide no marketing
> benefit over existing products. The future for Australian 
> firms is designing
> and developing specialist software.
> 
> 
> 4. Open Source Would Favour Commodity Firms
> -------------------------------------------
> 
> Government adoption of open source would favour less capable 
> firms that use
> free junk from the web instead of developing first class 
> software. Open
> source advocates typically cite a handful of flagship 
> products that work
> well, but there is a vast array of other tasks that software 
> must cater for.
> In those other fields, free software typically sucks. Examples are
> discussion forums that mix up posts from different years, 
> because they only
> sort on months.
> 
> As well, firms and organisations that use a lot of open 
> source software
> typically have no respect for the amount of work and 
> expertise required to
> develop advanced applications, and will refuse to pay fair 
> rates for that
> work, even where there are no free products that provide the 
> functionality
> they need. This a cargo-cult mentality.
> 
> 
> 5. "Many Eyes" Means Nothing If They're No Good
> -----------------------------------------------
> 
> The fact that source code is open to inspection really means 
> nothing in
> itself. Designing and fixing software, especially non-trivial 
> applications,
> requires skilled software engineers with the time to properly 
> analyse the
> whole design before making slap-dash changes. Random local 
> modifications are
> a recipe for disaster.
> 
> Promotion of open source as a development methodology usually 
> hinges on a
> few flagship products, typically Linux and Apache. However a 
> true assessment
> should compare the quality of the many thousands of other open source
> products against the tens of thousands of commercial applications.
> 
> Most open source products are amateur projects, with high bug 
> counts, clumsy
> operation and installation, excessive dependencies and unfinished
> functionality. For a list, see www.sourceforge.net. Netscape's Mozilla
> browser project was open sourced, and became one of the 
> biggest disasters in
> software history, running years behind schedule and leaving 
> Netscape without
> any competing product just when Microsoft caught up with its Internet
> Explorer browser.
> 
> A recent study by Oxford University theoretical physicists 
> Damien Challet
> and Yann Le Du purports to prove that open source development 
> fixes bugs
> faster than professional development[1]. However it uses 
> statistics from
> Linux archive files, essentially invalidating the study as a 
> study of open
> source methodology in general. Also, Linux, as a development 
> task, being an
> operating system, is relatively straightforward. Operating 
> systems are well
> defined and understood. There is not the same amount of 
> conceptual design
> required as for most greenfields software projects.
> 
> Also, it's worth bearing in mind that the popular notion in 
> open source of
> being able to fix things yourself is necessary because of the 
> dodgy nature
> of so many open source products in the first place.
> 
> 
> 6. Hidden Corporate Agendas
> ---------------------------
> 
> Open source advocates like to believe they're attacking big 
> business, but
> they're actually pushing an agenda that suits some elements 
> of big business.
> Competitors to Microsoft stand to gain handsomely from open 
> source, and are
> actively funding open source PR. Those competitors include foreign
> outsourcers, who will gain hefty consulting and support fees 
> from any switch
> to Linux.
> 
> Another class of competitor are hardware vendors such as IBM and Sun.
> Hardware vendors have a strategic interest in making software 
> and software
> development cheap, because it creates a need for more 
> computers. Software
> developers suffer, but computer makers benefit. This was part of the
> motivation behind Sun creating Java and it's why IBM funds a 
> lot of open
> source development.
> 
> For Australia, these are serious dangers, because our 
> technology industries
> depend on exporting software; we do not have the 
> manufacturing and design
> base to design and export computers.
> 
> 
> 7. Being Able to Verify Operation of the Software is a Red Herring
> ------------------------------------------------------------------
> 
> Lawyers, ever alert to more sources of fees, have recently joined open
> source advocates in claiming that government should use open 
> source software
> in order that government can verify the operation of that software.
> 
> However government routinely trusts other firms, including lawyers,
> accountants, engineers and PR firms, so it's difficult to see why it
> shouldn't also trust software firms. To do otherwise is a 
> form of cultural
> apartheid. Further, even if government had the source code, 
> it would still
> have to trust software engineers to assess and report on the 
> source code.
> Those software engineers would probably be in the employ of external
> auditing firms. There's no reason government should trust 
> accountants but
> not software firms.
> 
> 
> 8. Protection of Source Code is the Only Feasible Way of Protecting
> Copyright in Software
> --------------------------------------------------------------
> --------------
> -------------
> 
> Copied movies, books and articles can't be provided in public 
> without their
> origin being obvious, thus preventing widespread pirating. 
> With software
> though, once the source code is made available, freeloaders 
> can take that
> source code and build similar programs without doing all the 
> development
> work. The source code used to build a product is not visible 
> in the final
> product, so freeloaders can claim it to be their own work.
> 
> Even if expensive legal action is initiated to examine the 
> source code of
> the new product, and even if that source code was indeed 
> copied from that of
> the original product, and would not otherwise have been 
> created, it is still
> not certain that the copying will be provable. The copied 
> source code, while
> mimicking the important concepts and architectures in the 
> original,  might
> have different names and layout. It is for these reasons that software
> developers retain their important blue prints, or source 
> code, as a way of
> protecting their copyright and thus being able to carry on a business.
> 
> 
> 1. Damien Challet and Yann Le Du: Closed source versus open 
> source in a
> model of software bug dynamics, Submitted to Elsevier 
> Science, June 2003
> http://arxiv.org/abs/cond-mat/0306511
> 
> 
> 
> Regards, Tony Healy
> 
> -----------------------------------------------------------------
> Aus-Innovate, Sydney
> 
> Aus-Innovate is interested in the development of Australian R&D and
> software, and the furthering of professional software
> engineering interests, including workplace issues. If you 
> would prefer not
> to receive any further mailings like this, please tell me.
> 
> Ph (02) 9360 0988
> thealy at siliconchic.com
> 
> -- 
> Brenda Aynsley   ||  Mobile:+61(0) 412 662 988
> BA, DipSocSci(sociology), DipAppSci(computing), MACS, PCP
> Vice President Australian Computer Society
> http://www.acs.org.au/
> --
> Oz Business Partners - helping you do better business
> Web Design and Implementation : Technical Writing for Effect
> Multimedia Production : Training Development and Delivery
>     http://www.ozbusinesspartners.com/ ACN 069 346 327
>     Phone: +61(0) 8 8357 8844  Fax: +61(0) 8 8272 7486
> 
> _______________________________________________
> Link mailing list
> Link at mailman.anu.edu.au
> http://mailman.anu.edu.au/mailman/listinfo/link
>