Club Linux Coffs Harbour

Long ago, I used to read a magazine (no longer in print; a victory of the Web over dead tree publication) which carried Matt Groening's "Life in Hell" comic strip. At the end of each year, it was customary (I don't know if it still is) for Matt to list the "Forbidden Words" for the following year; terms relating to the fads, scandals, cause celebres, and all the cringe-inducing cliches of the previous twelve months.

Carrying on this tradition (good artists copy, great artists steal, bloggers carry on traditions), here's my provocatively heretical list for 2007, with accompanying ranting. Add your own in the comments.

• "Commercial" Software
• "Ready for the Desktop"
• Eric Raymond
• Open Source
• Linux
  

"Commercial" Software

I've racked my brains trying to recall a single piece of computer software which, all else being equal, could not conceivably be used in some commercial context. I can't think of an example. You might use or distribute any particular piece of software in a commercial or non-commercial manner, but the "commerciality" is a consequence of what you're doing with the software, not inherent in the software itself. As a descriptive term, it is therefore redundant.

As a propaganda term however, the word "commercial" is often used in place of "proprietary" to disguise the reality of the the terms under which you are able to use the software. Commerce is generally regarded as a good thing for society, therefore anything "commercial" must be also good for society, right?

So if we can say "commercial equals good", then in order to say "proprietary equals good" all we need do is demonstrate that "proprietary equals commercial". Let's try, shall we?

Proprietary software uses the copyright system to impose a monopoly on the publication and sale of the software. One person, typically a single corporation (corporations are persons in the eyes of the law), has the right to say who may or may not copy, sell, or otherwise redistribute the software. Free software waives these restrictions. Ask any economist, and they'll tell you that monopolies, while they may be justified in particular instances as a form of government intervention in markets to achieve certain social ends, are in general anything but "commercial". Monopolies restrict the total amount of commerce that may take place, and are generally held to make markets function less efficiently. Free software is in this sense more commercial than proprietary software (and more economically efficient, according to conventional economic wisdom).

Copyright also prevents you from creating and distributing software based on pre-existing proprietary software. Free software permits the creation and distribution of derivative works, thereby allowing innovation and the creation of new software, and hence the potential for more commercial activity.

Most proprietary software furthermore uses the contentious mechanism of End User License Agreements to restrict not only copying and redistribution beyond the restrictions set by copyright law, but also to restrict the ways in which you may use the software. If you use this software in a commercial context, this is in effect a restriction on how you can carry out your commercial activity with this software. Free software licenses in contrast, guarantee your freedom to use the software for any purpose.

End User Licence Agreements also mean that you are not buying a copy of a piece of proprietary software, you are buying a license to use that software. It is therefore not the software, but the license, that is commercial, in the sense of being something that may be bought and sold.

Unfortunately, it is not only the propagandists in the proprietary software industry who have been misusing the word "commercial" in this way. There is a marked tendency for people in the free software world to refer to proprietary software as "commercial software", implying that free software is somehow not commercial when in fact the very opposite is the case.

In 2006 for example, Canonical introduced a new "commercial" software package repository for Ubuntu. The content of this repository is of course software that you can't legally sell without permission from the copyright holder, and which comes with all the above restrictions on commercial use, unlike the free software which comprises (most of) the rest of the Ubuntu operating system. Canonical's "commercial" repository is in fact strongly anti-commercial, and a free gift to proprietary software propagandists.

"Ready for the Desktop"

Here is a fool-proof recipe for an article for any lazy tech pundits on a slow news week:

Is <insert name of free software product here> Ready for the Desktop?

Exactly what is supposed to be ready for whose desktop? How do you define ready? How do you define desktop? The trick is you don't, and hey presto you've got a FUD-filled article that writes itself.

I lost the time and inclination to play epic $100 computer games for Microsoft Windows about five years ago, and therefore stopped using any proprietary operating system on a regular basis. Does that mean Debian GNU/Linux was "ready for the desktop" five years ago? It was certainly ready for my desktop.

In fact long before then, and still today, there were things I could do with Debian that I couldn't do with the same ease, or at all, with a Windows or Mac system. Does that mean that Microsoft and Apple should be in a frantic panic to make themselves "ready for the desktop"?

No, of course not. There is some proprietary software which has no exactly comparable free software equivalent, and there is some free software which has no proprietary equivalent. The free software community leads in some areas and is playing catch-up in others. There is no reason to expect this will cease to be the case any time soon.

The real story that the "ready for the desktop" stories obscure is this:

In 2007 most people can have most of their software needs met by free software.

Vendor lock-in and marginal cases involving niche applications will persist for many years to come, but a significant number of people are now able to get by quite happily with no proprietary software at all. Whether they will exercise that ability is another matter, unrelated to the question of "readiness" on the part of the software.

Eric Raymond

Exactly why Eric Raymond gets the amount of press attention that he does is rather hard to fathom. Nearly a decade after splitting the free software movement into two factions (see below), his achievements have since been, perhaps thankfully, rather thin on the ground.

More puzzling are the lables that tech journalists attach to him that bear no relation to anything he says or does. The man persistently credited as an "open source guru" and "a champion of all things open" has spent most of the last year urging people to use proprietary software, culminating in his appointment to the board of Freespire, a GNU/Linux-based operating system with a larger than usual amount of proprietary components. The year before, he even declared that the GNU General Public License, the most prevalent free software license, and more specifically the "copyleft" provisions of the license which guarantee the continuing freedom of derivative works, are somehow unnecessary.

It would be more accurate to call him proprietary software's evangelist to the free and open source software communities, but I suppose such a title would render him less effective in that role.

Open Source

"Open Source" has been considered by different people at different times to be:

• an alternative term for "free software"
• a replacement term for "free software"
• a parallel movement to the free software movement
• a replacement movement for the free software movement

On it's introduction, the term was justified as being:

• less ambiguous than "free software", and
• less threatening to "corporate types",

and therefore merely a more effective way to sell free software to a particular audience.

The only ambiguity in "free software" lies in the dual meaning of "free", which is not an issue in most languages, which have the luxury of separate words for these two meanings ("libre" and "gratis", for example). To an English-speaking audience, once you have made it clear which meaning of the word "free" you are using, the ambiguity is gone.

The term "open source" on the other hand is not merely ambiguous, but so vague as to be almost meaningless. "Open" may indeed be intended to mean "free", but it can be easily be taken to mean anything from "free as in libre" to merely "open to scrutiny under certain limited conditions", and indeed many proprietary software vendors have misleadingly used the word in this way to imply some connection with the free software and open source communities. And exactly how introducing the concept of source code to the terminology is meant to make it more accessible and less threatening a non-technical audience is left unexplained.

If "open source" was indeed intended to be merely a re-branding of "free software" for certain audiences, why do we need an entirely new organisation to promote it, when the Free Software Foundation already exists? Why draft an Open Source Definition when the Free Software Definition already exists? Why couldn't you just you say a program is open source if it meets the criteria in the Free Software Definition? Perhaps because the intention from the start (among some people at least) has not been to make free software more acceptable to a wider audience, but to divorce the software from the ethical principles that made it possible, in order to re-integrate it into a system where code and the ideas expressed in code can be owned. The oft-expressed distaste for GPL-style "copyleft" licensing among many proponents of "open source" is significant.

Under the leadership of certain individuals (see above) who have no particular concern for the principles of the free software community, "open source" has been used as a trojan horse to undermine the free software movement from within. Whether or not the term ever had any use in popularising free software, we should now accept that it has been given more than enough time to do so, and move on.

In 1999, Open Source Initiative co-founder Bruce Perens declared that "It's Time to Talk about Free Software Again". In 2007, it is well past time.

Linux

To clarify, there is no problem with applying the term "Linux" to the operating system kernel software of that name.

However, there is not, nor has there ever been, an operating system called "Linux". The operating system kernel called "Linux", is used by a great many different operating systems. Lumping all these different operating systems together under the same name is undesirable for a number of reasons:

It's not factually correct

The Linux project has never set out to create or assemble all the software needed to form a complete operating system. There is a project with that aim, the GNU project. Since 1983 the GNU project (and later the Free Software Foundation) have been working with the aim of providing, either by writing from scratch or otherwise acquiring from third parties, all the software necessary for anybody to use their computer with complete freedom.

Linux was the last part of this collection of software necessary to create a minimal working operating system. For this reason the technically correct collective term for the systems which are commonly called Linux, composed of the GNU system and the Linux kernel, is GNU/Linux. This term accurately describes these systems, whilst excluding GNU-based systems which use other kernels (Debian GNU/NetBSD, GNU-Darwin, etc.), and Linux-based systems which do not use the rest of the GNU system (embedded operating systems for special-purpose devices, for example).

Over time technology advances and what users expect from their computers changes, so the GNU project continues to work to ensure that all users can do what they need in freedom. For example, for a long time there was no complete modern graphical 'desktop' interface available for GNU/Linux systems under free licensing terms, so the GNU project started the GNOME sub-project. The Free Software Foundation continues to maintain a list of high priority free software projects.

It confuses new users

Every convert to a free operating system I have ever found who has lapsed back to a proprietary system has done so because "Linux is too hard". This person has either tried to:

• install some third party software which is "for Linux", or
• follow some documentation "for Linux"

... but not for their particular Linux-based operating system, and consequently come to grief. This would not have happened if they weren't under the impression that there was an operating system called "Linux", but rather that there are many operating systems which use the Linux kernel.

It leads to pointless arguments

Hot on the heels of "not ready for the desktop" declarations are endless analyses of what's wrong with Linux, or what Linux "needs".

Often heard is the call to address "fragmentation" of Linux. This is tantamount to demanding that people should not be free to use the Linux kernel in an operating system unless the system they create conforms to certain strict guidelines. Unless you are under the misapprehension that Linux is the operating system as a whole rather than a small part of many different systems, this demand for suppression of non-conforming operating systems is absurd, and contrary to the principles of freedom which lead to the creation of these systems.

It's FUD-fuel

Whenever a security flaw is found in an operating system that uses the Linux kernel, no matter how marginal this particular system is, it's reported as a bug in the mythical "Linux" operating system, tainting in the public mind all operating systems that use the Linux kernel regardless of whether they are affected by the same flaw.

The popular myth that the "Linux operating system" was created by a single Finnish student on a whim is as dangerous as it is ridiculous. The myth has been used to bolster claims that "Linux" must be built on wholesale copyright infringement (notably in the frivoulous lawsuits brought by the SCO Group), because obviously one person can't write an operating system all by himself.

The truth is that Linus Torvalds started work on an operating system kernel in 1991 (eight years after the start of the GNU project), announcing to the world that bash and gcc (GNU software) seemed to work with it, and asking for suggestions on where to go next with his creation, which "won't be big and professional like gnu" [sic]. However, by publishing his work under the GNU General Public License, it was able to be rapidly improved by a growing group of contributors worldwide, and it soon became adopted by the GNU project as the official kernel of the GNU operating system. This lone Finnish student did not single-handedly create an operating system. He started a project involving many people to create a small but vital component of an operating system; a component that subsequently became widely-used, including in the GNU system and most of it's many derivatives.