Other LWN stuff:
 Daily Updates
 Calendar
 Linux Stocks Page
 Book reviews
 Penguin Gallery

 Archives/search
 Use LWN headlines
 Advertise here
 Contact us

Recent features:
- RMS Interview
- 2001 Timeline
- O'Reilly Open Source Conference
- OLS 2001
- Gaël Duval
- Kernel Summit
- Singapore Linux Conference
- djbdns

Here is the permanent site for this page.

See also: last week's LWN.

Leading items and editorials

The new draft W3C patent policy is out. This policy was, of course, the subject of a great fuss back in September when it turned out that it allowed patented technology to be included in W3C standards as long as it could be licensed under "reasonable and non-discriminatory" (RAND) terms. Quite a few people felt that the web had flourished as a direct result of having been built on open standards, and they did not welcome the change. The W3C responded by backing off and promising to reconsider the policy.

The result can be seen in the Current Patent Practice note published on January 24. The core of the new policy can be seen at the beginning:

This current practice has evolved in order to satisfy the goal held by a number of W3C Members and significant parts of the larger Web community: that W3C Recommendations should be, as far as possible, implementable on a Royalty-Free basis.

In fact, the policy does not, directly, allow for standards to contain anything but royalty-free technology. Should an "essential" technology arise which is governed by patents, and for which royalty-free licensing is not a possibility, the issue will be sent off to pasture in a "patent advisory group" (PAG). The PAG will have 90 days to figure out how to resolve the issue. One of the possible resolutions is RAND licensing, but the policy cautions:

Note that there is neither clear support amongst the Membership for producing RAND specifications nor a process for doing so. Therefore if a PAG makes a recommendation to proceed on RAND terms, Advisory Committee review and Director's decision will be required.

In other words, those who wish to push RAND-licensed technology into a W3C standard have to wander into uncharted territory, marked only "here be dragons." If there were any further doubt that the W3C would like to be done with this issue, the standard also says that "*It is also possible that a the PAG could recommend that the work be taken to another organization."

To summarize: it looks like we won. Let it not be said that this sort of obnoxiousness can not be overcome by speaking up. Congratulations are due to all of the (many!) people who let the W3C know what they thought when this issue first came up. (For those who have more to say, there is another comment period that will be opened up before the policy is submitted for its final review).

Patents and real-time Linux. Version 24.1.8 of the Real-Time Application Interface (RTAI) was released this week. This release includes a number of interesting new features, including a MIPS port and the latest Linux Trace Toolkit support. Many people, however, may be most interested in the note from FSF attorney Eben Moglen at the end of the announcement. Mr. Moglen addresses the question of whether RTAI users need to be worried about the real-time Linux patent held by Victor Yodaiken (covered in LWN last year).

With regard to RTAI itself, Mr. Moglen states that there is no difficulty in using it within the RTLinux patent license:

The new patent license allows the teaching of the patent to be practiced without any additional obligations in any program released under GPL. If RTAI is released under GPL, as it can and should be, it is fully protected against any infringement claim by the license we negotiated.

Note that the validity of the patent (and associated license) are not being questioned. As long as RTAI fits within the terms of the license, there is no need to look further. To ensure that RTAI is in compliance, the developers have, with this release, relicensed the core code from the LGPL to the full GPL.

Then, there is the question of whether proprietary applications running on top of RTAI infringe the patent. Mr. Moglen refers to the patent claims:

These eleven claims, like all claims in all patents, are written as broadly as possible, so as to bring as much behavior as possible within the scope of the patent. But none of the teachings of the patent, as specified in its claims, is practiced by any applications program running under any OS kernel. No application in a running RTLinux or RTAI system does any of the things the patent claims.

Meaning that:

No applications program is therefore potentially infringing, and no applications program is covered, or needs to be covered, by the license.

Victor Yodaiken's patent, then, should not be an issue for any RTAI users, as long as they comply with the GPL licensing requirements of the RTAI core. This will be good news for these users. This situation is a reminder, however, of the sort of traps that software patents can set for the unwary. The next patent may be more difficult to work around.

When the GPL is not the right license. Ximian announced on January 28 that the license for the Mono class libraries would be changed from the GPL to the X11 license. This license is, perhaps, the simplest free software license available; it says, essentially, "do as you will with this code; just preserve the copyright and don't hold us responsible." There are no restrictions on distribution of proprietary derived products, and no requirement to disclose source for any changes.

This change makes it easier for corporations contributing code to Mono (i.e. Intel and HP) and their customers to distribute proprietary versions of the class libraries. It will, says Ximian, "have the effect of expanding the pool of potential contributors to the project, further speeding its already impressive progress." This statement almost has to be true; it is easy to imagine (though impossible to confirm) that one or more of the corporate contributors insisted on this license change.

The momentum of the GPL is sometimes unstoppable. A look at this week's software announcements shows that the vast majority of projects use the GPL. So, when a project moves in the opposite direction, it attracts attention.

The immediate conclusion that can be drawn is an obvious one: there is no "one size fits all" free software license, and there probably never will be. When developers apply a license to a body of code, they are making a statement on how they want that code to be used. The choice of the GPL indicates a desire to see all versions of the code be free, and to block anybody who wishes to use the code in a proprietary manner. Those who chose a looser license, such as the X11 or BSD licenses, are making a choice to work with proprietary vendors.

That choice is mostly to be made when the resellers are seen as supporters of the code, rather than competitors who would like a proprietary advantage. In the Mono case, Ximian has concluded that the developer base will be increased by the adoption of the X11 license. Those who would sell proprietary products that include the Mono class libraries are not competing with Mono itself; it is in their interest to see Mono thrive.

A similar reasoning was behind last year's Ogg Vorbis licensing change. Allowing the Ogg Vorbis libraries to be used in proprietary products (and things like portable sound file players are likely to stay proprietary) should increase the chances of the Vorbis standard being widely adopted. And that can only help with the development of the code.

As free software works its way deeper into the software economy, it will not be surprising to see more thought given to the choice of licenses. To an extent, this can be a good thing; the right license can help a free software project to thrive. We can only hope, however, that we will not see a resurgence of custom corporate and "half free" licenses, which proliferated a few years ago.

The Linux Standard Base is complete. As LWN went to "press," the Free Standards Group was expected to announce the availability of the Linux Standard Base, version 1.1, for the IA-32 architecture. With this release, the LSB is essentially complete. The specification is in shape, the test suite is available, and there is a sample implementation and cross-development environment. A long development process has come to a conclusion.

As if that weren't enough, the 1.0 release of the "Li18nux" internationalization standard is also about to be released. Here, too, there is a specification, test suite, and sample implementation available.

At this point, these standards lack only one thing: a list of compliant distributions. It is time for that to change. These long-awaited Linux standards will help simplify the management of Linux systems, and they are crucial for independent software vendors who wish to support more than one Linux distribution. Their adoption can only help increase the user and development base for Linux and free software in general. We need LSB- and L18nux-compliant distributions.

The time has come for the Linux distributors to either announce their plans for standards compliance, or to explain why they feel this compliance is no longer necessary. The time for waiting is over.

LWN.net becomes independent once again. In April, 2000, we were pleased to announce that LWN.net had been acquired by Tucows, Inc. It was an optimistic time, and there appeared to be a great many ways in which Tucows and LWN could complement each other.

The times, of course, have changed. Tucows has found other lines of business which are serving the company well, but which have little to do with Linux and free software. Tucows has also found it difficult to earn money from LWN's operation. In these times, it is difficult for a company to carry a loss-making group for any time. For its own future, Tucows must concentrate in the areas where it can bring in revenue - and that's not LWN.

As a result, as of the beginning of February, LWN will operate, once again, as an independent publication of Eklektix, Inc., which will be owned by the current LWN staff. All of our financial issues remain, and they have only gotten more pressing over time, but we will have more freedom in how we try to address those issues. LWN's future is far from guaranteed, but we intend to try our hardest to continue to be a part of the Linux community.

We would like to take this opportunity to offer a whole-hearted "thanks" to the people at Tucows. They have treated us with great courtesy and respect throughout, they have always respected our editorial independence, and, not least of all, they have financed LWN for almost two years. We have no regrets for our decision to join with them. We will miss working with Tucows, but we look forward to watching that company thrive in the coming years.

Meanwhile, expect more announcements in the near future as we work out how we will turn LWN into a sustainable, profitable company. It is the beginning of a new and challenging period, but we intend to meet those challeges and be here for a long time.

There is no history page this week, and probably not again for the forseeable future. We know that some of our readers really enjoyed the "this week in Linux history" feature, but most people didn't manage to get that far. By putting the History page on hold for now, we can better concentrate on the pages that most of our readers enjoy. We do hope to be able to resume this page sometime in the future.

Inside this LWN.net weekly edition:

• Security: Distributor response to security problems; lots of updates.
• Kernel: Patch penguins; rmap, fork, and COW.
• Distributions: Distribution list progress report; New releases from Caldera and Monta Vista.
• Development: Eclipse C/C++ IDE, MySQL 3.23 stable, FreeWRL 0.30 3D browser, KDE 3.0 release plan, Open Motif 2.2, gphoto2-2.0beta4, TinyCobol 0.56.
• Commerce: LinuxWorld press releases from HP, LPI and many more.
• Letters: LWN's birthday; printer drivers.

This Week's LWN was brought to you by:

• Jonathan Corbet, Executive Editor

Next: Security