User: Password:
Subscribe / Log in / New account

Leading items

The USB that ate Linux

Robert X. Cringely has reported on a new threat to Linux: a Microsoft-driven version of the USB standard which will not be usable by Linux. The article is rather short on details, but the idea seems to be that only "trusted" USB devices could be written to, and the mechanism for identifying and communicating with these devices would be closed. You'll be able to install Linux on your future motherboard, but it will not be able to work with the new USB devices.

This sort of story comes around fairly regularly. Long-time LWN readers will remember some past worries:

  • Once upon a time, the "Merced" architecture from Intel was to be the future of computing. Unfortunately, Merced was under nondisclosure, and, in any case, getting gcc to generate code for that architecture was said to be beyond the capabilities of its developers. In the reality, Merced, later named Itanium, had top-quality Linux support from the beginning. We're still waiting for the "future of computing" part, though.

  • The I2O specification was kept under wraps for some time, and it looked like Linux would be unable to drive any I2O-based hardware. Richard Stallman called I2O "a broad plan to keep hardware specifications secret". As it turned out, the specifications were released, and Linux supports I2O without trouble.

In other words, we have seen this sort of thing before. Fears of Linux-killer hardware turned out to be misplaced even in the 1990's, when Linux was a far smaller commercial force than it is now. In the current climate, it is hard to imagine the hardware companies adopting a fundamental technology (a processor or bus architecture, say) that was deliberately closed to non-Microsoft operating systems. Not all vendors rush out to embrace Linux, BSD, and MacOS users, but few will see a business case in explicitly excluding them. Especially if that exclusion would consolidate the position of a company which has not always distinguished itself with its considerate treatment of its "partners."

On the other hand, proprietary hardware and digital restrictions management schemes do bear watching. The troubles Linux has had with playing DVDs have been well documented. The "broadcast flag" will restrict the ability of Linux systems to work with digital radio receivers in the future. "Trusted computing" schemes may keep Linux off some hardware altogether. There are threats out there, but an exclusionary USB specification is probably not one of them. Nobody besides Cringely seems to know much about this new USB standard, however, and the Linux USB developers are not particularly worried about it. For the time being, the rest of us probably need not worry either.

Comments (7 posted)

What's coming in Fedora Core 3

September 22, 2004

This article was contributed by Joe 'Zonker' Brockmeier.

The final release of Fedora Core 3 isn't expected until November 1, but with the release of Fedora Core 3 test 2 (FC3t2) on Monday (a week later than originally planned) we decided to check in and see what users could expect from the next release of Fedora Core. We also contacted Red Hat to see if Cristian Gafton or another representative would be available to talk about Fedora, and its relation to Red Hat's commercial products, but they were unable to provide a representative to speak to LWN by deadline.

This release marks the addition of the GNOME 2.8 release candidate, KDE 3.3.0, 6.8.0, and the udev device model.

We gave FC3t2 a try on an Athlon XP 2000 system with 1 GB of RAM. One thing we noticed was that the media check failed all of the disks we burned, but we were able to perform an install from the media without any problems. This seems to be an issue that came up during tests of FC3 test 1 as well. While bugs and glitches are to be expected in test releases, we note this particular issue so that users trying out FC3t2 do not burn through a stack of blanks in a futile attempt to burn four good disks.

Users will find that the default partitioning has changed a bit since Fedora Core 2. By default, the installer will attempt to set up LVM rather than the standard "simple" partitioning most Linux users are used to. There seem to be a few bugs left in the partitioning tool, as the installer informed us we were "probably out of disk space" when attempting to install. However, when we performed an install using a standard partitioning scheme, all went well. No doubt, this will be ironed out by the time that FC3 final is released.

Fedora Core 3 marks the Fedora team's second stab at SELinux, and they are asking that users give SELinux another try as well. According to Colin Walters, this release marks a scaled-back approach that should cause fewer problems while still providing additional security for "select system daemons."

Instead of the original "strict" policy which covered everything, a new "targeted" policy has been developed which only applies SELinux restrictions to a few select system daemons. Regular user login sessions are unrestricted.

The initial approach to SELinux was probably a too-radical departure for many users, so we're happy to see the Fedora team taking a more moderate approach that will (we hope) build support for SELinux over time.

However, the actual documentation and tools for SELinux leave a bit to be desired, as Matias Feliciano points out on the fedora-devel list. While the "targeted" policy is "mostly invisible" to the end-user, so is the documentation for users who want to customize and tweak their SELinux policy.

FC3t2 marks the introduction of the udev device model to Fedora. The udev device model implements devfs in userspace, creating a dynamic /dev that allows consistent naming of devices. Users upgrading from test 1 or installing udev on test 1 reported a few bugs, but we didn't see any problems with udev from a clean install.

Despite the occasional glitch in the test release, FC3 is shaping up nicely. It's not a radical change from FC2, most of the changes are package upgrades and further refinement of existing features. The udev device model is probably one of the most major changes that users will see in FC3.

It bears mentioning that the Fedora Core development process still seems to be shy on community involvement. However, Red Hat and the Fedora team have provided a usable Linux distribution with many of the cutting-edge technologies that users want to try. From that perspective, we think that Fedora has become a success.

Comments (4 posted)

Reforming WIPO

There is a movement afoot, initially pushed by Brazil and Argentina, to change the mission of the World Intellectual Property Organization (WIPO). An information page is available. There is also the text of a declaration (PDF) which will be debated in Geneva on September 30. "As an intergovernmental organization, however, WIPO embraced a culture of creating and expanding monopoly privileges, often without regard to consequences. The continuous expansion of these privileges and their enforcement mechanisms has led to grave social and economic costs, and has hampered and threatened other important systems of creativity and innovation.... We do not ask that WIPO abandon efforts to promote the appropriate protection of intellectual property, or abandon all efforts to harmonize or improve these laws. But we insist that WIPO work from the broader framework described in the 1974 agreement with the UN, and to take a more balanced and realistic view of the social benefits and costs of intellectual property rights as a tool, but not the only tool, for supporting creativity [and] intellectual activity."

Comments (3 posted)

LWN update

Occasionally we get a message noting that we have not been posting "LWN update" articles, and wondering how things are going. We are still trying to keep a lid on such articles, but we're about to hit an important anniversary. It is now two years since we began the subscription experiment, so the time seems right for a look at how things are going.

Our goal at the outset was 4,000 subscribers. As of this writing, LWN has just under 3,300 active, paid subscriptions - up from about 2,700 at this time last year. Things are clearly headed in the right direction, even if they are not yet where we would like them to be. The next big test will be to see what happens over the next month as the "great expiration" sets in. We got a big group of subscribers right at the beginning, and many of their subscriptions will expire (again) in the next few weeks. Last year's "great renewal" brought in enough cash to see through the slow parts of the year (we're sure glad we hung onto it at the beginning); with luck that will happen again. Our subscription renewal rate tends to be quite high, and you can be sure that we are grateful for it.

We're looking to add more new subscribers, of course. The external authors program has helped to fill out our content, but LWN could really benefit from another editor who could write original content and provide a bit of redundancy. We will continue to work to find those subscribers; going out and marketing LWN to new readers has proved to be a challenge, however.

Meanwhile, we plan to continue to do our best to provide top-quality, comprehensive coverage of the Linux and free software community. Many thanks for your continued support; it is a pleasure to write for this group of readers.

[As an aside: we have noted for a while a certain number of people creating accounts without giving us working email addresses, then trying to sign up for our mailing lists. That is clearly not going to work. If you do not get the mail you expect, please try going into the My Account area and making sure we're sending it somewhere useful.]

Comments (17 posted)

Page editor: Jonathan Corbet
Next page: Security>>

Copyright © 2004, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds