The Gazette grew quickly, attracting new readers and new authors. By the fifth issue, mirrors were necessary; these were contributed by Phil Hughes (of SSC, the publisher of the Linux Journal) and Alan Cox. Putting out the Linux Gazette took time, however, and the project lapsed for some months in early 1996. When Issue 8 came out in August of that year, it carried the following announcement:
Once it came under the Linux Journal's wing, the Linux Gazette thrived. Over 80 issues were produced, on an approximately monthly basis, and the range of authors and topics seemed to increase every month. The Linux Gazette carried early articles by a number of well known community authors, including Joe Barr, Miguel de Icaza Chris DiBona, Jon 'maddog' Hall, Michael J. Hammel, Dwight Johnson, Evan Leibovitch, Dave Phillips, Alessandro Rubini, Doc Searls, Jamie Zawinski, and many others. And, of course, the infamous "Answer Gang" - though the Gang started small with Jim Dennis as the Answer Guy. Over the years, the Linux Gazette has remained true to its roots, providing high-quality, noncommercial information aimed at making Linux more fun.
The Linux Gazette has reached a fork in the road, however, which threatens to make things somewhat less fun for a while. The volunteer core which puts together the Gazette has announced that the publication is leaving SSC's embrace, and is striking out on its own. This group has put out an Issue 96 which includes a fairly strongly-worded editorial:
The dissidents have set up shop at LinuxGazette.net. Meanwhile, SSC continues to operate LinuxGazette.com, which has published an Issue 96 of its own. There are, in other words, two competing publications using the same name and even the same issue numbering scheme.
The core of the dispute is a decision by SSC to move the Linux Gazette to a modern content management system with reader forums, a constant stream of articles, etc. Phil Hughes explained the reasoning for this change to us:
To many in the Linux Gazette organization, the changes to the site went against everything the Gazette had always been: a high-quality, edited, carefully-selected, monthly publication which can be mirrored worldwide. Rather than be part of a publication which, from their point of view, has been thoroughly compromised, these people decided to leave SSC - and to take the Linux Gazette with them.
What will happen now is unclear. Having two publications each claiming to be the "real" Linux Gazette seems unlikely to be good for either one of them. The departing contributors have asked that the LinuxGazette.com domain name be transferred to them, but that seems unlikely to happen. According to Phil Hughes, "SSC will continue to run Linux Gazette and it will continue to appear at www.linuxgazette.com." SSC does own that domain name, and it has a seven-year history of publishing the Linux Gazette (and employing its editor for most of that time); it could be hard to find anybody with a stronger legal claim to the right to use that name if it came down to a fight.
There have been signs that this disagreement could turn nasty, and some accusations have started to fly. These include stripped copyrights (since fixed, apparently) and censorship issues: the LinuxGazette.net Issue 95 mailbag contains a couple of letters which are missing from the LinuxGazette.com version. SSC has, by its own admission, been deleting posts on LinuxGazette.com that reference LinuxGazette.net, and has started making noises about trademark violations. Even so, most of the people involved seem to understand that neither the Linux community nor the Linux Gazette (either version) needs an ugly public feud. One can only hope that the relevant parties are able to keep that idea in mind as they carry their respective projects forward.the associated press release; others came out during the press conference and afterward.
SUSE was bought for $210 million in cash. The deal thus values SUSE at less than one tenth the market capitalization of Red Hat ($2.4 billion as of this writing, down somewhat after Novell's announcement). Since SUSE has never been a public company, information on its finances has been hard to come by; at the press conference, however, SUSE's revenues for this year were estimated to be between $35 and $40 million. Red Hat's revenue will be on the order of three times that figure. So SUSE truly is a smaller company, deserving of a lower valuation. The magnitude of the difference is striking, however.
The press conference was full of upbeat "forward looking statements" on how this acquisition positions SUSE as a proper competitor for Red Hat. Novell's large, global support and training operation was mentioned several times; indeed, having a business on that scale behind the distribution might just help nervous CIOs sleep better at night. Novell and SUSE also have high hopes for Novell's large sales (and reseller) channel. Of course, Caldera/SCO was also supposed to succeed based on its channel... Novell also wastes no opportunity to point out that it now has "the whole stack" of offerings, from the base operating system through its proprietary enterprise products.
SUSE currently has 399 employees. It appears that Novell plans, for now, to keep the technical staff around; there may be some reductions in the administrative area, however. Novell has stated that it is committed to maintaining SUSE's presence in Nuremberg.
Novell's management has learned to say the right things with regard to the open source community - though they accepted no questions from community publications at the conference. Novell, says CEO Jack Messman, "expects to learn a lot" from the SUSE engineers, and plans to continue to be a leader in the development community.
After the conference, we asked about Novell's plans in a couple of areas. Unlike Red Hat, Novell/SUSE does not plan to drop its retail distribution; instead, it will fire up its sales channels and try to create a much larger presence for all of SUSE's products, especially in the U.S. The situation with desktops is a little less clear. SUSE has long been a supporter of the KDE desktop, but Novell owns Ximian, which is rather firmly in the GNOME camp. Novell, apparently, doesn't yet know how it will resolve that difference; PR person Kevan Barney told us "We'll be evaluating how to proceed on the desktop front in the coming months."
The same press release announced that IBM is investing $50 million into Novell. The two companies will be negotiating other deals in the future for the continued support of SUSE Linux on IBM's platforms.
The long-term consequences of this deal could be large. Red Hat is no longer the biggest Linux distributor; it will now be competing with an established, large company with a huge installed base of customers. The upper end of the Linux distribution market looks increasingly like a duopoly controlled by two giants. Despite the wealth of distributions available to Linux users, only a very few of those distributions will ever develop the mass to be successful in the commercial arena.
It's also worth noting that, for the first time, one of the core Linux distributions is owned by something other than a community-based company. It is certainly possible for a large company like Novell to handle such a resource properly and not ruin SUSE's relationship with the development community that supports it. Novell does seem to be trying to do the right thing in this regard. This acquisition might just work, and it could turn out to be a good thing for everybody involved, but the Linux commercial landscape has a different look than it did last week.announced the end of the Red Hat Linux product line in July, and merged the Red Hat Linux Project and Fedora Project in September. Still, the end-of-life announcement sent to Red Hat Network (RHN) subscribers this week seemed to catch some by surprise:
While Red Hat will continue to sell and support its enterprise line of products, users who have grown accustomed to the (relatively) inexpensive Red Hat Linux line and RHN support are now looking for other options. Users have about six months to decide what direction they want to go. RHN channels with updates for discontinued versions will remain available for at least six months after the end-of-life, but the April 30 date will be the end of new errata for regular Red Hat Linux products. RHN subscribers who are paid-up past April 30 will receive an evaluation ISO for Red Hat Enterprise Linux WS and channel access to updates for that distribution until their subscription expires. Red Hat is no longer allowing subscribers to extend their subscription past April 30, though subscribers can renew up until April 30 for $20.
The first option for Red Hat loyalists is Red Hat Enterprise Linux. Red Hat is offering introductory pricing for Red Hat Enterprise Linux WS or Red Hat Enterprise Linux ES; the deal is 50% off either product for up to two years, putting annual cost of the workstation product (WS) at $89.50 per system and the server product (ES) at $174.50 per system. This pricing structure, while not overly expensive for a single system, may not be popular with Red Hat users who have been maintaining multiple systems with Red Hat Linux.
Another likely choice for hobbyists and users who have grown fond of Red Hat Linux is the Red Hat Linux Project's successor: Fedora (which has just made its first core release). Fedora will likely replace Red Hat on many systems as Red Hat Linux 9 approaches end-of-life. However, many users are likely to be a bit wary about adopting Fedora as the project is still in its infancy and it has yet to be seen how well the project will evolve. Fedora will will also be a more volatile distribution, with each release being, essentially, a "dot-zero" version.
Users might also choose to move to derivative products like Tummy.com's KRUD, or KRUD Server. KRUD is based on Red Hat Linux and users can opt for a monthly subscription with updates via CD-ROM. A one-year subscription to KRUD will run users $65, a one-year subscription to KRUD Server is $190. This may be an attractive option to many users, since Tummy.com does not require a per-machine subscription. Thus, a KRUD subscription is usable on any number of machines, unlike subscriptions to Red Hat Enterprise products.
There are, of course, other distributions which will be more than happy to pick up customers left behind by Red Hat. Red Hat's termination of the "consumer" line of products may be a blessing for other commercial Linux distributions with a strong interest in the retail Linux market. SUSE, Mandrake Linux, Xandros, Lindows and other commercial distributors may pick up some of Red Hat's audience still looking to buy a supported retail product. Non-commercial distributions like Debian might gain users as well; see this week's Distributions page. In the commercial arena, Mandrake is still working to emerge from bankruptcy, leaving SUSE as the strongest contender for the retail market at this point, particularly with the backing of Novell.
Joseph Eckert, SUSE's Vice President of Corporate Communications, told us that he is optimistic about SUSE's prospects in the retail channel. He noted that SUSE has seen a jump in sales with the 9.0 release, though it has not been available for very long. Unlike Red Hat, SUSE's retail products still account for a significant portion of their overall sales. According to Eckert, SUSE expects between €35 million and €40 million in sales this year, with the SUSE's desktop products accounting for more than 50 percent of their business.
Eckert also said that SUSE has no plans to cancel its desktop products. "As Red Hat continues to distance itself [from retail products] we consider it a service to the community to keep the desktop alive...it's not just about the enterprise desktop, it's about making sure that our community of developers and enthusiasts are satisfied."
Indeed, it may be important for any vendor interested in the enterprise to keep developers and enthusiasts happy. Red Hat's decision to abandon retail products and focus solely on its enterprise products may help boost Red Hat's rivals in the enterprise market as well. Red Hat found its way into many organizations because that was what IT staff used at home. With some of Red Hat's user base looking at moving to different distributions, they may decide to bring those distributions into the workplace with them.a news release and the actual order - both in PDF format.
Why do we need a digital broadcast flag? From the order:
So "mass indiscriminate redistribution" is not a problem now, but preemptive action is the way of things in the US these days, so we have to mandate copy protection mechanisms for transmissions on our public spectrum.
The actual broadcast flag rule, as found in page 40 of the order document, states that a digital TV demodulator cannot send unprotected content to any output, except in a set of specific cases:
The FCC repeatedly asserts that home recording will not be affected by the broadcast flag. The rules, however, do place significant constraints on digital recordings. In particular, the resulting recording cannot be transferable to another device, or the recorder must be explicitly "authorized" by the FCC. The MPAA had pushed hard for the "authorization" mechanism to require, among other things, approval by at least two "major studios," but the FCC, at least didn't buy that. Instead, there will be an involved bureaucratic process where manufacturers of recorders have to show the FCC how their product will implement copy protection schemes.
Much debate evidently went into the specification of "robustness rules." The MPAA wanted an extensive set of regulations on things like "how content may be transmitted on data paths within Demodulator Products" and such, an an effort to make circumvention as difficult as possible. The FCC, however, concluded that a level of robustness sufficient to defeat an "ordinary user" would be enough. Interestingly, the FCC uses the CSS scheme used on DVDs as an example:
One might have just as easily concluded that a copy protection (and "region coding" price support) scheme like CSS was unnecessary in the first place, but the FCC wasn't willing to go there.
The resulting "robustness requirements" say that the broadcast flag scheme must be implemented in products in a way that can't be defeated or circumvented by "an ordinary user using generally-available tools or equipment." Examples of such tools, as listed in the regulation, include screwdrivers, jumpers, clips, soldiering irons, EEPROM writers, debuggers, and decompilers.
This rule will have an obvious effect on free software - under the broadcast flag provisions, there simply cannot be a free TV demodulator system. Even if somebody wrote a free system which implemented the broadcast flag restrictions, a source-available system clearly would not meet the "robustness requirements." The FCC report does, at least, note this problem:
Given that the FCC seeks further comments, the free software community would be well advised to provide them with those comments. The Electronic comment filing system can be used for this purpose (the docket number for the report is 02-230). The chances of getting any sort of free software exemption to the broadcast flag requirements appear slim, however. The MPAA might not have gotten everything it wanted out of the FCC - thanks to the efforts of the EFF and many others - but that organization remains the driving force behind the FCC's rulemaking.
Page editor: Jonathan Corbet
Next page: Security>>
Copyright © 2003, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds