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

Eazel shuts down. It's official, Eazel has shut down. The company ceased operations on May 11, after the last attempts at raising money failed. This is an unfortunate development, but it is not particularly surprising.

After all, Eazel never really has presented a convincing story on just how it was going to make money. It has been big on flash and hype - a year or so ago the press was full of stories on how the original Macintosh developers were going to save the Linux desktop. Back when Eazel was being founded, an interesting idea and the name "Linux" were enough to get a pile of investment money and get a company started. Those were fun times.

But those times are over. Now a business has to come up with ways of separating customers from their money to survive. Eazel's plans along those lines seem to include:

• The Eazel Software Catalog, which is another service providing easy software downloads and installation.

• Eazel Online Storage, making 25MB of disk space available (via Nautilus) over the net.

The online storage service seems an unlikely money maker - 25MB of space (almost enough to hold a set of RPMs for Nautilus and its support code) isn't worth a whole lot. And people like to keep their files close at hand, so that, say, if the company goes out of business, they can still get their data.

The software catalog idea has drawn more interest, to the point that a number of companies are trying to make it work. Consider the Red Hat Network, Acrylis WhatIfLinux (see below), Aduva Manager, Ximian Red Carpet, and, yes, the Eazel Software Catalog. It is not surprising that the GNOME-oriented companies are providing this kind of service: anybody who tries to track a bleeding-edge GNOME application tends to end up in a maze of twistly little shared library updates, all interdependent. So an update service may make it easier for people to play with Nautilus or Evolution, but it is a hard business proposition for a few reasons:

• A great many users don't want to update their software frequently. Surprisingly enough, people tend to be happy when they make their systems work, and are uninclined to mess with them thereafter.

• Those who do want to upgrade their software tend to learn, early on, that they can do so without having to pay for the privilege.

• Those who are inclined to pay anyway will probably go to the source of the distribution they are using - that's where the bulk of the updates come from, anyway. So, for example, the Red Hat Network has an inherent advantage over third-party services in selling to Red Hat users.

Eazel has also worked some deals with companies like Red Hat, but it is hard to see them amounting to much. A couple of years ago, a company like Eazel could count on a new round of funding to keep it going while it figured something out (...someday...), but those days are gone.

Eazel the company may have failed, but the company has left behind a legacy: the Nautilus code, which is licensed under the GPL. The code will remain free and will grow without Eazel. That is one of the benefits of using free software: you are not marooned just because the company you have been dealing with goes out of business.

On the other hand, what if you are using proprietary software? The word also went out this week that Enhanced Software Technologies (EST) has been shut down. The word, in fact, is a bit premature: EST still is operational, it has just lost most of its staff. Its future is still being worked out, and the fate of the BRU product is currently unclear. This is bad news for a company that was an early supporter of Linux, and for the users of its products as well.

EST is the maker of the well-respected BRU (Backup & Restore Utility). Work started on the Unix version of BRU back in 1985, and the first Linux version was made available in 1994, when there was very little commercial software for Linux. EST was acquired by Atipa in February, 2000.

Atipa at that time looked very much like it was trying to follow the path blazed by VA Linux Systems. It was expanding its hardware business with software utilities like BRU, and a series of web sites under the "Linsight" name. Then, of course, the bubble burst. Atipa's CEO left abruptly (and has disappeared from the Linux community without a trace), and most of Linsight was shut down - only LinTraining and LinEvents remain, and they are not operated by Atipa. The Atipa logo was literally everywhere at the 2000 New York LinuxWorld conference; it would be a lot harder to find now.

In fact, Atipa does not make hardware anymore; that business was sold to Microtech Computers back in March. The company has now decided to make its bet in the network management arena, with the OpenNMS package (see this week's Development page) as its centerpiece. But Atipa enterprise backup software is evidently not considered to be a network management issue; thus, the decision to gut EST and try to sell it off.

The ultimate fate of EST and BRU is yet to be worked out; according to Atipa Director of Corporate Communications Darrek Porter, negotiations are currently in progress and should be completed shortly. It is possible (and rumored) that it will be purchased by a competitor, who will simply remove it from the market. Mr. Porter, of course, would not comment on that possibility. But the fact that the company was gutted before a deal was worked out gives a clue of where Atipa thinks things will go.

This move leaves a lot of BRU users in an interesting position. The software they depend on for their system backups could simply vanish. Changing backup systems on a large network is a painful (and expensive) experience. There is also the issue of being able to read old backup tapes when the software is no longer supported.

Here we see the legacy of proprietary software. Companies that depend on such software can find the systems that support their business pulled out from under them at any time. Depending on the good will and longevity of a software business is a risky endeavor. A few incidents like this one, and the corporate world may begin to really understand the risks it is taking.

That said, it is worth pointing out that, as far as we know, there is still not a free, top-quality large network backup and restore system available for Linux. Numerous commercial alternatives are out there, but the available free systems just do not have the same level of features and scalability. This could be a good project for somebody...

Caldera buys WhatIfLinux. Caldera International has announced the acquisition of the WhatIfLinux system from Acrylis. WhatIfLinux is one of the package management and update services mentioned above. It can track the software on your systems, point out those that could benefit from upgrading, handle dependencies, and make it all happen.

WhatIfLinux as part of Acrylis' business raised all of the concerns mentioned above - it looked like a difficult path to take. Evidently Acrylis thought so as well, and decided to sell. This service (now "Volution Online") as part of Caldera's offerings makes a lot more sense. As Caldera tries to hang on to all of those SCO enterprise customers, it will have another management service to sell them. This is a move that may actually make sense.

SGI sets the TPC-H benchmark record on Linux. SGI has put out a press release describing a database benchmark it ran recently. The system was an SGI 1450 server with four nodes, each of which has four processors, running a 2.4.3 kernel and IBM's DB2 database. The resulting performance set a new world record: yes, Linux is now at the top of the database performance chart.

This result is important - it shows that Linux can play in the "enterprise database" arena, in a language that companies can understand. It should help pave the way for more high-profile corporate deployments. Certainly it's not for household deployments - the system that ran this benchmark lists for just under $1 million.

Of course, the really nice thing will be when a Linux system running a free database management system takes that top spot. It may happen sooner than many people think...

Inside this week's Linux Weekly News:

• Security: A worm to fix cracked hosts, new vulnerabilities in CUPS and sendfile, lots of distribution patches.
• Kernel: The device number moratorium; A Linux kernel fork?; the shape of device naming in the future.
• Distributions: HP and Debian, Linux-Mandrake ISO image policy, Linux in Argentina, Openwall GNU/Linux.
• On the Desktop: KDE runs fast, Konqueror runs clean and Eazel runs away.
• Development: OpenNMS review, new GNU Ada, Omni printer driver questions answered, Lisa first beta, Perl Exegesis 2.
• Commerce: New websites from IBM, Caldera and Nokia/CollabNet, the last word on Mundie.
• History: How to fund open source software development, the future of Linux distributions.
• Letters: RMS on PriorArt.org; ESR on back doors; bloatware; Dell.

This Week's LWN was brought to you by:

• Jonathan Corbet, Executive Editor
• Elizabeth O. Coolbaugh, Managing Editor
• Michael J. Hammel, Senior Editor

Next: Security