First Look at Asianux 2.0

If you visit Asianux.com, you'll find yourself on a web site of a project with plenty of ambition. Perhaps the best indication of it is a chart on this page which attempts to define the sphere of influence of the main enterprise Linux distributions in the world, with Red Hat dominant in North America, SUSE in Europe and -- you've guessed it -- Asianux on much of the huge (and potentially very lucrative) Asian continent. Although the reality is a little more complex than the chart would like to us to believe, the Asianux project has succeeded in attracting much attention in the Linux media. With the recent release of Asianux 2.0 we decided to download the two ISO images for the i386 architecture (images for x86_64 and pSeries processors are also available) and install it on a Pentium 4 box to take it for a test drive.

Before we start, it is important to stress that Asianux is not meant to be used as a standalone product. Although this is not clearly stated on the project's web site, the lack of any security and errata pages makes it obvious; in fact, since the release of Asianux 1.0 in April 2004 the project has issued just a single service pack, rather than regular security updates to vulnerable applications, as one would expect from an enterprise Linux distribution. Instead, Asianux serves as a base for the three participating Linux vendors - China's Red Flag Software, Japan's Miracle Corporation, and Korea's Haansoft. Of the three, only Red Flag has so far released a product based on Asianux 2.0 - a development snapshot of Red Flag Linux 5.0 Desktop, which is available for free download from Red Flag's web site.

Asianux 2.0 is not an independently developed distribution, but rather obviously based on Red Hat Enterprise Linux 4. This is also true of the Anaconda installer, which, although heavily modified and themed, offers roughly the same steps as any recent Red Hat or Fedora installation program. Nevertheless, there are some differences; for example, the partitioning stage in Asianux also offers ReiserFS and XFS formatting options, but, on the other hand, it completely omits the option to select SELinux functionality, leaving SELinux disabled instead. The package selection is also simplified with the only two options being a "minimal install" and "install everything". Also unlike the Fedora/Red Hat installation program, where some configuration takes place after the first reboot, in Asianux, the package installation step is immediately followed by monitor setup, while the first reboot triggers the good old Kudzu for further hardware configuration. The step to add non-root users has been removed from the Asianux installation program.

The system can boot into a console login prompt or KDM. The only available desktop environment in Asianux 2.0 is KDE 3.2.1 with a "start button" and window decorations and widgets strongly resembling Windows 98. The set of available applications is rather limited, but this is hardly surprising given the fact that Asianux is designed to be a base to build upon, rather than an all-encompassing Linux distribution. As such, don't expect to find much beyond the simplest of tools for common tasks. The only area that has some interesting applications is system administration, which includes several graphical front-ends for analyzing SELinux policies, a tool for authentication configuration, Guarddog firewall configuration, and the usual printer and network setup tools. Also present is a "Control Panel" with several modules that are virtual copycats of their counterparts from Microsoft Windows.

Although the name "Asianux" implies that the operating system is designed for the large Asian continent full of diverse cultures, languages and writing systems, the truth is that Asianux only supports four Asian languages, or to be more precise, four Asian character sets - simplified Chinese (used in China and Singapore), traditional Chinese (used in Taiwan and Hong Kong), Japanese and Korean - these are often referred to as CJK languages. It also supports English. Somewhat surprisingly, the system locale defaults to the national language code set (e.g. Big5, GB2312) as selected during installation, rather than Unicode (UTF8). The input of CJK characters is offered via SCIM, an increasingly popular and intelligent utility (originally developed by Turbolinux) for typing the complex character sets of East Asian languages.

After spending an afternoon in Asianux 2.0 we found few reasons to complain about the operating system. Although the application set is a little outdated and we didn't particularly care for the Windows-like look and feel of the default theme, we found the system solid, responsive, and with trouble-free input of the four supported character sets. The extra graphical utilities included in the system provided for a pleasant surprise. A little less impressive is the Asianux web sites, which lacks documentation and any interactive community resources, such as mailing lists, user forums or Wikis. Also, the distribution is developed completely behind closed doors without any public participation and without any public development releases.

Now for the all-important question: are there any compelling reasons for an Asian company to choose Asianux over its main competitors - the enterprise offerings by Red Hat, Novell and even Turbolinux, which has been active on the Chinese and Japanese markets? After all, Red Hat, SUSE and Turbolinux have been supporting CJK languages for a long time. The latest release of Red Hat Enterprise Linux and the upcoming release of SUSE Linux Enterprise Server will also support a number of Indian languages, not to mention many other languages on the continent, including those of the Middle East and South East Asia. As such, one can argue that both Red Hat and SUSE are, in fact, more "Asian", at least in terms of language support, than Asianux itself.

If Asianux wants to become a leading Linux player on the continent, it needs to do two things. Firstly, it needs to invite other main Asian Linux players to join the development effort - Turbolinux is an obvious example of a highly experienced and relatively successful Linux vendor with good sales in China and Japan, but large Linux companies also operate in Thailand, India and other countries. Secondly, the project should open up to community participation - in a fashion similar to Fedora Foundation, OpenSUSE or OpenSolaris. This is a trend that has already started in other parts of the world and Asian Linux companies would be wise not to ignore it.

---

(Log in to post comments)