SuSE, MandrakeSoft to promote Linux NC architecture
But is Linbox's Linbox Network Architecture quite what the company says it is?
Linux distributors SuSE and MandrakeSoft are to back a little-known French IT operation Linbox's attempt to define the open source OS as a standard for network computing. The two distributors will promote something called the Linbox Network Architecture (LNA), a diskless client/server set-up. Right now, LNA comprises Linbox's NetServer hardware, which already ships with SuSE and Mandrake versions of Linux, and up to 32 diskless NetStation clients hanging off it. Users can run Linux applications on the client, loaded off the server, or on the server itself. However, LNA systems could easily be built around hardware from other vendors, and ultimately even using software from other Linux distributors. SuSE, MandrakeSoft and Linbox plan to promote LNA as an open standard. That said, since LNA is, according to Linbox's Web site, based on "free software", it could surely be little else? Because it appears to contain proprietary software -- software, more to the point, that Linbox doesn't own. Quite apart from LNA's ability to operate like Sun's Star NC architecture and the like, Linbox claims the system offers what it calls "300 per cent compatibility" -- in other words, it's 100 per cent compatible with three operating systems: Linux, Windows 98/NT and MacOS. Linux compatibly is obvious -- it's a Linux system -- and Windows support apparently comes from VMware, inc's VMware emulation software (SuSE also has a relationship with VMware). But MacOS? Again, emulation is the only option here. A closer look at Linbox's site reveals that its servers ship with software called the "original Macintosh ROM". Since Apple no longer licenses either the MacOS nor the Mac ROMs, we're not entirely sure how Linbox can ship this -- and certainly not as software. Linbox also claims LNA presents users with a 75 per cent cost saving over Windows-based client/server systems. However, since the saving is calculated largely on up front box and software costs, some questionable hardware lifespan parameters, and, more importantly, no independent verification, perhaps we should question this claim too. ®
Sponsored: 2016 Cyberthreat defense report