CyberKnights

Modern tools.
Traditional dedication.

Linux

CyberKnights prefers to base systems on the popular Linux™ operating system for a wide variety of positive reasons, & a few negative reasons.

Positive Reasons

Linux™ is extremely stable to run. Continuous uptimes of several years are common. Not quite to the same extent as, for example, some of the open-licenced BSD systems, but sufficiently stable that both systems are often run continuously until hardware or the entire system is replaced.
Linux™ is extremely efficient. It makes optimal use of all available resources, particularly memory & processor (CPU) power. Because it is so efficient, & halts the processor when it’s not actually doing things, Linux™ systems are also power-efficient & can often survive failure of CPU and/or power supply cooling systems.
Linux™ is extremely scaleable. The same operating system runs tiny embedded devices & multiprocessor super-servers, diskless workstations & mainframes.
Linux™ is extremely portable. While other operating systems are dropping support for processors, even going as far as specialising down to one single processor type, only NetBSD rivals Linux™ for the number of different kinds of processor it runs on.
Linux™ is widely supported. As well as running the vast majority of Posix & UNIX software with little or no adaptation, Linux™ has a large stable of software for which it is the primary or only development platform.
Linux™ has drivers for an extremely wide range of devices, & is relatively easy to write a device driver for. As one bright competitor said, “Any idiot could write a driver in 2 days with a book like Linux™ Device Drivers — there is no such thing as a 2-day device-driver for [his product].”
Linux™ distributions, with much in common, are available from a wide range of suppliers.
Linux™ is licenced under the GPL, which would make it extremely difficult to subvert or abuse.
Linux™ & the systems commonly run under it have a very fast issue turnaround time, which is especially important in the security arena. It is not uncommon for a tested Linux™ fix to appear within a few hours of problem discovery, & for proprietary competitors to spend weeks or months drumming up & testing (on good days) an answer to the same problem.
Linux™ supports a wide variety of filesystem & networking standards. This permits interoperability with many other systems, & also provides a range of (for example) fully journalling filesystems to suit particular applications.
Linux™ is contributed to by a large number of very competent people & organisations, & tested by an extremely diverse population, which affords it a solution scope & testing space unmatched by any more limited or artificial milieu.
Linux™ is understood & supported by many competent people & groups, which affords it a very broad support space, avoiding the many “lock-in” traps buried in the common closed-source, single-brand, proprietary-systematics approaches. Those competent people — an awesome array of competent people — can support your software, develop from it, extend it, without submitting to special licences, special closed-shop training sessions, exclusive or patent-encumbered documentation, et cetera.

Negative Reasons

Linux™ avoids the complex, expensive & ever-changing licencing conditions associated with some proprietary operating systems.
While the primary APIs in Linux™ are not completely frozen, they are highly interoperable with other UNIX-like systems (such as OS/2, BeOS & Apple’s new OS X), & much more stable than common specialised systems. It is also simple to run several generations of an API side by side, since shared libraries are not restricted to a short naming scheme or consequent ambiguity of contents.
Even systems which are generally hostile to & subversive to standards (such as Microsoft Windows) can often support, albeit in limited fashion, software written to a Linux™ API with little or no modification. CygWin is an example of such a support system, as is XMing. Services such as the Apache web server & the PostgreSQL database can be shoehorned into it.
Linux™ does not depend on a marketing or management structure with a history spanning several decades of frequent misleading statements or outright lies, but rather operates on a policy of full disclosure.

How can you make money from “free” software?

Directly: by selling it. The GPL explicitly permits selling & marking-up software released under it.
Indirectly: by selling systems, support or consulting based around it. See below.

Isn’t “freeware” low quality?

“Freeware” is indeed often of low quality, but “Free Software” & other Open Source Software is often of very high quality due to the refining nature of being directly accessible to your userbase.
One makes money from Free/Open Source Software by treating the software itself as a commodity, & engineering a business around how that software solves people’s problems. Consider these two quotes from an OSIA discussion of the topic. Both of the people quoted are paid by companies making money from FOSS in different ways.
“Anyone that can deploy YaCMaS wasn’t going to come to us for a site anyway, so we might as well have them help with development. ;-]” — Rodd Clarkson.
“That’s interesting, isn’t it. If a developer does not provide support, they are less likely to get significant (if any) revenue. Sounds to me like decent quality control!” — Arjen Lentz

Linux® is the registered trademark of Linus Torvalds in the United States & other countries. Here in Australia, it is a common-law Trademark in the process of becoming Registered.

ALSO, THERE IS NO WARRANTY OR CONDITION OF TITLE, QUIET ENJOYMENT, QUIET POSSESSION, CORRESPONDENCE TO DESCRIPTION OR NON-INFRINGEMENT WITH REGARD TO THE SOFTWARE PRODUCT.

They also assume an economic liability of $5 or software price, unless forced by some applicable law.

Is this the kind of customer support that makes proprietary software a better choice for business use? — nicke on LWN

Last changed: 09-Sep-2008 10:29:50 Find out who links to this page. Verify for yourself that this page is pure, standard HTML, not Ruby.

If you would like us to read email for USD$1000 per page, payable in advance, send it here.