> Bob, just my two bits as a linux-only user for over eight years. > > one has to use the software that one is comfortable with. for me, > personally that comfort zone includes software efficiency as well as > an acceptable (for me) ideology. and i am perfectly willing to accept > that your parameters may be different. > > for me open source is important because diversity is important. if > someone wants alternatives, there should be space for them. and > monopolies, by their very nature, aim for the opposite.. > > fwiw, being in touch with software developers from around the world, > making suggestions and seeing them implemented sometimes within days > is itself a pretty nice experience, for me... >
Using Microsoft for convenience because it matches what's used at work doesn't preclude people from also aligning themselves with free software and open source for other aspects of their computing. To keep myself sane and in touch with a high quality model of software development I also use a system called Oberon, developed by Niklaus Wirth and the clever people at ETH Zurich. It's far and away the most efficient and radical operating system I've ever had any experience with, and a great many of the spin-offs from the project have found their way over the years into more mainstream stuff like Java. Many of the people who were important in the development of Oberon are now working at places like Microsoft Research and you can see their influence making itself felt. Of course there must be space for diversity - I'm not arguing against that, just against making one's life difficult just because one doesn't like Microsoft. You imply above that linux is efficient in some absolute sense rather than efficient when compared to Windows. It may well be efficient compared to W (although I think that depends at what level you're measuring efficiency and whether that level matters at the level of the end user). But I don't think you could call it efficient in some absolute sense. I read something a few days ago (which unfortunately I can't find now) that was from a link posted here or from a link I followed from something posted here, which tried to put a value on Linux to show what people were missing by not using free software. It was a very flawed piece of reasoning, I have to say, but nevertheless the figures were quite interesting. Based on KLOC and function points it came up with a $ figure that they reckoned it would cost to develop from scratch one of the common distributions, and another smaller $ figure just for the kernel. The number just for the kernel was billions of dollars. Billions of dollars to develop an operating system kernel? Give me a break! Either the calculations are wrong by many, many orders or magnitude or that kernel is a complete rat's nest of inefficiency and waste. A good principle for system design is that it should be simple and clean enough to be understood in total by one person. A kernel should not be so complex that any suitably intelligent person (eg a PhD in Computer Science) cannot understand it as a whole. If a kernel costs billions to develop there's no way it can be kept in one person's head and I don't see any way of describing that as efficient. The Oberon operation system by contrast, was developed by 2 people in a couple of years. When I first acquired a copy it fit onto a 1.4 meg diskette containing the operating system, gui, compiler, network connectivity, all the source code and so on. Even I could understand it. So much so that I developed a small clone running under Windows just to prove to myself that I understood. And unlike Linux it doesn't spend its time trying to be Windows but not Windows. N. Wirth - Designing a System from Scratch - Structured Programming, 10:1, 10-18, 1989 http://www.oberon.ethz.ch/ Bob -- PDML Pentax-Discuss Mail List PDML@pdml.net http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
