Letters to the Editor

For several years, my son has been telling me about free software, and I have failed to understand completely what he meant. I associate “free” with “without cost”, and since I spend very little on software anyway, I have felt no need to use “free software”. I did not until recently understand that the freedom to change the software, which access to the source code provides, is far more important than lack of cost.

This misunderstanding is widespread. Therefore, I would like to propose an alternate term for software with source code—“liberated” software. Liberated software liberates the programmer. —Daniel L. Johnson, M.D., F.A.C.P johnsond@uwstout.edu

Once again I'm at work in front of my HP workstation, about to request more information on a software product I noticed in an LJ ad, that just might solve a problem a co-worker asked me about last week. A few months ago my system administrator bought BRU (from an LJ ad I gave him) for an HP workstation headed to a tele-commuting co-worker's house. I occasionally daydream about having time to explore my Linux system at home, but I consistently read LJ because its columns, articles, and advertisements give me information I use at work. —Greg Deitrick deitrick@shell.com

I would like to say that I was absolutely impressed with the May issue of the Linux Journal, especially with the articles which dealt directly with system administration such as how to set up a WWW site, the article of an ISP using Linux, the Majordomo setup/configuration article, etc.

Keep up the good work. Now I remember why I subscribed to this magazine in the first place! —John Coy jcoy@magic.yournet.com

Several times your articles have mentioned a guy from NASA who uses PVM instead of supercomputers, and that he gave a talk on it at a conference, etc.

Please have him, or someone in attendance, paraphrase what he is doing, and perhaps speculate that nnn Pentium 90's equals such and such Supercomputer, at so many gigaflops, etc.

Please also consider a `Porting Corner' article every month to summarize the progress of ports to other computers (i.e. Alpha, MIPS, PowerPC, etc). Only a quarter page or smaller would be required, with a line or two on each port. —Doug Fortune

In response to your first point, we do intend to have an article on this system. However, there is no way to say that nnn Pentium 90's is equivalent to any supercomputer; a loosely-coupled parallel system like that works well only on certain problem domains. We intend to have an article on the Beowulf system, but it is currently an ongoing research project, and is not ready for an article at this time.

The architects of Beowulf are about to build a second-generation system based on their current experience and research with their first system, and the results from that system will be more interesting and worthwhile to readers with a serious use for the technology. All the software that they are using will be released as a package, and when it is, we will certainly pursue an article.

In response to your second point, progress on the ports to various architectures does not progress in a way that facilitates monthly reporting. We will report on significant progress on the ports. In particular, our “Stop The Presses” article in this issue mentions that the Linux/Alpha port is now self-hosting. We will also report on the state of all ports from time to time, as we did recently.

Corrections