Letters to the Editor

In Phil Hughes' response to Thomas L. Gossard in the July LTE column, he said that Corel's license to Caldera for WordPerfect allows its use only on Caldera's version of Linux.

Corel's web page (http://www.corel.com/) has links which lead to a page on http://www.sdcorp.com/ for WordPerfect 7, showing it as certified for Red Hat, Slackware and OpenLinux, without mentioning any restrictions on using other versions of Linux. The current Corel port is WP 6, which is getting fairly old now. The Corel/SDCORP web pages say that WordPerfect 7 for Linux will be available in June (until a few days ago, it said May).

—Bob Nielsen nielsen@primenet.com

I was referring to the use of WordPerfect 6, the only one available at the time. All new Caldera licensing will allow use on other Linux platforms.

—Phil Hughes phil@ssc.com

“The first virus able to infect a Linux system has been found by McAfee Associates. The virus, named Bliss, has spread to Linux systems, as many Linux users play Internet games while logged in as root.” [“From the Editor”, May 1997]

Have you done ANY independent research about Bliss, or do advertising dollars simply obligate you to propagate untruths which the marketing-driven McAffe Associates has spread about this virus?

Bliss cannot be spread by playing Internet games whiled logged in as root.

Check your facts.

—Kirk Haines khaines@oshconsulting.com

For the record, McAfee Associates has never advertised in our magazine. Thanks for letting us know that Bliss cannot be spread in this fashion.

I just received LJ Issue 39 and have discovered some errors in the published article “Porting Scientific and Engineering Programs to Linux”, which I wrote with Charles Kelsey.

In the first line of the fifth paragraph, there is a reference to NCBO code. I don't know what NCBO means. In the manuscript that I submitted this line read “One thing that makes porting this code to a new platform somewhat challenging is that it is a safety related, pedigreed code.” Perhaps you intended for the final product to read, “One thing that makes porting the MCNP4a code to a new platform somewhat challenging is that it is a safety related, pedigreed code.”

Also, in Listing 1. FORTRAN Patch File, whoever retyped the patch file typed less than symbols (<) where there should have been commas (,). Humorous and quite understandable, since they're on the same key. Fortunately, the files offered for download at URL ftp://ftp.linuxjournal.com/pub/lj/listings/issue39/2177.tgz appear to be correct.

Thank you for your attention and for the opportunity to share with the Linux community.

—Gary Masters gmasters@devcg.denver.co.us

Please accept my apologies for these inadvertent changes in your text.

Regarding the Correction in the July issue, page 93, I would have thought that most Unix C programmers would know that the main() function in C is required to return an integer result. Thus, the correct first code line for Richard Sevenich's main() is:

int main()

not void main(). See Steve Summit's excellent comp.lang.c FAQ for a more in-depth explanation. There seems to be more than enough confusion about this point propagated by the DOS/Windows crowd without LJ contributing to it as well.

—Matthew Saltzman mjs@clemson.edu

In “Porting Scientific and Engineering Programs to Linux” [July 1997], the authors [Charles Kelsey and Gary Masters] write:

“The f2c has inherent limitations, leaving g77 as the only viable alternative for compiling large, complex applications written in FORTRAN 77 under Linux.”

I was amazed to read such a blanket dismissal of f2c. At Berger Financial Research, we've been using f2c to compile FORTRAN since kernel .99pl8. We write volumes of FORTRAN and use standard libraries such as SLATEC, LAPACK, MINPACK and various routines from TOMS. The last time we had a problem with f2c was a couple of years ago, and the maintainer fixed it within a week of reporting it.

What exactly are the “inherent limitations” of f2c that the authors refer to?

—Dr. Harvey J. Stein abel@netvision.net.il