I'd just like to pass along my praise for the article “Algorithms in Africa” by Wayne Marshall in the June 2001 issue. This is a definite upgrade over the typical Linux success story. Insightful, committed, poignant, experienced and informed, you should consider Mr. Marshall's perspective as paradigmatic for covering the emerging global presence of Linux. The principles and values here have definite application to domains such as India (where the FSF is opening a branch office), China (where the government has adopted free software, if not political freedom, as its own) and many other areas of the world such as Eastern Europe and South America. Please keep us up to date on global development. And thanks for a superb magazine.
—William G. McGrath
I just wanted to write to let you know that I've consistently found Linux Journal to have the highest quality content in the magazines that cover Linux and technology. I am constantly getting refresher courses, learning about new code and projects, and generally getting fantastic info from your publication.
I especially wanted to compliment you on your regular sections: At the Forge, Cooking with Linux and Paranoid Penguin. Much of the information is applicable to other *nix to boot, making your publication one I keep around for a long time (much to the consternation of my wife). Anyway, thanks folks, and keep up the good work.
In the July 2001 LJ article “Debugging Memory on Linux”, I noticed that the open-source memory checker I've been using was not listed in the article. The checker contains a replacement malloc library plus patches for gcc. The gcc patches wrap C++-like constructors around local variables and insert tests before memory references. This allows checked programs to detect memory overwrites of local variables and some global variables in addition to malloced buffers, and the checking catches overwrites as soon as they happen. You may freely mix object modules compiled with and without checking. The checker also includes replacements for mem* and str* routines and can detect invalid calls against checked memory objects, even from modules compiled without bounds checking.
There are links to the checker from the gcc extensions page at gcc.gnu.org/extensions.html.
In your article “Debugging Memory on Linux” in the July 2001 issue of LJ, you list Purify from Rational as a proprietary tool. As far as I can tell from their web site, they do not support Linux. Also, a while back I did talk to a Rational salesperson who said they didn't have any plans to support Linux. Do you know something else?
Sorfa replies: It looks like you are right. At the time of writing the article (early this year), there was a hint that Purify would be supported on Linux. I assumed (wrongly) that by the time the article made it to press, it would be available. It is a pity and I apologize for the incorrect info. It looks like the only proprietary alternative is Insure++.
I must respectfully disagree with Allan Hall in his letter of the July 2001 issue. Certification per se is certainly no substitute for experience, but it does show that a candidate at least took the initiative to attend some classes, read some books and pass some tests. It also usually requires putting a few hundred dollars up front.
I don't see how one could give a certified candidate anything but an edge over an uncertified one, experience levels in the two being equal.
Just want to write to let you know that Robin Rowe's article “MPEG-1 Movie Players” (May 2001) was very helpful and also convinced me to renew my subscription to Linux Journal. I wanted to play movies on my new notebook and had played with xanim before, but your recommendation of MPlayer was great. It compiles, installs and works like a charm. Thanks again.
It's articles like “CVS: an Introduction” (July 2001 issue of LJ) that keep me subscribed to Linux Journal. I've been doing basic RCS for years and knew there had to be a better way. But let's face it, the man page for CVS is a little overwhelming to the uninitiated. But the day after reading the article, I was using CVS at work (the magazine is opened on my desk to page 72 right now), and I'm feeling much better about long-range management issues now. Keep 'em coming! So many thanks to you and Ralph Krause for putting this together.
I just went through your article about Nessus in LJ (June 2001), and I just wanted to congratulate you, both on the accuracy of what is being said, as well as on the excellent step-by-step instructions you gave to readers.
Just to illustrate how well you explained: I only received a few installation-problem questions from (what seems to be) readers of your article, while the mailing list gained about 100 new users the two weeks after the release of LJ. (And congrats for not having taken the screenshots directly off nessus.org, as so many persons do.) A very fine job indeed.
—Renaud Deraison The Nessus Project
Choong Ng's review of Mandrake 8.0 in the August 2001 issue gives a good overall picture of the strengths and weaknesses of the release. However, he failed to mention one important issue. It seems that the stock kernel shipped with Mandrake 8.0 will not work with the PS/2 mouse pointer in many IBM ThinkPad laptops. On the ThinkPad mailing list, some people have not even been able to install Mandrake 8.0 because of this.
Izzet Agoren's Kernel Korner (May 2001) The line of code on page 28 that reads
cat /proc/sys/net/ipv4/ip_forwardingMitch Chapman's “Create User Interfaces with Glade” (July 2001) The link for the Glade home page listed in this article is obsolete. The new location is glade.gnome.org.