I wanted to contact you some time ago with a mention of a
person I think you should honour as someone contributing to the Linux
Community in Sydney, Australia. Forgive me if this has been mentioned
already by others, and I know that I am sending this using another
operating system, but I felt that it was significant to mention the
efforts of one Geoffrey Robertson who runs the Introduction to Linux
courses through the Technical College of Advanced Education (TAFE)
further education courses at Granville, Sydney. A longtime member of
the Sydney Linux User Group (SLUG, www.slug.org.au), he has taken
on some courses in basic usage and advanced usage of Linux, devoting
time to assisting people with getting started and some advanced
techniques in using the Linux operating system.
Actually, I would be surprised if his name hadn't already been mentioned
as a person who has done much for advancing Linux in Australia. I
know he has given much of his personal time both to attending the slug
group, and teaching related courses, such as the LPI courses.
I personally feel you should mention Geoffrey Robertson as amongst
those who contribute especially to the advancement of Linux and open
Well, you've really done it this time. You put out a fantastic January 2006 issue. Articles on video production and DVD authoring—loved them. The new Work the Shell series by Dave Taylor—love it. The focus on hardware projects—love it and want more of it.
You could really top off the gEDA article with a follow-up tutorial on using the gEDA suite to lay out, simulate and produce an actual circuit complete with a home-etched PCB. Maybe a scrolling LCD display or motor driver. The following month could cover the Linux driver to interface to the project.
I'd also like to see how to snoop the PCI bus to figure out how that darn Windows-only driver talks to the PCI card. I'll admit, I have a vested interest with that one. Also, you've run lots of firewall configuration stories, but nobody has ever talked about firewall rules. I'd like to see a listing and discussion of rules for protecting oneself from the various known attacks—more than just blocking incoming connection and spoofing.
Other than that, keep up the great work. I've just renewed for my ninth
In April of 2003, my mother asked me what I wanted for my birthday. I told her I would like a subscription to Linux Journal, so she got me one for two years. In April 2005, she asked me again, and I asked her to renew my subscription. She called the subscription line for Linux Magazine and asked them to renew my subscription (which didn't exist). They went along with her anyway, and now I have a two-year subscription to LM. The other magazine doesn't even hold a candle to Linux Journal. I finally decided to get another subscription after seeing the issue on MythTV in my local Fry's store. I thought I would hurry and sign up so I could get the issue. I used the email@example.com method, so I could guarantee the December 2005 issue. I spoke with Jon-Mark about this, and after some negotiating, he gave me a discount on my renewal and the December 2005 issue! I have never gotten that level of customer service from a company—EVER. His dedication to please the customer was admirable. I know he probably doesn't work directly for SSC, but you should rest easy in having SDS as your subscription partner.
I just hope my mom doesn't call Linux Magazine in April 2007!
Great article [“Monitoring Virtual Memory with vmstat” by Brian K. Tanaka, LJ, December 2005]! I had a thought that pestered me when reading it, and I thought I'd drop you folks a note about it.
I was under the impression (from the good ol' days) that paging was the manipulation of the instruction space and swapping was the manipulation of the data space.
Now, it might be that I'm showing my age, but as I remember it (I'm saying this without a full dose of coffee yet), if memory space begins to get used beyond a particular value as judged by the kernel, the first thing it'll try to do is page out unused instruction space (that is, a loaded program that is not running) but keep the data space in core (see, I'm old) until it can't hold out any longer. Then, the next thing it'll try to do, if paging doesn't give it enough space, is to swap out data to the swap device to gain more working room. (Now on Linux specifically, which uses free memory for disk buffering/caching, the order and priorities of who gets tossed out first may very well change.)
If I really stretch my brain, I seem to remember that paging out in some cases was a lightweight process of basically marking a page as no longer part of a program's instruction space, and putting it back on the “available” queue. (I'm simplifying here.) But when the system needs that page of instructions, it triggers a page fault to try to page in that piece of code from the disk, which is the really costly part of the process.
The granularity of how much to reclaim by paging or by swapping is a kernel parameter that seems to operate under the belief that if we're starting to swap, go for the gusto and swap lots of stuff now, as one big task to save having to do it again for a while.
There's also the question of who pays for this overhead work. Page outs are counted against the process trying to get more memory and page ins are counted against the code that needs the pages. This also applies to swapping. Again, this is based on what's left of some grey matter suffering from caffeine deficiency.
Now, unless my memory fails me...what was I saying...oh yeah, if I'm wrong, then the inevitable question is, where do paged-out pages go? To the swap space? If that's the case, then what's the difference between swap and page space?
Thanks for your time and keep on hacking!
Michael C. Tiernan
Thanks a lot for your article on Linux video editing [“Linux Video Production: the State of the Art” by Dan Sawyer, LJ, January 2006]--a very interesting read. I'm a bit surprised that you didn't address the legal complications surrounding mpeg-encoding at all—is this, in practice, no problem, especially considering that you presumably use these patented algorithms in a commercial setting? Not being a video geek, I don't know how good Theora is at present—I've heard high praise from some, but OTOH it's still very new, and not supported in any video (hardware) equipment I know of.
(My background: I am a Debian developer, and Debian's commitment to free
software means that software using patented algorithms can, in most
not be included in the Debian distribution—that's why I probably pay a
bit more attention to such legal matters than other Linux users.)
Adrian von Bidder
In the January 2006 issue, Dave Taylor's Work the Shell column “Exploring Pipes, Test and Flow Control” states, “Many modern shells have a version of the test command built in to the shell itself, considerably speeding up shell script execution. Using the [ symbol ensures you'll use the built-in version if available, but explicitly calling test means that you'll likely not have that performance enhancement when running your scripts.”
This seems intuitive, but it's incorrect. At least for the past
decade, all modern shells (bash, ksh, zsh) treat [ and test as
equivalents. In other words, if [ is a built-in, you'll find that test
is as well.
Great work with the new column Work the Shell and the vmstat article in the
December 2005 issue. I would like to see more articles for newbies like these. There
are many topics to cover, such as system startup, UNIX programming, useful
command-line tools and so on.
Not long ago, Ubuntu was just another distro hoping to make an impact. Like hundreds of others before it, Ubuntu had hoped to catch the eye of random Linux users. Maybe one day it would even become a Top 50 distro with a few hundred followers.
But now? According to distrowatch.com, Ubuntu is the
top distro on the planet! I must say I'm not
surprised. This is how Linux was meant to be. As far
as I'm concerned, Ubuntu is Linux.
The article “Wireless Home Music Broadcasting” by John
January 2006] caught my attention as it resembles a different take on a
project of my own: music in every room. Both my project and John's
utilize a large collection of ripped MP3 and OGG files on a NAS. However, in my case, the plan is to hang flat computer speakers from the wall in
each room, and plug them in to my roving PDA (a Zaurus SL-5500). The PDA
will access the NAS via an 802.11b compact Flash card. At least, that's
the theory; in practice, Opie Media Player 2 terminates with signal code
SIGSEGV when I try to play any audio file, either mounted locally or NFS.
Perhaps I'll find another MP3 player that will work; I'm not giving up.
I have recently published a nonprofit Web site
that tries to promote the use of OpenOffice.org formats by placing a banner
on Web sites and/or registering on the Web site for listing and linking
the logo. This is inspired only because I know quite a lot of people
using OpenOffice.org but using MS Office formats for general compatibility
reasons, even though you may be sometimes sharing documents with people
with access to computers with OpenOffice.org.
I think it's a good idea to have a more proactive community of users
willing to use the OpenOffice.org format more commonly, and that's why I am
writing to you in the hope that you find this idea as good as I believe.
In this case, a small note in your magazine would help much to try to develop
If you want to have a look at the site you'll find it at
|Where's That Pesky Hidden Word?||Aug 28, 2015|
|A Project to Guarantee Better Security for Open-Source Projects||Aug 27, 2015|
|Concerning Containers' Connections: on Docker Networking||Aug 26, 2015|
|My Network Go-Bag||Aug 24, 2015|
|Doing Astronomy with Python||Aug 19, 2015|
|Build a “Virtual SuperComputer” with Process Virtualization||Aug 18, 2015|
- Concerning Containers' Connections: on Docker Networking
- Problems with Ubuntu's Software Center and How Canonical Plans to Fix Them
- A Project to Guarantee Better Security for Open-Source Projects
- Where's That Pesky Hidden Word?
- Firefox Security Exploit Targets Linux Users and Web Developers
- My Network Go-Bag
- Doing Astronomy with Python
- Build a “Virtual SuperComputer” with Process Virtualization
- Three More Lessons
- diff -u: What's New in Kernel Development