Friday, July 24, 2009

antiX M8.2 has been released and I recommend it!

The announcement for the release of antiX M8.2 can be found here on the antiX forum site, and as usual, I recommend it.

Those of you who like to pretty much concoct your own system might want to give antiX BASE a try. It is very close to the size of SLAX in image size, does not come with much pre-installed, so you can pretty much rip it apart and put in what you want, whether that's something light or heavy.

For those of you who want a complete, yet lightweight system likely to work on older Pentium III and IV hardware, this is a great one to try out in the full version. I have used both versions in the past and I have already downloaded, and I am about to try out, both the BASE and FULL versions in Virtualbox OSE, then install one or both of them on my systems in place of the current system.

Thursday, July 23, 2009

Choosing a trilogy of distributions

I happen to choose a trilogy of distributions because I am looking for three different styles of systems:

1. Cutting edge
2. Lean and fast
3. Stable and relatively non volatile

For those, I personally picked:

1. sidux
2. antiX
3. SimplyMEPIS

Others may want to pick:

1. Arch Linux
3. Debian Linux

Still others may choose:

1. Gentoo Linux
2. Tiny Me
3. Slackware

What's nice is that there are choices. Notice that my choices are Debian based. That's my preference, but clearly it does not have to be everyone's preference, nor would I attempt to muscle my own interests and preferences upon others. It is nice when someone shares a few common interests, but it is also nice when others investigate other options, too. It makes for nice interactions, sharing what we have found useful in the systems we choose.

At the end of the day, each person has their own needs and interests. Some people may not want to bother with a trilogy of systems at all; others may want to go well beyond that! (Actually, I do myself, it's just that my trilogy of systems are my core systems that I come back to more than anything else).

My categorizations and descriptions of the systems are simply an attempt to stimulate conversation and encourage others to do the same. If a few people want to try out the same software, that's fine, but it is just as well when people try other things, and I always encourage that.

As far as antiX goes, what I like about it is that it is fairly light, yet it contains enough software to consider it a complete system. Other alternatives are even smaller, but for me, antiX is a great compromise between small size and flexibility. Having two versions, a BASE version, which does not contain a complete system, just a framework from which you can create your own idea of a complete system, and FULL, which offers a IceWM and Fluxbox (window manager) base with moderate sized applications, makes it ideal for me as a framework for setting up systems that I can either use at home or on the go. When on the go, such as in an Internet Cafe, antiX leaps to the top of my list. In thirty seconds, I can have a desktop up and running, and in two minutes or less, I can be computing in an Internet Cafe.

SimplyMEPIS and Debian Lenny are two of the most stable systems around. SimplyMEPIS just gets you there a bit faster because it has a nice choice of applications and some "non-free" drivers, codecs, and plug-ins to get you up and running quicker. Otherwise, they are more common than different, because they are based on 95% of the identical software.

Slackware is right in there with Debian as a fantastic, stable system, and it is a great learning environment too. The convenience factor is somewhat less with Slackware than with Debian, but there are several Slackware derivatives that get you around that, if it is an issue for you in time or in understanding. Vector Linux, Zenwalk Linux, and Absolute Linux are three good Slackware based derivatives.

Lots of guys here are gung ho on Arch. It is the premier binary distribution for those who want to tailor things exactly the way they want them. It can take longer to set up - without help - but with our setup checklists, which Dave Crouse originated, and Tim Miller and I discussed and tweaked, you can get an Arch Linux system going in twenty to thirty minutes that is just what you want.

Jester and a few others are big fans of Gentoo Linux, and you have to give the crown for the most complete source based infrastructure to the Gentoo team. Sabayan Linux, also source based, is a good alternative, as is Funtoo, but each of these are actually based on Gentoo, so credit should go to the Gentoo effort.

Linux from Scratch is the ultimate, you build it from whatever you want kind of source based system. You do not hear about this one all that often. You truly pick every component, every aspect of the system from where you want and how you want to do it. Linux from Scratch is based on a book series, which is regularly updated. I follow it from time to time, but it is not for me. I know how to do enough of the stuff, and if ever pushed to do it, I could, but I have nothing to prove there, and talk about a time sink, that is the ultimate. Great for hermits who have bleached white skin! Smile

Corporate systems like Red Hat Enterprise Linux are very good, and the test versions of that are Fedora based, and the free version is based on Red Hat Enterprise Linux source code, called CentOS - not related to Red Hat - just uses Red Hat source code.

Ubuntu and its derivatives are for those who want a regular release instead of a release when ready version like Debian, upon which it is based. Mint is a more stable (and to some, more visually appealing) derivative of Ubuntu. They both have many of the attributes of a Debian system, but their main attraction is that they are easier to set up and they are widely available, so it is easy to find information about them. Mint makes things easier because it includes more non-free stuff, similar to the value add that MEPIS provides to core Debian code.

That's a quick wrap. Of course, there are numerous other systems. I have not even touched upon Mandriva or any of its derivatives or on SUSE or its derivatives. Clearly they are very viable as well, as are many other distributions. What they all share in common are the core kernel and the primary GNU utilities. They vary in which precise versions of code are used and in which devices and configurations are applied by default.

Tuesday, July 21, 2009

sidux 2009.02, waiting for antiX M8.2

I was going to wait for both sidux 2009.02 and antiX M8.2 to come out, since there were test versions of both out earlier this month. antiX has been in public testing longer, but anticapitalista wants to get a few things just right before releasing. Meanwhile, the public test versions, including the Pre Final 1 for antiX M8.2 look very good.

There was some sad news on the sidux front, because there were unfortunate conflicts within the development ranks that caused quite a stir, and I was accidentally involved in some of it. I hope that the fences have been mended, and peace is being restored in the community, because for the past two years, sidux has been my favorite distribution. It is cutting edge, fast, and it has done a good job of "taming" Debian Sid. The test team for sidux does a great job looking for defects and packaging mismatches, and they advise on which packages to put on hold.

The smxi project, which provides a tool that works with these package holds, was once an integral part of the unofficial, yet very common extended software that was often added onto sidux. Unfortunately, that has been an area that has become increasingly contentious, and, as it turns out, completely unknown to me, that was the major, and probably primary, source of the fireworks that recently touched off on sidux irc channels and in the sidux forums.

The unfortunate fact is that sometimes these things have to happen as the directions of projects and individuals take on differing priorities. At one time, Harold Hope, also known to many as h2 on sidux forums and as Tech Admin over at Tech Patterns, was considered a "developer" at, something that the sidux team recently denied. Harold has been finding himself at odds with the objectives of the sidux project for some time now, but it recently came to a head.

Things were said and done that I hope can one day be mended, but suffice it to say that sidux and smxi have absolutely no ties any more, and the sidux community and h2 have parted ways.

We've seen an awful lot of this kind of thing happen this year - sparks flying at Puppy Linux, where a project manager resigned after leading only one release, a number of developers leaving the PCLinuxOS project after poor communication during the absence of the founder, Texstar, then the sidux conflict, and it seems I am forgetting one or two others.

h2, I wish you well in what you do; you know how fond I am of smxi. sidux developers, know that I am also fond of the work that you do. May each project find new energy and direction as you take different paths.